TECHNICAL FIELD

The present invention relates to a method for estimating the wheelbase length of at least one trailer of a vehicle combination comprising a towing vehicle and at least one towed trailer. The method is especially suited for vehicle combinations having more than one towed trailer.

BACKGROUND ART

In order to reduce the number of heavy vehicles on the roads, longer vehicle combinations comprising more than one towed trailer are proposed for the use on regular roads in some countries. Apart from reducing the required number of towing vehicles for a specific load, the energy consumption and the emission of exhaust gases will also be reduced compared with traditional vehicle combinations. Normally, the length and the weight of the vehicle combination are controlled by law and regulations. One vehicle combination that at the moment is allowed in some European countries is a truck-dolly-semitrailer combination, which is slightly longer than a normal truck-trailer combination. Such a combination allows a truck to tow a semitrailer using a small trailer often referred to as a dolly or dolly trailer. A dolly is a small trailer that can be coupled to a truck or trailer in order to support a semi-trailer. The dolly is equipped with a fifth wheel to which the semi-trailer is coupled.

In some countries, longer and/or heavier vehicle combinations are also allowed under restricted conditions. Such vehicles may comprise several trailers and may be over 50 meters long and more. They are often used in remote areas and for specific purposes. In Australia, road trains comprising more than 4 trailers are used in some states and on some roads. Longer combination vehicles (LCV's) are also used in e.g. USA, Canada and Argentina. All these LCV's are used under strict regulations. One problem that may arise when longer and varying vehicle combinations are used, is to determine the effective wheelbase of each towed trailer. Knowledge of the effective wheelbase is e.g. advantageous to have when slow speed driving of the vehicle combination is to be assisted by an autonomous or semi-automatic function, e.g. when marshalling, in order to simplify the positioning of the vehicle combination at a loading ramp. It is especially important to know the effective wheelbase when automatic or semi-automatic reversing is to be implemented. The effective wheelbase is a well-known expression. When a vehicle is provided with more than one axle at one end, these axles can be approximated with only one virtual axle. A vehicle having only this virtual axle would in the ideal case behave almost exactly in the same way as the original vehicle with more axles. For a vehicle combination, i.e. a towing vehicle followed by one or several towed units, each vehicle unit will thus have one effective wheelbase. In Fig. 1 , a vehicle combination 1 comprising a truck 2, a dolly 3 and a semitrailer 4 is shown. The effective wheelbase is the distance between the connection point of the trailer and the virtual wheel axle. For the dolly 3, the effective wheelbase is the distance L ₂ between the trailer coupling 5 and the virtual axle 6 of the dolly, which is derived from the regular axles 7 of the dolly. For the semitrailer 4, the effective wheelbase is the distance L ₃ between the kingpin 8 of the semitrailer and the virtual axle 9 of the semitrailer, which is derived from the regular axles 10 of the semitrailer. In Fig. 1 , the virtual wheel axles are shown with dashed lines.

The easiest way of determining the effective wheelbase of a trailer is to measure the distance from the connection point to the wheel axles and to determine the tyre characteristic and the axle loads. The effective wheelbase can then be calculated using available explicit formulas. This may be a reliable method when the same trailer is used exclusively and is running with the same load, or when trailer axle longitudinal position, tyre properties and load is continuously made available to all control units that require the information. In reality this is most often not the case. The effective wheelbase, i.e. the position of the virtual axle, is dependent on each axle load and the tyre characteristics, which means that the effective wheelbase may vary depending on how the trailer is loaded. Further, for a trailer having a liftable axle, the wheelbase will change when the axle is raised.

US 6,301 ,548 discloses a method for determining the wheelbase of steerable vehicles when cornering, in which the wheelbase is determined from at least one defined track width, measured wheel circumference speeds and/or measured steering angles. In this method, the radius of the curve is used as an input to the method which means that the vehicle must drive through a curve with a constant radius. EP 2 324 323 B1 discloses a method for determining a wheelbase of a vehicle having at least two axles by which measurement signals of the vehicle are recorded at least during travel of the vehicle through a curve, and where the wheelbase is determined from at least the measured signals. A yaw rate and the rotational speeds of the wheels for determining the wheel speeds are measured and the wheelbase is determined from geometrical calculations. In this method, the radius of the curve is used as an input to the method which means that the vehicle must drive through a curve with a constant radius.

These methods are mainly adapted to determine the wheelbase for a vehicle combination comprising one towing vehicle and one towed trailer driving through a continuous curve. There is thus still room for an improved method and arrangement adapted for vehicle combinations having at least one towed trailer and where steering is not restricted to steady state cornering. DISCLOSURE OF INVENTION

An object of the invention is therefore to provide an improved method for estimating the effective wheelbase length of a trailer in a vehicle combination comprising a towing vehicle and at least one towed trailer. A further object of the invention is to provide an improved arrangement for estimating the effective wheelbase length of a trailer in a vehicle combination comprising a towing vehicle and at least one towed trailer.

The solution to the problem according to the invention is described in claim 1 for the method and in the characterizing part of claim 11 for the arrangement. The other claims contain advantageous further developments of the inventive method and arrangement.

In a method for estimating the effective wheelbase of a trailer in a vehicle combination comprising a towing vehicle and at least one towed trailer, the steps of driving the vehicle combination forwards, recording the speed of the vehicle combination, recording the steering angle of the towing vehicle, recording the yaw rate of the towing vehicle and of the towed trailers or recording the articulation angle for each towed trailer, recording the steering angle of each towed trailer, and using the determined values to calculate a value for the effective wheelbase for each towed trailer are comprised.

By this first embodiment of the method, the method can estimate the effective wheelbase of the at least one towed trailer. The estimation of the effective wheelbase is done when the vehicle combination is driven forwards for a predetermined time interval, where the time interval may be between a few seconds up to several minutes. The duration of the time interval may e.g. depend on the type of road on which the vehicle combination is driven. Since the steering angles and the yaw rates and/or the articulation angles are used for the estimation, it is advantageous that the vehicle combination at least drives through one curved portion of the road. During the time interval, the different values are determined and recorded with a predetermined sample frequency, such that a series of recorded value sets, which may be arranged in matrixes, are obtained. One or more of these recorded value sets or matrixes are then used to calculate the effective wheelbase for the towed trailers of the vehicle combination.

It is possible to sort the value sets before the effective wheelbase is calculated. When the vehicle combination drives absolutely straight, the recorded values for e.g. the steering angles or the articulation angles of the towed trailers will be substantially zero, which means that there is no point in calculating an effective wheelbase using these value sets. Thus, such value sets can be discharged before the effective wheelbase is calculated. It is also possible that such values are not recorded. It is e.g. possible that some or all values are compared with threshold values before a value set is recorded. If one of the values is below or above the threshold level, the value set is not recorded. When the vehicle combination has travelled for the predetermined time interval, a number of value sets have been recorded. One or more of these value sets are used to calculate an effective wheelbase. It is possible to calculate a number of wheelbases using different value sets and to average these effective wheelbase values.

The estimation of the effective wheelbase is preferable done by fitting a vehicle model using a least square condition. In this way, the deviations in the used values are minimized. Deviations in the values may e.g. come from disturbances or noise in the measured signals. This may e.g. arise from the road conditions, which may not always be perfect. A pot hole may e.g. influence the yaw rate of the truck or the steering angle value of a trailer. Other types of road defects may also influence the estimation. By using a least square approximation, and by performing several estimations for different value sets, an effective wheelbase estimation that differ a few percent or less from the actual effective wheelbase value can be obtained.

The different recorded values are either obtained by measuring the values directly using sensors or by estimating the required values with the use of other measured signals.

One measure that can be estimated is the yaw rate. The yaw rates of the vehicles are e.g. coupled to the articulation angles of the following vehicles. The yaw rate of the first towed trailer can e.g. be approximated by using the yaw rate of the towing vehicle and adding the time derivative of the articulation angle between the towing vehicle and the first towed trailer.

The yaw rate of the towing vehicle can e.g. be determined either by estimation or by a measurement. The estimation may be done by using an estimation method based on vehicle combination properties, such as wheel speed of the vehicle, vehicle mass, vehicle length, steering angle etc. Yaw rate measurements are done by using a yaw rate sensor.

In an arrangement for estimating the effective wheelbase length of a trailer in a vehicle combination comprising a towing vehicle and at least one towed trailer, comprising means for obtaining the speed of the vehicle combination, means for obtaining the steering angle of the towing vehicle, means for obtaining the steering angle of each towed trailer, the object of the invention is solved in that the arrangement further comprises means for obtaining the yaw rate of the towing vehicle and each towed trailer or for obtaining the articulation angle for each towed trailer, means for recording the obtained values in a data set, and means for calculating a value for the effective wheelbase for each towed trailer.

By this first embodiment of the arrangement, the arrangement comprises means for estimating the effective wheelbase of the at least one towed trailer. The estimation of the effective wheelbase is done by recording appropriate values obtained either by measurements or estimations when the vehicle combination is driven forwards during a time interval, where the time interval may be between a few seconds up to several minutes. The duration of the time interval may e.g. depend on the type of road on which the vehicle combination is driven. Since the steering angles, the yaw rate and/or the articulation angles are used for the estimation, it is advantageous that the vehicle combination at least drives through one curved portion of the road. However, the curve must not be continuous with a specified radius. During the time interval, the different values are determined and recorded with a predetermined sample frequency, such that a series of recorded value sets, which may be arranged in matrixes, are obtained. One or more of these recorded value sets or matrixes are then used to calculate the effective wheelbase for the towed trailers of the vehicle combination. The arrangement comprises means for calculating a value for the effective wheelbase, and the means preferably uses a least square fit to a one track linear vehicle model for the estimation of the effective wheelbase.

BRIEF DESCRIPTION OF DRAWINGS

The invention will be described in greater detail in the following, with reference to the attached drawings, in which

Fig. 1 shows a schematic vehicle combination comprising a towing vehicle and two towed trailers

Fig. 2 shows an example of a linear vehicle model, and

Fig. 3 shows a schematic flow chart of an inventive method for estimating the effective wheelbase of a vehicle combination.

MODES FOR CARRYING OUT THE INVENTION The embodiments of the invention with further developments described in the following are to be regarded only as examples and are in no way to limit the scope of the protection provided by the patent claims.

Figure 1 shows a schematic vehicle combination 1 comprising a towing vehicle 2 and two towed trailers 3, 4. The towing vehicle is preferably a regular truck adapted for commercial highway use, but could also be a tractor having a fifth wheel or a bus. The first towed trailer 3 is in the shown example a dolly having a drawbar connected to the trailer coupling of the truck. The dolly is provided with two wheel axles 7. The second towed trailer 4 is a semitrailer, which is provided with a kingpin 8 that is connected to the fifth wheel of the dolly. This example shows a common type of a longer vehicle combination, but it is also possible to estimate the effective wheelbase for other types of vehicle combinations having other types of towing vehicles and other types and numbers of towed vehicles. One such example is a vehicle combination comprising a tractor and a regular trailer or a tractor, a semitrailer and a regular trailer. A regular trailer having both a front axle and one or several rear axles can be modelled as two towed trailers. The first towed trailer comprises the drawbar and the first axle and the effective wheel base extends from the drawbar connection to the front axle, or the virtual front axle if the trailer is provided with more than one front axle. The second towed trailer comprises the trailer body and the rear axle or axles, where the effective wheel base extends from the pivot point between the trailer body and the first axle to the virtual rear axle of the trailer. In order to estimate the effective wheelbase for each towed trailer, a number of measures are recorded in value sets when the vehicle combination is driven forwards. The measures are preferably recorded when the conditions corresponds to certain predefined conditions, e.g. when the speed of the vehicle combination is within a predefined interval and when the road condition corresponds to a predefined road profile. It is e.g. advantageous to perform the recordings when the road is relatively even and plane. The time of a recording period is preferably within a predefined time interval, which may be a few seconds up to a number of minutes or more. The time period may either be preset or the recording may stop when a predefined number of approved value sets have been recorded. The recorded values are stored in a memory, preferably comprised in a control unit of the vehicle.

The measures that are recorded are the speed v _x of the truck, the steering angle of the truck, the yaw rate ri of the truck, the steering angles <¾ ¾ of each trailer, the yaw rate r ₂ , r ₃ of each trailer and the articulation angles φ _? φ ₃ of each trailer. The speed of the truck is measured in a conventional way, e.g. with a tachometer or a GPS. The steering angle of the truck is measured by a steering angle sensor. The yaw rate of the truck is either measured with a yaw rate sensor or estimated by using known truck properties. The steering angle of a trailer is the angle with which a steerable axel of the trailer steers, if the trailer is provided with a wheel steering function. The steering angle is measured by a steering angle sensor provided on the steerable axle of that trailer. The articulation angle for a trailer is the angle between the centre-line of the trailer to the centre- line of the vehicle ahead. The articulation angle is a measure of the articulation between the vehicles. The articulation angle may be measured with an articulation sensor mounted in the coupling between the vehicles or with another type of sensor. The articulation angle may e.g. be measured with optical sensors or by the use of a camera. The yaw rate of a trailer is either measured with a yaw rate sensor or is estimated by using e.g. the articulation angle.

When a suitable amount of value sets have been recorded, the effective wheelbase, L ₂ or L ₃ , is calculated for each towed trailer. The effective wheelbase is preferably estimated by using a weighted least square approximation. arg = min(L ₂ , L ₃ ) ∑f ₌₀ {i O* - ^)| + W _r (r _t - r W ² } (1) where φ _ι is a two elements vector comprising the measured articulation angles, for time-step /, and <Pi is a two elements vector comprising the estimated articulation angles using a linear vehicle model, for time step /. Analogously, r _t is a three elements vector comprising the recorded yaw rates and f _t is a three elements vector comprising the estimated yaw rates using the linear vehicle model. Further, _φ and W _r are arbitrary diagonal weighting matrices. The recorded steering angles S _1: <½ and <% serves as inputs to the linear vehicle model. The summation is made over the complete time series for which recordings have been made.

In the method, either the yaw rate of the towing vehicle and the towed trailers are required, or the articulation angles of the towed trailers are required. One special case is when all elements in W _v are set identically to zero, which means that the articulation angles are no longer part of the optimization. Thus, the measurements of articulation angles are no longer required. Analogously, when all elements in W _r are set identically to zero, which means that the yaw rates are no longer part of the optimization. Thus, the measurements of yaw rates are no longer required. Thus, either the yaw rates or the articulation angles are required in order to obtain a good estimation of the effective wheel bases of the towed trailers. However, using both the yaw rates and the articulation angles may give a result with a higher accuracy.

With the estimated effective wheelbase of each towed trailer, it is possible to perform an automatic reversing with the vehicle combination without knowing the effective wheelbase of each towed trailer in advance. Such estimation is advantageous in that it takes account of the actual properties of the vehicle combination. The effective wheelbase for a trailer will vary e.g. depending on the load and the position of the load. By this method, the effective wheelbase will be adapted to the actual load of a trailer. Further, when a multiple of different trailers are used at different occasions, the effective wheelbase will be estimated for the actual vehicle combination being used. The driver does not have to measure the different trailers or insert wheelbase values for the specific trailer into the control system.

The estimation can be performed automatically at specific occasion, such as every time the ignition is started, every time the vehicle has been stopped for a predetermined time, when the load has been changed by a predetermined amount, or at specified geographical positions such as loading docks. A further advantage of the method is that all required measures are already known to the control system of the vehicle, except for the articulation angles. Thus, the only required additional component is an articulation sensor for each towed trailer. An articulation sensor may e.g. be integrated in the coupling or the fifth wheel where the following towed trailer is attached to the previous vehicle. A device for measuring the articulation angle can also be placed on the forward vehicle, such as the truck, and can measure from a distance. Thereby it is possible to obtain information about the effective wheelbase of a trailer without providing the trailer with an extra sensor. In Fig. 2, a linear vehicle model of a truck-dolly-semitrailer vehicle combination is shown. The vehicle combination contains three rigid bodies attached by joints which have one rotational degree of freedom. The following state equations are derived for lateral and yaw motion: ^m i (v _y] + v _x r _l ) =∑F _yl

I ₂ A =∑F>,

mi {Vy2 + v _x r ₂ ) = F _y2 where m, and / _z/ is the mass and yaw inertia for unit /. The v _yi and η are the lateral velocity and yaw rate of each unit. The total lateral forces and yaw moment are built up by a linear tyre force model. By eliminating the joint forces and the kinematic constrains, a linear vehicle combination model in state space form can be written as x = Ax + Bu u = [δ δ ₂ £> ₃ ] where u in this case represents the driver steering input, i.e. the steering angle of the truck, the steering angle of the dolly, and steering angle of the semitrailer. The matrixes A and B are depending on forward speed, tyre characteristics and the effective wheelbases. This type of model is well known as a linear one track model for an articulated vehicle. Similar models are extensively used in the field of vehicles and are therefore not described in more detail.

Given the vehicle model, equation (1) can be solved using an existing solver technique, such as e.g. the Monte-Carlo method. Fig. 3 shows a schematic flow chart of the method for estimating the effective wheelbase length of a trailer in a vehicle combination comprising a towing vehicle and at least one towed trailer. The method is performed while driving the vehicle combination forwards for a predetermined time interval. In step 100, the speed of the vehicle combination is obtained and recorded. The speed of the truck is known to the control system and is preferably measured by the tachometer.

In step 110, the steering angle of the towing vehicle is recorded. The steering angle is known to the control system of the truck and is preferably measured by a steering angle sensor.

In step 120, the yaw rate of the towing vehicle is recorded. The actual yaw rate can be determined either by measuring the actual yaw rate with a yaw rate sensor or by estimating the actual yaw rate by using other vehicle properties, such as wheel speed and steering angle.

In step 130, the steering angle of each towed trailer is recorded. The steering angle is the angle with which the steerable axle is steered, if the trailer is provided with a steerable axle. If a trailer is not provided with a steerable axle, this value will be set to zero. In step 140, the articulation angle for each towed trailer is recorded. The articulation angle is the angel of the coupling between two vehicles, and is a measure of the difference between the centre line of the forward vehicle and the following vehicle. The articulation angle can be measured with a specific articulation angle sensor arranged in the coupling, or by using another type of sensor, based e.g. on optical principles.

In step 150, the yaw rate for each towed trailer is recorded. The actual yaw rate can be determined either by measuring the actual yaw rate with a yaw rate sensor or by estimating the actual yaw rate by using other vehicle properties, such as the articulation angles of the towed trailers. In step 160, a value for the effective wheelbase for each towed trailer is calculated by using the recorded values. The calculation of the effective wheelbase is preferably done by using a least square approximation, although other approximations may also be used. The steps 100 to 150 are repeated for a predetermined number of times during a predetermined time interval. When the predetermined time interval is ended, the calculation of the effective wheelbase in step160 may be performed one or several times, using one or more recorded data sets from the measurements. It is e.g. possible to set one or more threshold values for the data set such that only values that will give a good estimation are used. It is e.g. not possible to obtain a good effective wheelbase estimation when an articulation angle is 0.1 degrees, while such a small value may actually be induced by noise or other disturbances. It may thus be preferred to select and pick the recorded data sets that will give an adequate effective wheelbase.

The invention is not to be regarded as being limited to the embodiments described above, a number of additional variants and modifications being possible within the scope of the subsequent patent claims.

REFERENCE SIGNS

1 : Vehicle combination

2: Truck

3: Dolly

4: Semitrailer

5: Trailer coupling

6: Dolly virtual axle

7: Dolly axles

8: King pin

9: Semitrailer virtual axle

10: Semitrailer axles