FOR A HEATING SYSTEM AND CONTROL METHOD

The present invention relates to a radiator or convector and to a smart thermostatic valve for use with the radiator or convector, to a method of operating the radiator or convector and software for carrying out such a method.

Background

The current generation of electronic TRVs (Thermostatic Radiator Valve) are equipped with a temperature sensor and control the radiator based on a fixed control scheme and the measured temperature. The proportional factor of the controller is mostly l°K or 2°K which means that the valve is completely opened when the measured room temperature is l°K or 2°K lower than the desired temperature. Some advanced types are equipped with additional features like anti-frost, open window detection, party mode, ..., or are able to communicate with other valves in the same room (e.g. via Bluetooth).

A single pipe loop central heating system has a single loop of pipework running from a boiler and returning to the boiler. The connections of each radiator are made to the same pipe. The radiators are located above the pipework and as the heated water from the boiler is fed along the pipe, if the valve on the radiator is open, natural convection causes the heated water to rise into the radiator displacing cooler water back into the pipe. A major disadvantage of this arrangement is that the first radiator gets hotter than the second one and the last radiator will be considerable cooler as the water will have given up most of its heat to the previous radiators along the pipe run. The alternative is to keep the temperature high in each radiator, e.g. by a high flow, but then the return pipe temperature is high and the system is inefficient and is now rarely used.

In a more efficient system, heated water from the boiler is fed to one side of every radiator (the supply pipe) while another end of each radiator is connected to a separate common return pipe. This means that the temperature of the water entering all the radiators is more or less the same and the water in the return pipe does not enter another radiator.

A microbore heating system uses normal pipework for the supply from the boiler to manifolds and from manifolds back to the boiler on the return side. From each manifold, small diameter pipework is connected to a number of radiators. Both supply and return microbore pipes can be connected to the same end of each radiator or can be connected to separate ends of the radiators. The small diameter pipes contain less water so less heat is lost along each pipe.

Some advanced types of heating systems are equipped with additional features like anti frost, open window detection, party mode, ..., or are able to communicate with other valves in the same room (e.g. via Bluetooth). One known system is for example the HeatSave system described at the website (http://www.heat-save.com/how-does-it-work/). In such system a central micro-processor hub gets inputs from different sensors, including temperature - / light-level - / humidity- / occupancy sensors to control the working of the boiler in matching its heat output to the average heat transferred by the radiators. In matching this heating balance, the HeatSave system measures the supply and return temperature and only fires the boiler when differential temperatures exceed a preset limit.

HeatSave is said to have a state-of-the-art microprocessor connected to a number of temperature sensors measuring the external outside temperature, room, hot water tank and heating flow and return temperatures. HeatSave has a four layer, state-of-the-art, surface mount microprocessor controller with eight temperature/light level/humidity analogue sensor inputs, eight occupancy sensor/switch digital inputs and eight main relay outputs.

HeatSave is said to perform:

•Accurately switching on/off of heating systems using a heating diary.

• Use a unique“heat loss algorithm” comprising multiple quadratic equations for each heating zone (max. of fourty zones per controller) to switch the heating on as late as possible in order to still reach a required temperature by the required time.

• Not overheating a building by applying adaptive outside temperature compensation using twelve different compensation curves that can be simply changed by the user.

• Using occupancy sensors to reduce/switch off heating when nobody is around.

• Intelligently using the residual heat of a boiler rather than always switching it on.

• Intelligently applying frost protection, using pumps to distribute residual heat.

• Flow/retum balancing so that the heating is matched to the boilers heat exchanger.

Within HeatSave each heating zone is managed independently; a heating zone being defined as area within a building, a hot water cylinder(s) or even the whole building itself. A problem can occur when the outside winter temperature at say 6am on a winter’s morning can be -5°C on one day and +l2°C the next. HeatSave has a mathematical algorithm comprising a series of quadratic equations, which are said to have been designed to calculate and learn the heat loss profile of any heating zone. So long as the plumbing allows the control of the flow of heat into the zone, HeatSave is said to be able to calculate the heat loss profile of the zone. Armed with this information, the HeatSave controller is said to work out the latest time to switch on the heating for any given conditions, so as to reach the temperature that has been set for a given time. In practice, HeatSave will start the boiler later on warm days and/or if the zone has a large amount of residual heat, and starts the boiler earlier if the building is cold or it is cold outside. The reverse happens at the end of a heating period. Here, heating of a zone is switched off early if the residual heat within the zone can be maintained at the required temperature until the end of the heating period.

If the difference between the flow and return temperatures is insufficient, the boiler will still continue to try and heat the primary circuit at a faster rate than it can be absorbed. Boilers are unintelligent, so that this will continue to happen until either the boiler thermostat temporarily reaches the required temperature or, alternatively, the room thermostat reaches its upper limit. Residual heat within the radiators will raise the temperature further, thereby creating an additional temperature overshoot. Under these conditions, fuel is wasted by attempting to force heat into the building faster than it can be absorbed.

HeatSave is said to counteract the above problem because it measures the flow and return water temperature and only fires the boiler when the differential temperature exceeds a preset limit. This action matches the boiler heat output to the average heat transferred by the radiators. The boiler pump is kept switched on throughout to dissipate heat into the building. HeatSave constantly monitors the flow and return temperature, adjusting the boiler temperature to keep the boiler running efficiently.

A problem with such a system is that it interferes with the working of the boiler and, accordingly, requires installation of additional sensors on the boiler, like temperature sensors on the supply and return lines of the boiler. In addition, active interference with the boiler is not always possible with older heating systems. It would, accordingly, be desirable to have an alternative method of heating regulation which can easily be implemented in existing systems.

For larger systems such as district heating systems, increasing the efficiency of energy conversion and using sustainable sources can benefit from a low return temperature, e.g. more condensation in gas boilers (see Figure 3), higher Coefficient Of Performance (COP) of a heat pump, higher thermal efficiency of a Combined Heat and Power (CHP) system, higher efficiency of a solar collector, more heat extraction per unit brine in a geothermal context.

WO2012/116322 discloses a radiator surrounded by and housed in an insulating cover. A vent 115 can be positioned at the top of the insulating cover 110 to heat the surrounding space when needed. The radiator cover 110 can include a passive louver that can operate as a mechanical temperature control without the need for an electronic actuator. The temperature of the radiator can result in a vent in the radiator cover 110 to be opened or closed automatically in response to the temperature of the room.

Temperature sensing can be done using one or more sensors placed in specific locations around the space 200, which can be for example exterior windows, locations in the space 200 furthest from the radiator, or a living or sleeping areas of the space 200. In operation, if the room is cold, as indicated by sensor 220, and the radiator 235 is hot, as indicated by sensor 230, the microprocessor 215 can control a vent actuator 240 in the radiator cover 245 such that warm air exits the radiator cover 245 through a vent therein.

In step 305 a controller can measure the temperature in a space or room. In step 310 a controller can measure the temperature of a radiator located in the space or room. In step 320 the controller can check if the room or space temperature is below a minimum set point temperature.

• If the room is not below the minimum set point temperature, in step 325 the controller can check if the room temperature is above a maximum set point temperature. If, in step 325 the room temperature is above the maximum set point, then in step 330 the system can close a vent in the insulating radiator cover, thereby preventing additional heat from entering the space. In step 340 the controller can transmit an over-heat notification to a building server indicating that additional heating is not currently needed in the space.

• If, in step 320 if the room temperature is below a minimum set point, in step 345 the controller can open the vent above the radiator, thereby allowing warm air to circulate around the radiator and warm the room. If, in step 350, the controller determines that the temperature in the room has stayed below the minimum set point temperature for the predetermined period of time, an under heat notification can be set to a building server indicating that additional heating capacity is needed in the room.

The system described by WO2012/116322 has the following disadvantages

1. a radiator is mounted in a closed volume and the air in the closed volume is heated.

The closed volume is opened to permit hot air to escape, in response to a low temperature measured by a sensor in a room surrounding the closed volume. WO2012/116322 therefore discloses a system for indirectly heating a room, by hot air stored in volume which surrounds a radiator.

2. WO2012/116322 does not disclose a radiator comprising means to reduce the temperature of the liquid exiting from the radiator. This is of particular importance if the radiator forms part of a heating system which also contains a boiler, and if maximizing of the efficiency is envisaged.

Summary of the invention

In one aspect a radiator or convector and a controllable radiator or convector valve are provided, the radiator or convector having an inlet for hot liquid and an outlet for exiting liquid from the radiator or convector, the radiator or convector valve comprising a valve driver operable to controllably drive a valve member between an open position and a closed position; and a plurality of sensors, one at least located in a zone or room to be heated by the radiator or convector wherein valve control means are in operative communication with the plurality of sensors and are operable in response thereto to drive the valve member to a more open or closed position if the sensed temperature exceeds or is lower that a first threshold temperature, further comprising means to reduce temperature of the exiting liquid.

The means to reduce temperature of the exiting liquid can be a pulsed operation of the controllable radiator or convector valve, or one or more fans directing air against the radiator or convector.

The radiator or convector can have a plurality of plates and the means to reduce temperature of the exiting liquid is a means to produce a labyrinth flow through the radiator or convector.

One of the plurality of sensors measures the temperature of return pipe to a boiler, and the controllable radiator or convector valve is adapted to activate one of the means to reduce temperature of the exiting liquid on receipt of a signal derived from an output of the sensor measuring the temperature of return pipe.

Means can be provided for reacting to heat sources located in a zone or room or external to a zone or room,

The means for reacting to heat sources located in a zone or room or external to a zone or room is an inflatable airbag within the radiator or electric current sensors of electric currents entering a room or zone through cables for the driving of electrical equipment, or a sensor of noise level generated by persons in the room, or from the output of an infra-red camera, placed above places for persons e.g. placed on the ceiling, in the centre of a room or zone pointing downwards. The occupancy of infra-red sources in the images from the camera can be related to the number of persons in the room.

Means for executing a learning algorithm (i.e. a means or method of machine learning) can also be provided. A relationship between the occupancy of heat sources in images from the camera and heat input into a room or zone can be learnt by the means for executing a learning algorithm.

In one aspect, embodiments of the present invention include sensors in order to measure the internal and external heat gain to control better the room temperature. In embodiments of the present invention, sensors can be included in a radiator or convector thermostat or a thermostatic radiator valve. Including the sensors in the thermostatic radiator valve has the advantage that control of the temperature of the returning liquid from the radiator or convector is better. The return temperature has a direct impact on the efficiency of the heat conversion system such as a gas boiler, combined heat and power, heat pump, deep geothermal plant, etc. A further advantage is also that energy consumption can be reduced as the system reacts faster to changes in external and internal heat gain. The present invention provides a radiator or convector and a controllable radiator or convector valve, the radiator or convector having an inlet for hot liquid and an outlet for exiting liquid from the radiator or convector, the radiator valve comprising a valve driver operable to controllably drive a valve member between an open position and a closed position; and a plurality of sensors at least located in a zone or room to be heated by the radiator or convector wherein valve control means are in communication with the plurality of sensors and are operable in response thereto to drive the valve member to a more open or closed position if the sensed temperature exceeds or is lower than a first threshold temperature, further comprising means to reduce temperature of the exiting liquid. The sensors may be provided in a sensor stack. Each valve control means can be in communication with at least three different sensors. The radiator or convector does not need to be housed in an insulating cover. For example, the present invention does not foresee an insulating confinement and does not control a hot air release within such a confinement. Instead, in embodiments of the present invention valve settings of the radiator or convector are controlled. Embodiments of the present invention rely on a plurality of sensors instead of just one temperature sensor.

In a preferred embodiment, the present invention relies on a mathematical representation (grey box or black box model) and the plurality of sensor readings to find a setpoint through predictive assessment. Embodiments of the present invention aim to minimize the return temperature instead of just trying to conserve heat.

In embodiments of the present invention, a radiator or convector does not need to be positioned in a closed volume. Instead, direct heating of the room or zone is envisaged. Embodiments of the present invention comprise means to reduce the temperature of the exiting liquid from the radiator or convector, i.e. the temperature of the liquid leaving the radiator or convector. This is advantageous if the radiator or convector forms part of a heating system which also contains a boiler. An advantage is the maximizing of the efficiency. Means for executing a learning algorithm can be provided. Using a valve in a network with other valves is advantageous when these valves are Smart TV (Smart TRV or Smart TCV). The Smart TV can be in communication with a plurality of sensors connected to embedded or central computing equipment (IT system). In one aspect, the computing equipment contains a first or higher order software representation of generic heating zones and their interactions. The software representation of one or more heating zones can predict the effect of environmental changes as detected by the plurality of sensors and the effect of given settings on the valves or the network of valves with a given supply temperature. A learning algorithm records the values from the plurality of sensors over time and applies parameter estimation to tune the models of the heating zones to continually improve the accuracy of the software representation. An optimized control algorithm chooses valve settings according to their respective predicted outcome and according to the predicted influence of environmental changes detected by the plurality of sensors. The prediction-based algorithm optimizes the return temperature and prevents oscillations of the controlled system.

In another aspect, an IT system includes algorithms for computational statistics / machine learning - based on regression or (deep) neural networks - able to infer correlations between recorded values for valve settings, supply temperatures, and readings from the plurality of sensors as well as carrying out control actions. The GG system can be capable of pre-emptively executing control actions to optimize the return temperature and avoid space temperature oscillations.

In another aspect of the present invention, a controllable radiator or convector valve is provided for fitting to a radiator or convector having an inlet for hot liquid and an outlet for exiting liquid from the radiator or convector. The radiator or convector valve can comprise a valve driver operable to controllably drive a valve member between an open position and a closed position. Valve control means are adapted to be in communication with a plurality of sensors at least located in a zone or room to be heated by the radiator or convector and the valve control means are adapted to receive signals from the plurality of sensors and are operable in response thereto to drive the valve member to a more open or closed position if the sensed temperature exceeds or is lower that a first threshold temperature, whereby the valve control means can be adapted to activate means to reduce temperature of the exiting liquid.

In another aspect of the present invention, a method of operating a radiator or convector and a controllable convector radiator valve is provided, the radiator or convector having an inlet for hot liquid and an outlet for exiting liquid from the radiator or convector, the radiator valve comprising a valve driver, and a plurality of sensors at least located in a zone or room to be heated by the radiator or convector, the method comprising

controllably driving a valve member between an open position and a closed position; and wherein valve control means communicate with the plurality of sensors and operate in response thereto to drive the valve member to a more open or closed position if the sensed temperature exceeds or is lower that a first threshold temperature, further comprising activating means to reduce temperature of the exiting liquid.

The means to reduce temperature of the exiting liquid is a method step of pulsed operation of the controllable radiator or convector valve or activating one or more fans directing air against the radiator or convector.

The radiator or convector can have a plurality of plates and the means to reduce temperature of the exiting liquid is producing a labyrinth flow through the radiator or convector.

The method can include reacting to heat sources located in the zone or room or external to a zone or room.

The step of reacting to heat sources located in a zone or room or external to a zone or room comprises any of inflating an airbag within the radiator or convector, measuring electric currents entering a room or zone through cables for the driving of electrical equipment and measuring noise level generated by persons in the room.

Brief Description of the figures Figures la and lb show a number of rooms with radiators or convectors and sensors whereby Figure lb is a less extensive embodiment of the present invention.

Figure 2 shows a heating system schematically including a boiler as used in embodiments of the present invention.

Figure 3 shows the relationship between boiler efficiency and boiler return temperature. Figure 4 shows heat output of a radiator versus flow rate.

Figure 5 shows schematically a Smart TV (Smart TRV or Smart TCV) in accordance with an embodiment of the present invention.

Figure 6 shows schematically an RC equivalent circuit for use with Smart TV (Smart TRV or Smart TCV) in accordance with an embodiment of the present invention.

Abbreviations and definitions

“District heating” will be referred to as DH

A“Smart Thermostatic Radiator or Convector Valve” will be referred to as a Smart Thermostatic Valve (Smart TV (Smart TRV or Smart TCV)). Any reference to a“Smart Thermostatic Valve” or a“Smart TV (Smart TRV or Smart TCV)” discloses both a radiator valve (Smart TRV) and/or a convector valve (Smart TCV). The term“smart” refers to local electronic digital processing power such as provided by a microprocessor, a microcontroller, ASIC etc.

Where reference is made to“a plurality of sensors at least located in a zone or room", the meaning is that sensors are at least in one room but they could be in several rooms. For example a large room or an air curtain may divide up a room into heating zones or several rooms may be grouped together in one heating zone. In a heating zone there does not need to be a one-to-one relationship between a zone and walls.

Where reference is made to "labyrinth flow" this means flow through a labyrinth. This will reduce the flow rate. For example, U.S. Pat. No. 4,060,200 to Mehouder, the disclosure of which is incorporated herein by reference, describes a labyrinth channel comprising two opposing arrays of equally spaced baffle“teeth” that extend out towards each other from opposite walls of the channel. Each tooth has a cross section perpendicular to the wall substantially in the shape of a truncated isosceles triangle, i.e. the apex of the triangle is“cut off’. The arrays of baffle teeth are substantially mirror images of each other but are displaced relative to each other along the channel by half a repeat period of the baffle teeth, i.e. by half the distance between adjacent baffle teeth. A tooth in one baffle array therefore faces a point in a space, substantially half way between adjacent baffle teeth in the other array. The tips of two adjacent baffle teeth in one baffle array in the labyrinth and the tip of the tooth in the opposing baffle array that faces the space formed by the adjacent baffle teeth, are substantially coplanar.

Another example is disclosed in US Patent Publication 2003/0150940, the disclosure of which is incorporated herein by reference, which shows a labyrinth channel comprising two opposing rows of equally spaced baffle“fingers” that extend out towards each other from opposite walls of the channel. The tips of the finger baffles are terraced so that tips of the fingers decrease step-wise in size with height of the fingers off the floor of the channel. The labyrinth channel does not comprise a through-flow channel and tips of fingers in each row extend into spaces between fingers of the other row, i.e. the fingers mesh. All the fingers appear to be tilted at a same angle towards a downstream direction of water flow.

Another example is disclosed in PCT publication WO 00/01219, the disclosure of which is incorporated herein by reference, which describes a“sawblade-shaped zig-zagging” pattern comprised in a fluid flow regulatory channel of an irrigation pipe. The zig-zagging pattern is embossed on a relatively thin web of flexible plastic material. The web is folded over so that longitudinal edges of the web overlap and regions of the overlapping edges are welded to form the irrigation pipe and regulatory channel comprising the zig-zagging pattern.

Where reference is made to "a zone or room to be heated", the skilled person understands that a building may be divided in different ways. For example an air curtain may divide up a room into heating zones or several rooms may be grouped together in one heating zone. In a heating zone there does not need to be a one-to-one relationship between a zone and walls.

Where reference is made to "valve control means are in communication with a plurality of sensors" this means that a control means (i.e. a controller) can communicate with the sensors and obtain a benefit from the different outputs from the sensors. Hence, the outputs of the sensors can be transmitted to the valve control means and received and processed there. Hence, the valve control means can be in operative communication with a plurality of sensors.

Where reference is made to "return pipe of a boiler” this refers to a return pipe in a heating system that returns the circulating fluid to the boiler to be re-heated as shown in Figure 2.

“Radiator” refers to any device that provides heat by some radiation. A radiator may have a combination of radiation and convection, for example. The radiator will be heated by circulating liquid such as water.

“Convector”” refers to any device that provides heat by some convection. A convector may have a combination of radiation and convection for example. The convector will be heated by circulating liquid such as water. A HVAC is classified as an active convector and is typically cooled by a liquid or gaseous coolant. Underfloor heating is considered to be a convector, in this case a passive convector.

A’’sensor” is a device for detecting or sensing a physical event or change in its environment and can output information concerning the physical event or change to other electronics, such as a Smart TV (Smart TRV or Smart TCV) or to a computer which communicates with the Smart TV (Smart TRV or Smart TCV). The sensor can be a camera such as an infra-red camera.

Description of the illustrative embodiments

In buildings connected to a District Heating (DH) system equipped with radiators, or convectors increased efficiency requires operation at high supply temperatures and limited flow, requiring an advanced control strategy capable of dealing with the varying circumstances. A critical component is the thermostatic radiator valve that must be adapted so as to guarantee these low (primary side) return temperatures.

Instead of active interference with the working of the boiler, the efficiency of the heating system can be increased by assuring a high temperature difference between the supply and return temperature by controlling and maintaining the return temperature as low as possible, whilst maintaining a high supply temperature. In order to achieve a low return temperature it is important to keep the flow in the radiator or convector as low as possible. That means that also in periods with low heat demand it is important to keep the supply temperature to the radiator or convector high, but then with a low flow through the radiator or convector.

Maximizing this temperature difference with limited flow, results in longer cooling periods in which the radiator or convector can dissipate heat and ultimately a higher system efficiency. In the case of a district heating network it is also desirable to have a low return temperature to be sent back via the return pipe to the heating network. A high/maximum supply temperature guarantees a high/maximum temperature difference across the radiator or convector but as a consequence the flow needs to be (very) low.

However, under low heat demand, these high supply temperatures can cause problems: any faults in the control of the Thermostatic Radiator or Convector Valve (TRV) result in high fluctuations in the average radiator or convector temperature and therefore room temperature.

Embodiments of the present invention can provide one, some or all of the following:

• Low flow operation: reduce even small deviations that have a high impact

• Taking into account other internal and/or external supplies of heat which could be of the similar or higher as the radiator or convector power and this can potentially have a high impact on the room temperature if not taken into account

Reducing impact of the supply and return piping itself, both of which dissipate heat as well.

• Allowing different types of radiator or convector to be used such as casted radiators with high thermal mass compared to plate radiators with a lower weight/power ratio.

• Implementing the TRV or TCV mechanically for low flow rates. This is thought to result in‘pumping behavior’ of the TRV OR TCV; a rapid open/closed operation increases chances of faults/failure over time.

In one aspect, embodiments of the present invention include sensors in order to measure the internal and external heat gain to control better the room temperature. In embodiments of the present invention sensors can be included in a radiator or convector thermostat or a thermostatic radiator or convector valve. Including the sensors in the thermostatic radiator or convector valve has the advantage that control of the temperature of the returning liquid from the radiator or convector is better. The return temperature has a direct impact on the efficiency of the heat conversion system such as a gas boiler, combined heat and power, heat pump, deep geothermal plant, etc. A further advantage is that the energy consumption can be reduced as the system reacts faster to changes in external and internal heat gain. Alternatively or additionally, the sensors or one or more sensors may be not located in the TRV OR TCV but may be located in a stack or individually in the room or zone or on other parts such as on the ceiling or the walls of the rooms or zones.

Embodiments of the present invention provide a radiator or convector and a controllable radiator or convector valve, the radiator or convector having an inlet for hot liquid and an outlet for exiting liquid from the radiator or convector. The radiator valve comprises a valve driver operable to controllably drive a valve member between an open position and a closed position. A plurality of sensors are at least located in a zone or room to be heated by the radiator or convector wherein valve control means are in communication with the plurality of sensors and are operable in response thereto to drive the valve member to a more open or closed position if the sensed temperature exceeds or is lower than a first threshold temperature. In addition, means to reduce temperature of the exiting liquid are provided.

Embodiments of the present invention do not need to be housed in an insulating cover. For example, the present invention does not foresee an insulating confinement and does not control a hot air release within such a confinement. Instead, in embodiments of the present invention valve settings of the radiator or convector are controlled. Embodiments of the present invention rely on a plurality of sensors instead of just a temperature sensor.

In a preferred embodiment, the present invention relies on a mathematical representation (grey box or black box model) and the plurality of sensor readings to find a setpoint through predictive assessment. Embodiments of the present invention aim to minimize the return temperature instead of just trying to conserve heat.

In embodiments of the present invention a radiator or convector does not need to be positioned in a closed volume. Instead, direct heating of the room or zone is envisaged. Embodiments of the present invention comprise means to reduce the temperature of the exiting liquid from the radiator or convector, i.e. the temperature of the liquid leaving the radiator. This is advantageous if the radiator or convector forms part of a heating system which also contains a boiler, and an advantage is the maximizing of the efficiency of the heating system.

Embodiments of the present invention can include means for executing a learning algorithm, for example using a valve being in a network with other valves connected to appropriate embedded or central computing equipment. In one aspect, the computing equipment contains a first or higher order software representation of generic heating zones and their interactions. The software representation of the heating zones can predict the effect of environmental changes as detected by the plurality of sensors and the effect of given settings on the valves or the network of valves with a given supply temperature. A learning algorithm records the values from the plurality of sensors over time and applies parameter estimation to tune the models of the heating zones to continually improve the accuracy of the software representation. An optimized control algorithm chooses valve settings according to their respective predicted outcome and according to the predicted influence of environmental changes detected by the plurality of sensors. The prediction- based algorithm optimizes the return temperature and prevents oscillations of the controlled system. In another aspect, an IT system includes algorithms for computational statistics / machine learning - based on regression or (deep) neural networks - able to infer correlations between recorded values for valve settings, supply temperatures, and readings from the plurality of sensors and control actions, equally capable of pre-emptively executing control actions to optimize the return temperature and avoid space temperature oscillations.

It is not self-evident to control the room temperature by regulating the flow through the radiator or convector. Such regulation in accordance with embodiments of the present invention can make use of minute control of the flow, and anticipation of the system to environmental parameters like changing external conditions (e.g. solar radiation in the room) or changing internal heat sources (e.g. number of persons in the room, heat production from electrical equipment, other heating systems like electrical heaters).

In accordance with embodiments of the present invention one or more Smart thermostatic valves (e.g. Smart TRV or Smart TCV) can reduce the return temperature in heating networks. A building can include a heating network such as rooms or zones provided with heating such as radiators, or convectors or underfloor heating or HVAC all of which are heating systems that can be used with embodiments of the present invention. To reduce the return temperature in the heating network with radiators or convectors a high supply temperature is necessary also even if the heat demand is low.

Current electronic radiator valves might not provide a stable room temperature if the supply temperature is high and the heat demand is low (i.e. compared to the designed power) which would result in operating the valve at a low flow rate. Whereas the current electronic radiator valves have only a temperature sensor and react only if there is a difference between the measured temperature and the set-point temperature, embodiments of the present invention provide Smart TVs (Smart TRV or Smart TCV) (e.g. Smart TRV or Smart TCV). Smart TVs (Smart TRV or Smart TCV) address one of the problems with conventional valves, namely that when the valve is opening, the flow through the radiator or convector is higher but there is some delay before a temperature rise is detected. Also, if the supply temperature is high, the radiator or convector is filled with water at a high temperature and even after closing the valve the radiator or convector is still emitting heat leading to a room temperature that is higher compared to the designed temperature. In order to better control the temperature in a room, embodiments of the present invention make use of extra information comprising not only the temperature of the room, but also internal and external heat gains and their changing values like, for example, one, some or all of :

Solar radiation in the room which not only depends on the sun shining but also on the orientation of the windows in the room, the use of shadings, the orientation of the sun, the type of glazing, etc.

Internal heat gain by persons in the room: depending on the activity a person is a heat source (both sensible and latent). Thermal power is around 100 W depending on the activity.

Internal heat gain by appliances like laptops, radios, televisions, cooking devices, etc.

Embodiments of the present invention make use of and, if necessary, gather more information of the internal and external heat gain, and the control of the temperature in the room or zone can be carried out based on a predictive model and, hence, without the need to wait until there are detectable changes in room tor zone temperature. For example, if a lower room temperature is measured compared to a set-point, but it is detected by one or more sensors that people are entering the room and are starting up electronic equipment such as computers, laptops, printers, projectors, etc., the predictive controller can open the radiator valve less than if you have only the room temperature available, avoiding to have overheating of the room.

To provide extensive and exhaustive direct links between the measured values from sensors and the power of external and internal heat gain would require an expensive installation. Instead, embodiments of the present invention include a predictive model. Preferably, the predictive model has self-learning capabilities. Providing self-learning to a predictive model such as a lumped circuit equivalent model of the one or more heating zones or rooms means that, in the beginning, the smart thermostatic radiator or convector valve leams the effect of each input on the room or one temperature. Based on this learning sequence, the future room temperature control can be done more accurately via a more reliable prediction. Electronic radiator valves can also be used for convectors; i.e. device that emit most heat via convection, with no or less radiation heat transfer.

Embodiments of the present invention can also work with HVAC of buildings. Independent of whether the HVAC is heating or cooling a zone or room, embodiments of the present invention can function also for heating or cooling a room or zone. In this case, more detailed information about the internal and external heat gain can provide a better prediction of the heat energy required.

In most newly built offices the outside temperature is less determining the heating and cooling load. More important can be the internal and external heat gains, such as how many people are in the room or zone, how many electronic devices such as computers or laptops etc. are running, how many lights are on, what is the effect sunlight shining in the room?

Adding extra sensors to a Smart Thermostat or extra sensors being provided to work with a Smart Thermostat in each room or zone, which sensors sense directly or indirectly internal and external heat gain, can also increase the efficiency of heating and cooling the building.

For example, if one or more sensors detect that people are coming into the room (e.g. by using radar and/or infra-red surveillance sensors), the heating power to the room can be reduced in order to avoid overheating by making use of a predictive model. For example, it can be predicted whether it is not necessary to heat up the room before people are entering the room or the room could be cooled before people are entering in order to avoid uncomfortable temperatures.

In accordance with embodiments of the present invention sensors can be included in the Smart thermostatic valve. These sensors would be in addition to a conventional temperature sensor and/or humidity sensor included in the thermostat. The room Smart thermostatic valve can be extended with sensors related to the internal and external heat gain.

Embodiments of the present invention can be used with floor heating or any other thermally activated building elements.

Extra sensors, as used in embodiments of the present invention, can be used in dwellings / buildings with floor heating or TABS (thermally activated building elements). As the floor or ceiling is used as heating and cooling element and these elements have a high thermal inertia, the reaction time is very slow. Hence, if the supply temperature for the heating system is increased, the effect can only be seen after a while. The operation temperatures for heating can be low to very low (e.g. 30°C or 25°C) and the operation temperatures for cooling can be high (e.g. l6°C).

For controlling these slow systems, especially when used in well insulated building, the outside temperature is not so important for determining the heating / cooling load. More important can be internal and external heat gains. If the power of the heat gains is increasing, it is only seen after a while in the room temperature. Embodiments of the present invention preferably add extra sensors to measure indirectly the heat gains and their output is supplied to the predictive model of one or more rooms or zones and are used to predict more accurate expected heating or cooling demand.

Embodiments of the present invention can be used for thermal use of building mass providing a flexibility indicator.

If it is desired to integrate more RES (renewable energy sources) in the grid, more flexibility is needed on the demand side. Heat pumps, combined heat and power, chillers, etc. can deliver a variable output of heat energy, i.e. can provide“flexibility”. Thermal energy storage can be included to decouple heat/cold delivery and heat/cold demand. The building mass can be used as a thermal energy storage system. Thermal energy can be stored in walls, floor, ceiling, furniture during a certain period. In order to control the heat of a heat pump or similar towards availability of RES it is important to know how much energy could be stored in de building mass. Current systems use a simplified model of the room together with a temperature sensor of the room. Internal and external heat gains can only be detected indirectly via the temperature measurement. Embodiments of the present invention make use of additional sensors to measure, directly or indirectly, internal and external heat gain and the output of these sensors can be used in and can improve the active control of the heat pump or similar with thermal energy storage via the building mass. For example, if at a certain moment people are entering a room or zone, the available flexibility for the heat pump or similar will be reduced, due to the fact that extra‘heat’ is brought into the room. If this information is available, e.g. derived from the output of extra sensors, it can be used for a more optimal control.

Embodiments of the present invention find useful application in office buildings, hotels, hospitals, houses for elderly people, schools, rooms of individual dwellings, etc.

Figure la shows a number of rooms R1-R4 in a building. Each room has at least one radiator Ral to Ra4. Each wall, ceiling, floor of a room has a transfer function (Tl ....T12), for flow of heat energy across the wall, ceiling or floor depending on the temperature drop across the wall, ceiling and floor and the thermal resistance thereof. Internal room temperature sensors (Templ-4), radiator surface temperature sensors (Temp 5-8) and external temperature sensors (Temp 9). Incident solar energy sensors (Sl, S2, S3, S4) can also be provided. Boiler supply temperature sensor (Temp 11) and return pipe temperature sensor (Temp 10) can also be provided. The rooms or the building may also include other sensors or detectors such as presence detectors to detect presence of persons, e.g. by detecting movement such as with infrared or radar or infrared and radar motion detectors. Wind speed detectors may be located on the building which can be used to estimate wind chill.

All the sensors and/or detectors can have their outputs linked together by a sensor network (not shown in detail) which can be a wireless network, e.g. a CAN (controller area network). The network may include a main controller (not shown and optional) and each radiator Ral to Ra4 of Figure la has its own Smart TV(Smart TRV or Smart TCV) Cl to C4, each of which has local intelligence and can receive inputs from sensors or detectors in the same room or from other rooms or from outside. All controllers can be connected into the sensor network and can be controlled locally and/or optionally remotely from the main controller.

One aspect the present invention comprises the Smart TV (Smart TRV or Smart TCV) which is capable of managing a local radiator or convector system e.g.in one room or zone such that a low, a lower or a lowest return temperature (Temp 11) in the return pipe 8 is obtained under some or all conditions while maintaining low high heat demand or on the other hand high heat demand. Embodiments of the present invention can take into account user behaviour and occupancy, changing external conditions (e.g. solar radiation in the room) or changing internal gains (e.g. persons in the room, heat production from electrical equipment), etc. The Smart TV (Smart TRV or Smart TCV) according to embodiments of the present invention can communicate with additional sensor equipment, such as sensors for solar radiation, electromagnetic radiation, infrared radiation, presence detection, wind speed. The Smart TV (Smart TRV or Smart TCV) according to embodiments of the present invention can use learning (preferably self-learning) algorithms to maximize its performance and to predict impact of input signals from sensors on the room temperature. Using self-learning algorithms, the smart thermostatic radiator or convector valve according to embodiments of the present invention can also be capable to identify the type of radiator or convector and typical requirements of the room in which the radiator or convector is placed, without the need to customise the control individually and manually. At least some of the sensors can be placed in the room where the radiator or convector is integrated into the thermostatic radiator or convector valve itself.

Fig. lb shows a further floorplan which is for use with embodiments of the present invention.

The heating and/or zones are labeled 1 through 7. The radiators or convector are labeled A through H. There are two radiators or convectors D and E in zone 4 and F and G in zone 5. There is no radiator or convector in the room which communicates with zone 3 via a door. Heating and/or cooling zone 3 has no radiator or convector i.e. is a passive heating or cooling zone, nor does it have a sensor bank. Zones 1, 2, 6, 7 have one radiator or convector.

Smart TV (Smart TRV or Smart TCV)’s with internal or directly connected sensor bank (i.e. additional sensors at least two or more or preferably three or more) are labeled b, c, d, e, and h. Valve a uses an external sensor bank (i.e. additional sensors at least two or more or preferably three or more) in the same room (also labeled a). Valves f and g share a central sensor bank i (i.e. additional sensors at least 2 or more or preferably 3 or more). Valves a, f and g have their own communication system. Valves b, ..., e, and h have communicating sensor banks (illustrated as example by antenna symbols, other physical layer choices are possible).

One central or multiple individual or shared IT units (embedded computers, or server) are integral part of the system, but not shown in the figures. These IT units communicate with all the elements having antenna symbols via a wireless network such as a CAN. The wireless network transmits and receives through the antennas shown as antenna symbols allowing communication to the IT unit(s).

Heating and/or cooling zones 1, 2, 3 are single room, single zone (3 being the hallway). Zone 4 and 5 are single rooms, with multiple zones (4 being governed by radiator or convector E and D, 5 by radiator or convector F and G).

Zone 6 and 7 are multiple rooms, single zone.

Additionally, the smart thermostatic valve according to embodiments of the present invention can be adapted to allow for low flow operation. Figure 4 shows a graph with the power of a radiator in function of the flow at fixed supply temperature. The power of the radiator decreases only in case of (very) low flows.

The Smart thermostatic valve (Smart TV (Smart TRV or Smart TCV)) according to embodiments of the present invention can also be equipped with extra sensors such as pressure sensors for detecting mechanical malfunctioning like clogging and fouling. Based on the position of the valve and the pressure difference over the valve, more detailed information becomes available.

Accordingly, smart thermostatic valve according to embodiments of the present invention can read input from a variety of sensors in order to optimize radiator or convector output. This reduces the risks of overheating/undercooling and adapts the room to the required user comfort. Any of the sensors or detectors can be separate electronic temperature sensors

In another aspect of the present invention, a complete heating system is disclosed having a plurality of radiators and smart thermostatic valves according to embodiments of the present invention configured to keep the radiators or convectors return temperature as low as possible by having means for controlling the flow of heating medium through the radiator or convector, especially in the range of low flow rates. In order to control said flow, the smart thermostatic valve according to embodiments of the present invention can comprise a temperature sensor and can be configured to communicate with additional sensor equipment; including sensors for solar radiation, electromagnetic radiation, infrared radiation, presence detection, outside temperature, wind speed, other TCV’s or TRV’s, room temperature sensors, etc. The smart thermostatic valve according to embodiments of the present invention can use the input of one, some or all of these sensors. These sensor inputs can be used with either pre-programmed heat loss algorithms, or a self-learning algorithm to optimize the radiator or convector output, i.e. realize a low or the lowest possible return temperature. The additional sensors can inform the smart thermostatic valve according to embodiments of the present invention about additional internal and external heat gains, e.g. body heat based on the number of occupants in the room or solar radiation heat entering the room. The smart thermostatic valve according to embodiments of the present invention can receive input from a sensor capable of determining the internal heat gain and the external heat gain, and at least presence detection and solar radiation.

In case of pre-programmed algorithms, the smart thermostatic valve according to embodiments of the present invention can select an appropriate heat loss algorithm based on the sensor input or inputs, e.g. based on the calendar day and the outside temperature; or based on the outside temperature, window orientation and outside temperature forecast; and the like. Hence, in a further embodiment, the smart thermostatic valve can be configured to receive input from certain parameters, e.g. desired room temperature, window surface and orientation, radiator or convector type, nominal radiator or convector power, inner to outer wall surface ratio for the room, room dimensions, insulation characteristics of the room / building, air tightness of the room, room functionality (e.g. bedroom, kitchen, bathroom, living room, ...), cumulative and/or individual power of the internal heat sources (e.g. kitchen supplies, household electronics, ...), etc. either through a direct user interface or remote communication device. The heat loss algorithm can determine the latest time possible for the valve opening in order to start flow to the radiator or convector as well as controlling the flow rate with the earliest possibility of closing the valve to use the residual heat within the room or zone to maintain it at the required temperature until the end of a heating period. In case of a self-learning algorithm, it can use the aforementioned inputs to the smart thermostatic valve algorithm to optimize the radiator or convector output, i.e. to realize a low or the lowest possible return temperature, as well as self-adapting to changing environmental conditions.

The Smart TV (Smart TRV or Smart TCV) according to embodiments of the present invention can operate with more than one threshold temperature, for example:

TH1 - an ideal temperature for a room during the day that is being used. The“day” can be defined by a time period such as 09.00 to 18.00.

TH2 - a temperature for a room during the day which is not being used. An example is a meeting room which may often be empty and, hence, there is no need to heat the room to temperature TH1. Accordingly, TH2 is less than TH1.

TH3 - a night time temperature. Usually TH3 < TH2 < TH1.

TH4 - a temperature above freezing temperature of water that is maintained to prevent frost damage, e.g. during holiday periods or weekends in winter.

In modern, and better insulated houses, it becomes more important to understand the internal heat gains, to control the room temperature, the external temperature used in most heating systems together with a heating curve is getting less important in such well insulated environment. Using input from additional sensor equipment the Smart TV (Smart TRV or Smart TCV) is capable to cope with such internal and external heat gains and adjust the flow through the radiator or convector, accordingly anticipating on the effects of these heat gains on the interior temperature.

Advantages of embodiments of the present invention are: economical to apply, simple to control, having limited components. The technique is not very intrusive: a valve can be quickly installed with limited user hindrance and at low cost. Chances of equipment failure are a lot smaller: the self-learning technology is also capable of detecting changes in performance and can act as early fault handling equipment and can provide an alarm. Data processing and analysis are also very straightforward.

A Smart TV (Smart TRV or Smart TCV) 10 according to embodiments of the present invention is shown schematically in Figure 5. It comprises a housing 7 with a display 3. An entry system 1 is provided for manual customising of the Smart TV (Smart TRV or Smart TCV). The Smart TV (Smart TRV or Smart TCV) 10 is adapted to receive signals from a number of detectors or sensors 2, 4, 6, e.g. wireless signals, of Figure la. The sensors or detectors can be in the same room or zone as the Smart TV (Smart TRV or Smart TCV) 10 or may be outside the room or may be outside the building. The Smart TV (Smart TRV or Smart TCV) 10 preferably has communications interfaces as I/O ports to receive the signals from the sensors 2, 4, 6. The Smart TV (Smart TRV or Smart TCV) 10 has a processing engine such as a microprocessor or FPGA and memory. The Smart TV (Smart TRV or Smart TCV) may have a CPU or GPU. The Smart TV (Smart TRV or Smart TCV) 10 may be adapted to transmit signals through the communication interface, e.g. to other Smart TV’s (Smart TRV or Smart TCV) in other rooms or to an optional main controller. The Smart TV (Smart TRV or Smart TCV) 10 may have a clock and a calendar information stored in a memory, e.g. to be able to track weekends, public holidays, works’ holidays, etc. The Smart TV (Smart TRV or Smart TCV) 10 can be battery powered but can include other forms of power, such as a means for scavenging energy from the environment such as from wireless radiations, infrared, etc or a photovoltaic power supply, or a connection to mains power. The Smart TV (Smart TRV or Smart TCV) 10 can have a screw or bayonet connector 5 which fits to standard dimensions of ports such as ports with screw threads available on existing radiators or convectors.

The Smart TV (Smart TRV or Smart TCV) 10 can be a controllable valve with a movable valve member. The radiator or convector valve can be fitted to a radiator or convector to heat a room or zone within which it is situated. The radiator or convector valve can comprise a valve member, a valve driver and optionally an on-board temperature sensor. The valve member is adapted to be movable between a fully closed position which blocks the flow of heated water to the radiator or convector 2 to a fully open position wherein water is free to flow to the radiator or convector. The driver can drive the valve member to intermediate positions between the fully open and the fully closed position, in which positions the valve member restricts the flow to the radiator or convector. The valve driver can move the valve member by a temperature sensitive expansion element or by deformation of a bimetallic strip with change in temperature. Alternatively, the valve driver may comprise an electric motor. The Smart TV (Smart TRV or Smart TCV) 10 can have a rechargeable battery and/or energy scavenging means for charging said battery. Typically, the energy scavenging means might include thermoelectric means or RF scavenging means.

To improve the efficiency of such a system further, the boiler (Figure 2) can be controlled in response to the state of the Smart TV (Smart TRV or Smart TCV)’s. In particular, when the valves are closed, it is conventional to have an automatic bypass, so that no hot water will flow into the radiators or convectors and, thus, at such times there is no need for the boiler 30 to continue to heat water. As such, the system is designed to shut down the boiler when a number of Smart TV’ s (Smart TRV or Smart TCV) are closed and to re-start boiler operation once some of the Smart TV’s (Smart TRV or Smart TCV) have returned to normal operation. The Smart TV’s (Smart TRV or Smart TCV) 10 can each be provided with suitable wireless link to an optional main or boiler controller. The link may take the form of any suitable wired or wireless link. The Smart TV’s (Smart TRV or Smart TCV) 10 may be provided with suitable transceivers to enable the wireless link. The Smart TV’s (Smart TRV or Smart TCV) 10 may be adapted to send information as to their current status and optionally any information they have received from any of the sensors or detectors with which they are communicating.

Improving return flow temperatures

Embodiments of the present invention make use of means to lower the return temperature while maintaining heat output.

In another aspect, a computer program product is provided which, when executed on the processing engine of a Smart TV (Smart TRV or Smart TCV) 10 according to any of the embodiments of the invention, carries out any of the methods of the present invention. The computer program product may be stored on a non-transitory signal storage means such as an optical disk a magnetic disk, a magnetic tape, or a solid state memory such as a flash drive.

According to an embodiment of the present invention, one way to lower return temperatures is a method and means for providing local intelligence to the Smart TV (Smart TRV or Smart TCV) 10 containing or being in a network with other valves such as other Smart TVs (Smart TRV or Smart TCV) and in communication with a plurality of sensors. The Smart TV (Smart TRV or Smart TCV) 10 can be connected to embedded or central computing equipment (IT system) as described below.

For use in any of the embodiments of the present invention, the computing equipment contains a first or higher order software representation of one or more heating zones and their interactions. A heating zone, e.g. as shown in Figure la or lb, may be represented, for example, by a first order or higher order lumped circuit equivalent model, such as an RC model that abstracts various parameters in analogy to an electric circuit (see Figure 6). The RC model or higher order model defines the relationships between the temperatures of various parts of the zone and/or of the radiators or convector among each other and the parameters. The model has values Rn, R _f 2, R _wi , Q, and C _{w i} which can be individually determined per heating zone. The circuit has a representation of the ground temperature T _g , the internal wall temperature 7 _Wi , and the air temperature T _{Ά ir} . Furthermore, it has a representation of the thermal mass of the internal walls C _{w i} and the thermal mass of the floor C _f as capacities in analogy to capacitors in an electrical circuit. Finally, the model has a representation of thermal conductivities Rn which represent the flow between ground and indoor temperature and Rn and the flow between floor and ground temperature as well as R _wi representing the thermal conductivity between internal walls and indoor air in analogy to resistors in an electrical circuit. The RC model defines the relationships between the temperatures among each other and the parameters.

An implementation of the RC model according to an embodiment of the present invention includes software that may be implemented as a computer program product which has been compiled for a processing engine to carry out any of the methods of the present invention or is compiled to execute in an interpretative virtual machine such as the Java™ Virtual Machine in a Smart TV (Smart TRV or Smart TCV) 10. The software, when executed on a processing engine, may provide representations of one or more heating zones, and this representation may be of the RC model. The software can be embodied in a computer program product adapted to carry out its functions when the software is loaded onto one or more Smart TVs (Smart TRV or Smart TCV) 10 and executed on one or more processing engines such as microprocessors, ASIC’s, FPGA’s, etc. A Smart TV (Smart TRV or Smart TCV) 10 may comprise logic encoded in media for performing any step of the steps of the methods according to embodiments of the present invention. Logic may comprise software encoded in a disk or other computer-readable medium and/or instructions encoded in an application specific integrated circuit (ASIC), field programmable gate array (FPGA), or other processor or hardware. A device will also include a CPU and/a GPU and memory, the CPU and/or GPU having a processing engine able to execute software of the present invention.

The computer program product may be stored on a non-transitory signal storage medium such as an optical disk (CD-ROM or DVD-ROM), a digital magnetic tape, a magnetic disk, a solid state memory such as a USB flash memory, a ROM, etc.

The software representation of the one or more heating zones e.g. as shown in Figure la or lb comprises a prediction-based algorithm that predicts the effect of environmental changes as detected by the plurality of sensors and the effect of given settings on the Smart TV (Smart TRV or Smart TCV) 10 and/or the network of valves with a given supply temperature. The software may include a control algorithm that controls, i.e. chooses, valve settings of one or more Smart TV (Smart TRV or Smart TCV) 10 according to the respective predicted outcome for each Smart TV (Smart TRV or Smart TCV) 10 according to the predicted influence of environmental changes detected by the plurality of sensors. The prediction-based algorithm can optimize the return temperature. Preferably, the prediction-based algorithm prevents oscillations of the controlled system (“hunting”).

According to an embodiment of the present invention, another way to lower return temperatures is to place one or more fans below a radiator or convector so that air passes over the radiator or convector surfaces and picks up heat energy in this way. This results in the water in the radiator or convector lowering in temperature faster. Such a fan or fans can be operated manually, or from the Smart TV (Smart TRV or Smart TCV) 10 or optionally remotely e.g. if it has a wireless switch. In the latter case, the fan can be included in the sensor network and can be controlled by the optional main controller.

In an embodiment the Smart TV (Smart TRV or Smart TCV) 10 receives signals from or derived from the return pipe temperature sensor Temp 10 of Figure la. If this temperature rises above a threshold the Smart TV (Smart TRV or Smart TCV) 10 activates the one or more fans. This increases the rate of heat transfer to the air. This will cause an increase in temperature in the room which will be detected by a room temperature sensor such as the sensors Templ to Temp4 and when the room temperature exceeds TH1, this will be communicated to the Smart TV (Smart TRV or Smart TCV) 10 which will throttle the flow of hot water to the radiator or convector. The combination of lower flow rate and the action of the fans will lower the exit temperature from the radiator or convector.

The learning algorithm running on the processing engine of the Smart TV (Smart TRV or Smart TCV) 10 can learn a relationship between the activation of the fans and the increase of room temperature. For example, the time lag may be modelled by an exponential function and a half-life. Based on this learned half-life, the Smart TV (Smart TRV or Smart TCV) 10 may throttle the flow to the radiator or convector before the room temperature sensor reacts e.g. after a certain time after fan activation based on the model in order to prevent an overshoot of temperature in the room.

As indicated in Figure 4, the flow rate has to be dropped to low levels to change the rate of heat loss from a radiator or convector. The Smart TV (Smart TRV or Smart TCV) according to embodiments of the present invention can be adapted for pulsed operation. Pulsed pumps are widely used for fuel pumping in automobiles and are reliable. The Smart TV (Smart TRV or Smart TCV) 10 according to embodiments of the present invention can be adapted for pulsed operation at low flow rates whereby the pulse frequency can be low, e.g. some seconds. Depending upon the duty cycle, the average amount of flow of hot water can be lowered. More time is available for the hot water to cool between pulses, thus the outlet temperature is lowered. Thus, on receiving a signal derived from the temperature sensor Temp 10 that the return temperature is too high, the Smart TV (Smart TRV or Smart TCV) 10 can initiate pulsed operation either alone or in combination with fan activation. Pulsed operation allows the Smart TV (Smart TRV or Smart TCV) to operate at mid-range for low flow rates improving control. The learning program running on the Smart TV (Smart TRV or Smart TCV) 10 can leam a relationship between what pulsed operation duty cycle is required to achieve a specific drop in exit temperature in order to achieve the desired return temperature.

Generally, a radiator has four connection points of which only two or three are used. In scheme 1 in Figure 4, the top and bottom connections on one side are used with the bottom one for out and the top one for in. In scheme 2 connections at either end of the radiator are used. In scheme 3 two connections at either end of the radiator are used. In scheme 4 in and out is at one position. The graph indicates that scheme 4 provides the best control and this arrangement is often used in practice. However, for schemes 1 and 2 improvement is still needed. For all schemes a better control is preferred.

For most installed radiators there is always one connection point free. Many radiators have more than one plate. The plates are linked together hydraulically usually in parallel. In accordance with this embodiment, a controllable restriction is introduced into the plate linking pipes which is controlled by the Smart TV (Smart TRV or Smart TCV) 10 or by the optional main controller to restrict the flow to flow in series through the plates. This will be called labyrinth flow. If there are N plates, this means that the hot water has to travel up to N times longer to reach the exit of the radiator and, hence, has N times longer to cool. The connections of radiators are standardised and the modifications for labyrinth flow do not need to seal off the flow to the different plates completely. For example, a rubber housing with an inner valve can be lodged into the connection pipes, whereby the flow is mainly serially along the plates.

Reacting to internal and external heat sources

Internal and external heat sources can alter the temperature in rooms rapidly, e.g. from solar power, in meeting rooms with large numbers of persons. Embodiments of the present invention can reduce the output of radiators or convector quickly without stopping flow completely. For example, initiating labyrinth flow will lower the output of radiators or convector.

In a further embodiment through at least one unused connection point an air bag is introduced which can be achieved on existing installations easily as the diameter of such connections is larger than 10 mm. The air bag is connected to a tube which passes through the connection point through a connector designed therefor. The tube is connected to an air pump, e.g. battery or mains operated. The air pump can be operated manually or remotely, e.g. may be connected to the sensor network and controlled by the optional main controller or by the local Smart TV (Smart TRV or Smart TCV) 10. Inflation of the air bag causes the volume of the radiator or convector to reduce and, hence, the heat output of the radiator or convector is reduced, even with the same flow rate. The inflation has a larger effect on the heat released than lowering the flow rate.

The learning algorithm running on the Smart TV (Smart TRV or Smart TCV) 10 can leam to react to alterations in signals derived from solar power sensors SI, S2 in order to control heat output from radiators or convectors in advance of an increase of room temperature due to solar power.

The Smart TV (Smart TRV or Smart TCV) 10 can also be adapted to receive signals from electric current sensors of electric currents entering a room through cables for the driving of electrical equipment. The Smart TV (Smart TRV or Smart TCV) 10 can take this energy input into a room into account in deciding on regulating valve operation of a radiator or convector. The learning algorithm running on the Smart TV (Smart TRV or Smart TCV) 10 can learn to react to alterations in electrical energy liberated in a room.

The more persons there are in a room, the larger the heat output from them. The learning algorithm running on the Smart TV (Smart TRV or Smart TCV) 10 can leam to react to changes in heat energy liberated in a room by persons. The learning algorithm can leam, for example, from the noise level generated by persons to estimate the number of persons in the room. For this purpose, one or more sensors can be provided in a room or zone which measures the noise level, especially in the frequencies of human voices.

A method and means for executing a learning algorithm can also be provided, with the Smart TV (Smart TRV or Smart TCV) 10 containing or being in a network with other valves such as other Smart TVs (Smart TRV or Smart TCV) connected to embedded or central computing equipment (IT system). As described above in one aspect, the computing equipment contains a first or higher order software representation of one or more heating zones and their interactions e.g. zones or rooms as shown in Figure la or lb. A heating zone may be represented, for example, by a first order or higher order lumped circuit equivalent model such as an RC model that abstracts various parameters in analogy to an electric circuit (Figure 6). The circuit captures a representation of the ground temperature T _g , the internal wall temperature 7 _Wi , and the air temperature T _{Ά ir} . Further, it captures the thermal mass of the internal walls C _{w i} and the thermal mass of the floor C _f as capacities in analogy to capacitors in a circuit. Finally, the model captures the thermal conductivities Rn as the flow between ground and indoor temperature and Rn as the flow between floor and ground temperature as well as R _{w ;} as the thermal conductivity between internal walls and indoor air in analogy to resistors in a circuit. The RC model defines the relationships between the temperatures among each other and the parameters. The values Rn, Rn, R _wi , Q, and C _{w i} are found individually per heating zone from the measured temperatures T _g , 7 _Wi , and F _air over time in a process called parameter estimation. In this process, an optimal valuation of the parameters is found by systematically exploring possible parameter values through an optimization algorithm, so that the predicted values of the temperatures over time based on these parameters best match the actual, measured temperatures in that heating zone. Over time, by regularly re- estimating the parameters, the mathematical representation of the heating zone can be improved or adapted to physical changes in the setup of a heating zone in an automated fashion. In another aspect, the GG system includes algorithms for computational statistics / machine learning - based on regression or (deep) neural networks - able to infer correlations between recorded values for valve settings, supply temperatures, and readings from the plurality of sensors and control actions, equally capable of pre-emptively executing control actions to optimize the return temperature and avoid space temperature oscillations.

An Implementation

In accordance with another embodiment of the present invention, software may be implemented as a computer program product which has been compiled for a processing engine to carry out any of the methods of the present invention or is compiled to execute in an interpretative virtual machine such as the Java™ Virtual Machine. A TRV according to embodiments of the present invention may comprise logic encoded in media for performing any step of the steps of the methods according to embodiments of the present invention. Logic may comprise software encoded in a disk or other computer-readable medium and/or instructions encoded in an application specific integrated circuit (ASIC), field programmable gate array (FPGA), or other processor or hardware. A device will also include a CPU and/a GPU and memory, the CPU and/or GPU having a processing engine able to execute software of the present invention.

The computer program product may be stored on a non-transitory signal storage medium such as an optical disk (CD-ROM or DVD-ROM), a digital magnetic tape, a magnetic disk, a solid state memory such as a USB flash memory, a ROM, etc.

The software can be embodied in a computer program product adapted to carry out the following function when the software is loaded onto the respective device or devices such as a TRV having one or more processing engines such as a microprocessor, ASIC, FPGA etc., and executed on the one or more processing engines such as microprocessors, ASIC’s, FPGA’s, etc.: a method of operating a radiator or convector and a controllable radiator or convector valve, the radiator or convector having an inlet for hot liquid and an outlet for exiting liquid from the radiator, the radiator or convector valve comprising a valve driver, and being in operative communication with a plurality of sensors, the plurality of sensors being at least three sensors at least located in a zone or room to be heated by the radiator or convector.

The software can be embodied in a computer program product adapted to carry out the following function when the software is loaded onto the respective device or devices such as a TRV having one or more processing engines such as a microprocessor, ASIC, FPGA etc., and executed on the one or more processing engines such as microprocessors, ASIC’s, FPGA’s, etc.: controllably driving a valve member between an open position and a closed position; and wherein valve control means are in communication with the plurality of sensors and operate in response thereto to drive the valve member to a more open or closed position if the sensed temperature exceeds or is lower than a first threshold temperature, further comprising activating means to reduce temperature of the exiting liquid.

The software can be embodied in a computer program product adapted to carry out the following functions when the software is loaded onto the respective device or devices such as a TRV having one or more processing engines such as a microprocessor, ASIC, FPGA etc., and executed on the one or more processing engines such as microprocessors, ASIC’s, FPGA’s, etc.:

pulsed operation of the controllable radiator or convector valve or activating one or more fans directing air against the radiator or convector, receiving and processing outputs from the at least three sensors arranged as a bank of sensors located in or on the radiator or convector valve or remote from the radiator or convector valve but in communication therewith, measuring a temperature of the room or zone and measuring internal and external heat gain with respect to the room or zone, measuring directly or indirectly the effect of solar radiation inputting heat energy into the room or zone, internal heat gain being heat generated by electronic appliances in the room or zone.

The software can be embodied in a computer program product adapted to carry out the following functions when the software is loaded onto the respective device or devices such as a TRV having one or more processing engines such as a microprocessor, ASIC, FPGA etc., and executed on the one or more processing engines such as microprocessors, ASIC’ s, FPGA’s, etc.: executing a predictive model that predicts a setting for the controllable radiator or convector valve, the predictive model can be executed as a self-learning mode, the predictive model can be based on a lumped equivalent circuit RC model.

The software can be embodied in a computer program product adapted to carry out the following functions when the software is loaded onto the respective device or devices such as a TRV having one or more processing engines such as a microprocessor, ASIC, FPGA etc., and executed on the one or more processing engines such as microprocessors, ASIC’s, FPGA’s, etc.: the self-learning model recording output values from the plurality of sensors over time and applies parameter estimation to tune the model of the room or zone, reducing temperature of the exiting liquid being a pulsed operation of the controllable radiator or convector valve.

The software can be embodied in a computer program product adapted to carry out the following functions when the software is loaded onto the respective device or devices such as a TRV having one or more processing engines such as a microprocessor, ASIC, FPGA etc., and executed on the one or more processing engines such as microprocessors, ASIC’s, FPGA’s, etc.: one of the plurality of sensors measures the temperature of return pipe to a boiler, and the controllable radiator or convector valve is adapted to activate one of the means to reduce temperature of the exiting liquid on receipt of a signal derived from an output of the sensor measuring the temperature of return pipe, reacting to heat sources located in a zone or room or external to a zone or room.

The software can be embodied in a computer program product adapted to carry out the following functions when the software is loaded onto the respective device or devices such as a TRV having one or more processing engines such as a microprocessor, ASIC, FPGA etc., and executed on the one or more processing engines such as microprocessors, ASIC’s, FPGA’s, etc.: capturing a representation of a ground temperature T _g , an internal wall temperature 7 _Wi , and an air temperature T _{Ά ir} , capturing a thermal mass of the internal walls C _{w i} and a thermal mass of the floor C _f as capacities, wherein the predictive model is a lumped equivalent circuit model in analogy to capacitors in an electric circuit, capturing the thermal conductivities Rn as the thermal flow between ground and indoor temperature and Rn as the thermal flow between floor and ground temperature as well as R _Wi as the thermal conductivity between internal walls and indoor air of the lumped equivalent circuit model in analogy to resistors in a circuit, the predictive model defines the relationships between the temperatures and finds Rn, Rn, R _wi , C _f , and C _{w i} individually per heating zone from the measured temperatures T _g , 7 _Wi , and T _air over time by parameter estimation.

The computer program product or software as described above may be stored on a non- transitory signal storage medium such as an optical disk (CD-ROM or DVD-ROM), a digital magnetic tape, a magnetic disk, a solid state memory such as a USB flash memory, a ROM, etc.