Technical Field

The present invention relates to an accident prevention system and method to be used in railways transport, that informs user on whether there is a failure or not by checking conditions of the rails.

The present invention particularly relates to an accident prevention system and method to be used in artificial intelligence-based rail systems, which allows for checking whether there is failure or not by way of performing a classification based on the change in regular sound intensity and spectrum occurring in the project speed (optimum speed foreseen to be used in railways) of rails by being positioned on the rail vehicle and also for informing user on the classification result together with GPS (Global Positioning System) position and time information as a railway condition map.

State of the Art

Nowadays, the method commonly used in checking the rails is individual observation by an officer. In this method, units in charge examines the rails by means of the devices in a certain periodical maintenance. The main drawback in the method is that every point of the rail line cannot be monitored continuously. Furthermore, since the method depending strictly on experience and perception of the person in charge varies from person to person, standardization cannot be achieved in the checking processes. A failure considered to cause no problem until a next examination period by an inexperienced officer may lead disastrous accidents due to climate conditions, weather and raining conditions and use frequency of rails and this may cause not only goods losses and but also losses of lives.

Furthermore, even though the operation logic and the failures are different from each other, there are devices mounted between wheels and body of the rail vehicle by welding. Said devices monitors conditions of wheels and rail vehicle, while the rail vehicle moves. Temperature of oil in the wheels and accordingly, maintenance and condition of rail vehicles are determined by said devices. However, the most important different of the present invention is that it examines the rail and railway condition, not the rail vehicle. Further, while the devices examine parameters such as the oil temperature, viscosity etc., the examination is performed over a sound analysis through the present invention.Moreover, a condition map of the whole railway is presented to user by means of the sound analysis.

Another application used in the state of the art measures the vibration values of the line by means of rods penetrated in certain intervals or fiber system application. The most important difference of the present invention from those applications is the technology employed, performing the examination over the sound analysis and afterwards, the technique for informing user on the condition map of a whole railway by means of the sound analysis. This eliminates to penetrate rods along an entire railway or to establish a fiber system as proposed by the mention applications and the rail can be checked by means of a single system. Furthermore, the fact that there is a crack that may occur in the systems necessities to check the whole line.

In the state of the art, giving a signal is available based on the signal power received through mounting a receiver and transmitter to the rail vehicles to avoid crashes. However, this system avoids only crashes and does not offer a solution for failures on the rails.

In the document numbered TR 2012/09466 in the state of the art, it is disclosed a rail crack and fracture sensing method used in detecting railway rail failures in the rail systems technology. The invention particularly relates to a method of detection of rail cracks or fractures that can be used for the determination of rail breakdowns in the field of railway systems technology and incorporates a power supply, an electromechanical vibration application module that periodically applies a vibration signal to the rail body with the energy taken from the said power supply, a vibration application module drive card which drives this module and a vibration sensing module that is oriented on the same rail at a certain distance from the electromechanical vibration application module, and an electronic sensing card for vibration sensing module with a capability to communicate to the outside world with various means and process the incoming signal, evaluating it for the detection of rail cracks and fractures, a vibration sensor. However, in the application, it is not disclosed to perform an examination over the sound analysis and informing user on the entire railway condition map by means of the sound analysis and as stated in the above, a measurement is performed through the vibration.

In the document numbered TR 2015/16844 in the state of the art, it is disclosed a rail checking system for avoiding accidents to occur by means of detecting deformation on the railway, on which the rail vehicles travel. In the application, an inertial system uses coordinate data of global positioning system and detects the position of the rail vehicle and measures the acceleration thereof. The acceleration measures are afterwards transferred to the electric control unit and compared to the reference acceleration thresholds. If there is an acceleration value exceeding the threshold, it is interpreted that there is a deformation and the location data is detected. The related application comprises location data differently from previous applications. However, it is not disclosed to perform an examination over the sound analysis again and informing user on the entire railway condition map by means of the sound analysis and as stated in the above, a measurement is performed through the vibration again.

A rail vehicle moving on the line moves with a stable speed determined in the line project and defined to the vehicle. Academical studies conducted in the state of the art have shown that a rail vehicle moving with the project speed (optimum speed foreseen to be used on railways) on the rail lines generates the same sound intensity regularly (with a stable frequency). In the application numbered CN110470736 (A) in the state of the art, it is disclosed a wireless network-based steel rail monitoring system and the control method thereof and can be evaluated as the closest document. The related application concerns steel rail safety monitoring technique and particularly, relates to a steel rail monitoring system based on a wireless network and control method. The system comprises a monitoring module, a central processor and a remote terminal. The monitoring module is connected with the central processor and used to detect the current condition of a steel rail and transfer the condition information to the central processor; the central processor is used to process the condition information received, to analyze the same based on different condition information and then to transfer the analysis result to a remote terminal by means of a wireless transfer module; the monitoring module comprises an ultrasonic sensing module, a camera module and a sound sensing module; the ultrasonic sensing module detects whether the steel rail is cracked or not in a real-time manner during a train travel, the camera module is used to collect a present steel rail pictures in a real-time manner and to transfer a picture to the central processor and the sound sensing module is used to detect the sound between wheels and steel rail while the train moves on the rail, thereby evaluating the condition of the steel rail and wheels. Further, in the invention, it is also disclosed to transfer the picture together with the location data. However, it is to be noted that the invention does not comprise any road condition map information and an artificial intelligence-based software forming the information. In the application numbered CN110470736 (A), the detection is performed through the sound and afterwards, the image taken through the camera is combined and sent with the location data and in such case, the train already will have approached the region with failure until the range that can be measured by the sound waves and may experience an undesired delay. Yet, in the present invention, since the railway condition map is formed, the railway operator may observe the case and take precautions and inform the officers in charge by waiting at the previous station, if necessary and thus, the operator will not face a surprise delay. This advantage provided to the related technical by means of the software and railway condition map allows the operator for observing the rail condition at every point he/she operates, thus planning the travel and also eliminating the surprise delays and accident risks.

As a result, due to the aforementioned drawbacks and insufficiency of the existing solutions regarding the subject matter, the need for a development in the related technical field has arisen.

Object of the Invention

The invention is created by being inspired with the present situations and aims to solve the problems mentioned above.

Object of the present invention is to provide an accident prevention system and method to be used in railways transport, that informs user on whether there is a failure or not by checking conditions of the rails and thus detects the failures in advance occurring on the rails by listening continuously the rail line and prevents potential accidents. By means of the present invention, structural failures occurring on the rail lines will be detected the formation phase, thereby preventing disastrous accidents.

Another object of the present invention is to provide an accident prevention system and method to be used in artificial intelligence-based rail systems, which allows for checking whether there is failure or not by way of performing a classification based on the change in regular sound intensity and spectrum occurring in the project speed of rails by being positioned on the rail vehicle and also for informing user on the classification result together with GPS position information as a railway condition map. While the present invention detects the railway failures, the present invention is a product that performs detection based on the change in the sound intensity and spectrum values occurring while the rail vehicle moves and outputs the railway condition map in the GPS system and presents the map to the user.

Another object of the present invention is to provide an accident prevention system and method to be used in rail systems, that can be attached easily on the rail vehicle without a significant mounting process such as welding and the like without interfering with the rail vehicle structure. The present invention is more advantageous compared to the present systems in terms of its applicability .

Another object of the present invention is to provide an accident prevention system and method to be used in rail systems that allows for monitoring continuously the structure condition of the rail and railway without the need for another maintenance vehicle by means of the rail vehicle regularly operating on the line.

So as to achieve the objects above mentioned, the present invention relates to an accident prevention system in rail systems that allows for informing user on whether there is a failure or not by checking the rail conditions on which the rail vehicles moves in railways and thereby preventing a potential accident, comprising:

• at least one sound sensor for collecting sounds resulting from the interaction between said rail vehicles and rails during the movement,

• at least one data outputting board for outputting the sound data collected by means of said sound sensor together with the time and location data by means of the GPS network,

• at least one data collecting and outputting board for outputting the sound collected from said data outputting board and the time and location data collected from GPS network so as to be processed,

• at least one sound pool wherein the failure sounds that may result from the interaction between said rail vehicle and rails are predefined and saved,

• software for detecting rails with failure, when there is a match-up, by comparing the sound data collected from said data collecting and outputting board to sounds in the sound pool and forming the railway condition map indicating whether there is a failure in rails or not by means of visualizing the location and time data collected from the data collecting and outputting board and the comparison results, at least one imaging unit for transferring said railway condition map to user.

So as to achieve the objects above mentioned, the present invention relates to an accident prevention method in rail systems that allows for informing user on whether there is a failure or not by checking the rail conditions on which the rail vehicles moves in railways and thereby preventing a potential accident, comprising:

• collecting sounds resulting from the interaction between said rail vehicles and rails during the movement by means of the sound sensor,

• transferring the sounds collected to the data outputting board in the rail vehicle,

• transferring to the data collecting and outputting board by adding the time and location data into the sound data in the data outputting board by means of GPS network,

• transferring the sound, time and location information in the data collection and outputting board to the imaging unit,

• detecting rails with failure in case of a match-up by comparing the sound data collected with the data in the sound pool, in which the failure sounds that may result from the interaction between the rail vehicle and rails are predefined and saved, by means of the software operating said imaging unit.

• forming the railway condition map indicating whether there is a failure in rails, by way of visualizing the location and time data collected from the data collecting and outputting board and the comparison results by means of said software, presenting the railway condition map to user by means of said imaging unit,

Structural and characteristic specifications and all advantages of the invention will be understood more clearly by means of following figures and detailed description written by making references to these figures and therefore, assessment should be done by taking into consideration these figures and detailed description.

Figures for a Better Understanding of the Invention Figure 1 illustrates a representative view of the inventive accident prevention system in rail system.

Description of Component References

1. Sound sensor

2. Data outputting board

3. GPS network

4. Data collecting and outputting board

5. Software

6. Railway condition map

7. Imaging unit

8. Rail vehicle

9. Rail 10. Sound pool

Detailed Description of the Invention

In this detailed description, the inventive accident prevention system and method in rail systems are described only for a better understanding of the subject, such that it constitutes no limiting effect on the scope of the invention.

A rail vehicle (8) moving on the line moves with a stable speed determined in the line project and defined to the vehicle. As seen in academical studies conducted in advance, the rail vehicle (8) emits regularly (with a stable frequency) the same sound intensity while with a stable speed. That is, the rail vehicle has sound values that can be used as reference in the alignment area of the line and the curve area of the line. However, when a failure is started to occur on the line structure, regular sound values vary based on the failure type starting to occur on the line structure. It is started to exhibit difference in the range of 1-15 dbA based on the failure, differently from the standard sound values, while the rail vehicle is passing on that area.

Figure 1 illustrates a representative view of an accident prevention system in rail systems that allows for informing user on whether there is a failure or not by checking the rail (9) conditions on which the rail vehicles (8) moves in railways and thereby preventing a potential accident.

Said system comprises at least one sound sensor (1) for collecting sounds resulting from the interaction between said rail vehicles and rails (9) during the movement.

In a preferred embodiment of the present invention, there is a sound sensor (1) being two microphones in total, namely one on the right rail (9) and one on the left rail (9), on which the rail vehicle (8) moves.

Said system further comprises a data outputting board (2) for outputting the sound data collected by means of said sound sensor (1) together with the time and location data collected by means of the GPS network (3) and a data collecting and outputting board (4) for outputting the sound collected from said data outputting board (2) and the time and location data collected from GPS network (3) so as to be processed. As can be seen, the outputting process for the data outputting board (2) refers to outputting the sound data to the data collecting and outputting board (4). Data collecting and outputting board (4) outputs the sound, location and time data to the imaging unit (7) to be described in detail in the following so as to be processed and interpreted. GPS network (3) is a spatial navigation system for providing location and time information under every sort of weather conditions through satellites at any line of sight around the world. Location and time information are obtained through GPS network (3) in the invention.

In a preferred embodiment of the present invention, said data collecting and outputting board (4) features DAQ. It can be ensured by means of DAQ feature to process the data obtained by the rail (9) line and to output to the imaging unit (7) by converting into a format suitable for the software interpretation .

The inventive system comprises a sound pool (10), wherein the failure sounds that may result from the interaction between said rail vehicle and rails (9) are predefined and saved. Sound pool (10) comprises sound equivalents of the structural failures that may occur on rails (9).

The system also comprises software (5) for detecting rails (9) with failure, when there is a match-up, by comparing the sound data collected from said data collecting and outputting board to sounds in the sound pool (10) and forming the railway condition map (6) indicating whether there is a failure in rails (9) or not by means of visualizing the location and time data collected from the data collecting and outputting board (4) and the comparison results, Software (5) performs a comparison with the data in the sound pool (10) based on the data collected and it is an artificial intelligence algorithm that can perform interpretation pertaining to the type and intensity of the failure, because it is detected what kind of sound it will emit in case of which failure (frequency, dbA etc.) and it is sampled in the sound pool (10).

The most important point of the present invention, which is novel and distinguishes the present invention from the state of the art, is to blend the sound values and location information by using the sound identification technology being the technology of the time and to present an updated railway condition map (6) of the entire line condition to the final user. Thus, the user is able to observe the condition of all rails (9) along the forward railway and accordingly take precautions. When the rail vehicle (8) itself passes on that rail (9), the data is updated accordingly by the software (5).

The inventive system comprises an imaging unit (7) for transferring said railway map (6) to user. As can be seen in Figure 1, said imaging unit (7) is connected to the data collecting and outputting board (4). It operates said software (5) thereon and keeps the sound pool (10) in its memory.

In a preferred embodiment of the present invention, said imaging unit (7) may be mobile phone, computer or PLC (Programmable Logic Controller) - HMI (Human Machine Interface) pair.

As can be seen, the present invention that can be easily mounted on the rail vehicle (8) can operate with a significant mounting process such as welding or the like without interfering with the structure of rail vehicle (8) and also advantageous in terms of its applicability. The rail (9) and road condition can be monitored regularly without the need for another maintenance vehicle by means of the present invention that already operates regularly on the line.

Accident prevention system in rail systems that allows for informing user on whether there is a failure or not by checking the rail (9) conditions on which the rail vehicles (8) moves in railways and thereby preventing a potential accident comprises the following process steps of;

• collecting sounds resulting from the interaction between said rail vehicles and rails (9) during the movement by means of the sound sensor (1),

• transferring the sounds collected to the data outputting board (2) in the rail vehicle (8),

• transferring to the data collecting and outputting board (4) by adding the time and location data into the sound data in the data outputting board (2) by means of GPS network (3),

• transferring the sound, time and location information in the data collection and outputting board (4) to the imaging unit (7),

• detecting rails (9) with failure in case of a match up by comparing the sound data collected with the data in the sound pool (10), in which the failure sounds that may result from the interaction between the rail vehicle and rails (9) are predefined and saved, by means of the software (5) operating said imaging unit

(7),

• forming the railway condition map (6) indicating whether there is a failure in rails (9), by way of visualizing the location and time data collected from the data collecting and outputting board (4) and the comparison results by means of said software (5), presenting the railway condition map (6) to user by means of said imaging unit (7).

In a preferred embodiment of the present invention, sound data in the data collecting and outputting board (4) are transmitted to the imaging unit (7) by means of the global serial data.

By means of the present invention, said railway condition map (6) may be presented to user colorfully by means of the software that can perform interpretation and the GPS network (3) providing the location and time information. For example, in a common manner, rails (9) without failure are indicated in green, rails (9) with slight failure in yellow and rails (9) with failure in red.

Main failure instances targeted to be detected by means of the inventive system and method before causing an accident may be listed as connection looseness in rails, sliding on the line, loose ground, damaged welding, rough rail, undulated rail, rail failure with gasket. The most important case that may be caused by those failures and is desired to be prevented by the present invention is derailment.

Thus, it is achieved an accident prevention system and method to be used in artificial intelligence-based rail systems, which allows for checking whether there is failure or not by way of performing a classification based on the change in regular sound values occurring in the project speed of rails (9) by being positioned on the rail vehicle (8) and also for informing user on the classification result together with the location and time information received from the GPS network (3) as a railway condition map (6). ABBREVIATIONS

GPS : Global Positioning System PLC : Programmable Logic Controller

HMI : Human Machine Interface DAQ : Data Acquisition- dbA: A sound evaluation unit, in which medium and high frequencies are particularly stressed, to which the human ear is sensitive mostly.