Method for carrying out a test run on a vehicle test bench

Abstract

The aim of the invention is to make carrying out a test run (P) with a vehicle (2) on a vehicle test bench (1) more reliable using GNSS signals (GS). This aim is achieved in that the vehicle (2) determines a comparative position (VP) of the vehicle (2) from sensor data (SD) from at least one further vehicle sensor (23) while the test run (P) is being carried out, and the vehicle (2) transmits the determined comparative position (VP) to the vehicle test bench (1), a change in position (ΔP) of the vehicle (2) is determined from the specified test run (P), and the target value (SWS) for the GNSS signal generator (22) is determined from the comparative position (VP) and the change in position (ΔP).

Description

[0001] Procedure for carrying out a test run on a vehicle test bench

[0002] The invention relates to a method for carrying out a predetermined test run with a vehicle on a vehicle test bench, wherein the vehicle is arranged in a stationary manner on the vehicle test bench and a drive train of the vehicle is connected to at least one load machine on the vehicle test bench, and when carrying out the test run according to the specifications of the test run, a temporal sequence of at least one setpoint value for the operation of the drive train is determined, which is set by the drive train, and a temporal sequence of at least one setpoint value for the load machine is determined, which is set by the load machine, and a temporal sequence of at least one setpoint value for a GNSS signal generator is determined, from which the GNSS signal generator generates a temporal sequence of GNSS signals and transmits them to the vehicle test bench, wherein a position sensor is used on the vehicle which receives the transmitted GNSS signal,to determine the vehicle's own position from the received GNSS signal. The invention also relates to a corresponding vehicle test bench.

[0003] Modern vehicles incorporate a multitude of control systems, driver assistance systems, and other systems, such as entertainment systems, navigation systems, charging station reservation systems for electric vehicles, etc. These are generally referred to as vehicle systems below. Some of these vehicle systems require the precise position of the vehicle to function properly. An example of such a vehicle system (control system) is a hybrid drive control system, which specifies a driving strategy for the hybrid drive, for example, to anticipate the energy requirements of the route ahead (e.g., uphill or downhill gradients) and incorporate this into the selection of the hybrid drive's operating point. A navigation system, as a vehicle system, naturally also requires vehicle position data.In the case of driver assistance systems as a vehicle system, uninterrupted knowledge of the vehicle's own position allows the vehicle to support functions such as automatic speed selection, automated steering, etc. This often occurs in combination with the vehicle's environmental sensors, such as ultrasound, radar, lidar, or camera, and often also in combination with a recorded driving state that essentially describes the vehicle's dynamics, such as position, speed, acceleration (including in different directions), and angular acceleration (including in different spatial axes). Driving state sensors, such as a position sensor, acceleration sensor, attitude sensor, etc., are also typically installed on the vehicle.If detailed (often centimeter-accurate) map information (so-called HD maps) is stored in the vehicle, the safety of vehicle control can be increased by the driver assistance system, such as a lane keeping assistant, autonomous emergency braking system, etc., with the help of the exact and reliably determined position of the vehicle, even in poor visibility or lighting.

[0004] Thanks to the driving environment sensors and driving condition sensors provided on the vehicle, the vehicle has continuous knowledge of the vehicle environment and the driving condition during normal operation.

[0005] In order to test such functions of vehicle systems developed for modern vehicles, tests are carried out on a vehicle test bench, for example. The vehicle or the part of the vehicle to be tested is spatially fixed, for example, on a chassis dynamometer or powertrain test bench, and the vehicle's powertrain (or the part of the powertrain to be tested) can actually accelerate and brake. In the case of a chassis dynamometer, the vehicle is mounted on rollers, at least with the driven wheels. Steering is also sometimes possible on the vehicle test bench, although this obviously does not result in any steering movement of the vehicle.To simulate the loads of real-world operation, a load machine (dyno) is provided that acts on the drivetrain, for example, in the form of a roller on a chassis dynamometer or a load machine connected to a semi-axle of the drivetrain on a drivetrain test bench. Such vehicle test benches are well known and require no further explanation.

[0006] In order to realistically test a vehicle system, such as a control and driver assistance system, on a vehicle test bench, it is necessary to provide the vehicle's driving environment sensors and driving state sensors with information that matches the image of the vehicle's environment and driving state available internally in the vehicle. For example, for the vehicle's camera sensors on the vehicle test bench, this is achieved through suitable image projections in front of and / or next to the vehicle, or for radar sensors through suitable echo simulation.

[0007] To determine their own position, vehicles generally use a position sensor that detects GNSS (Global Navigation Satellite System) signals, such as GPS (Global Positioning System) or Galileo. Due to the static fixation of the vehicle on the vehicle test bench, the GNSS signals received on the vehicle test bench do not match the journey simulated as a test run, or often no GNSS signals can be received within the vehicle test bench at all. This can be remedied on the vehicle test bench, for example, with GNSS signal generators. These generate GNSS signals on the vehicle test bench that match the test run performed on the vehicle test bench. If, for example, a journey from a starting point to a destination over a certain distance is simulated as a test run, the GNSS signal generator generates GNSS signals that correspond to the simulated route.The GNSS signal generator on the vehicle test bench is usually controlled by a simulation of the test run, in which the (virtual) movement of the vehicle is reproduced in a simulation, and from this the (virtual) position of the vehicle at a specific time during the test run is determined. This virtual position is passed on as a target value to the GNSS signal generator, which then generates the GNSS signal corresponding to the target position, which can then be received and evaluated by the vehicle. From the signal generated by the GNSS and received by the vehicle, the vehicle derives its position (supposedly measured in reality via GNSS), which is used by a vehicle system when carrying out the test run. The vehicle therefore always has an internal "idea" of its own current position. A procedure of this kind is known, for example, from EP 3 121 729 A1, as well as from Geneder S., et al., "Development of Connected Powertrains at the Power Test Bed", ATZ worldwide 116(6): 14-19. CN 117647404 A shows a chassis dynamometer for a vehicle with a GNSS signal generator that generates a GNSS signal for the vehicle on the chassis dynamometer according to a specified test run. CN 111949544 A also shows the testing of an autonomous vehicle using GNSS signals generated by a GNSS signal generator. CN 115 112394 A shows a test bench for a vehicle on which position signals can be generated.

[0008] The fundamental problem with this approach is determining and specifying the target values ​​of the GNSS signals to be generated in such a way that they do not grossly contradict the driving environment and driving state assumed in the vehicle based on other driving environment sensors and driving state sensors. If this fails, and the vehicle detects a position inconsistency, this can have negative effects on the test run currently in progress, for example, because the vehicle then assumes an error and enters an emergency program that prevents the test run from continuing and thus the testing of the control or driver assistance system. Currently, attempts are being made to solve this problem using heuristic methods. However, such heuristic methods are complex and not very reliable.

[0009] It is therefore an object of the present invention to make the execution of a test run on a vehicle test bench using GNSS signals more reliable.

[0010] This object is achieved according to the invention in that, during the test run, the vehicle determines a reference position of the vehicle from sensor data from at least one other vehicle sensor and transmits the determined reference position to the vehicle test bench. A change in the vehicle's position is determined from the specified test run. The reference position and the position change are used to determine the target value for the GNSS signal generator. The target value for the GNSS signal generator is, of course, determined at each point in time in order to determine the temporal sequence of target values ​​for the GNSS signal generator.The inventive procedure ensures that, during the test run, the self-position determined from the GNSS signal and the comparison position are always kept within an acceptance interval (essentially defined by the accuracy of the determined position change, which can be influenced), so that no inconsistencies can arise between the self-position and the comparison position in the vehicle. This ensures that, even for vehicles with a plausibility check of the self-position, the plausibility check during a test run on a vehicle test bench does not lead to any inconsistencies that could interrupt or influence the test run.

[0011] The present invention will be explained in more detail below with reference to Figures 1 to 2, which show exemplary, schematic and non-limiting advantageous embodiments of the invention.

[0012] Fig.1 a vehicle test bench according to the invention and

[0013] Fig.2 is a flow chart of a test run sequence according to the invention.

[0014] Fig. 1 shows a vehicle test bench 1 in the form of a roller dynamometer for carrying out a test run with a vehicle 2. A test run P is essentially a simulated drive with the vehicle along a specific route. In the embodiment according to Fig. 1, at least one vehicle wheel 3 of the vehicle 2, driven by a drive train 6 of the vehicle 2, is arranged on the roller 4 driven by a roller drive 7. It is possible to assign the vehicle wheels 3 of a driven axle of the vehicle 2 to a roller 4 or each driven vehicle wheel 3 to its own roller 4, whereby each roller 4 can be driven independently by a roller drive 7. A roller dynamometer is also conceivable in which vehicle wheels 3 of different axles are arranged on at least one roller 4 per axle (as indicated by dashed lines in Fig. 1). The vehicle 2 is arranged in a stationary manner on the vehicle test bench 1, as shown in Fig. 1 by a holder 5.Due to its stationary arrangement, the vehicle 2 essentially does not move relative to the vehicle test bench 1. A movement of the vehicle 2 in the form of a journey with the vehicle 2 is simulated during the test run P by the vehicle wheel 3, driven by the drive train 6 of the vehicle 2, rolling on the roller 4. The roller 4 forms a loading machine (dyno) 8, which generates a load, usually a load torque, for the drive train 6 via the roller drive 7. The drive train 6 comprises at least one drive component 9, such as an internal combustion engine or an electric motor, which acts on the vehicle wheel 3 via transmission components such as a clutch, gearbox, differential gear, and half axles in order to transmit a torque and set a wheel speed (virtual vehicle speed).

[0015] Instead of a roller dynamometer, the vehicle test bench 1 can also be designed differently. Often, a load machine 8, such as an electric motor, is connected to each half-axle of the drive train 6, or at least to the driven half-axles, for example, via wheel hubs on the half-axles. This allows each half-axle to be individually and independently subjected to a load, such as a load moment. Such a vehicle test bench 1 is described, for example, in the paper by Geneder S. et al. mentioned above.

[0016] In general, at least one load machine 8 is connected to a component of the drive train 6 on the vehicle test bench 1, wherein the drive train 6, during operation of the vehicle when carrying out a test run P, generates a drive torque and a drive speed by means of the drive component 9, which is counteracted by the load torque of the load machine 8 acting on the drive train 6.

[0017] However, the design of vehicle test bench 1 is irrelevant to the invention. It is also irrelevant whether vehicle 2, with all its components such as superstructures, body, etc., is mounted on vehicle test bench 1. Vehicle 2 is therefore understood to be the vehicle required for performing the respective test run P.

[0018] Vehicle 2, which also refers to only a part of vehicle 2 (such as a drive train 6), is tested on vehicle test bench 1 with a test run P, whereby test run P simulates a real journey with vehicle 2 along a route. The simulated route can be a real route or a freely defined route.

[0019] A test run P is generally carried out by simulating a journey with the vehicle 2 along a predetermined route on a test bench automation unit 10. For this purpose, a driving model 11 can be used on the test bench automation unit 10 in a well-known manner, which simulates the behavior of the vehicle 2, the driver, and the environment of the vehicle 2. The driving model 11 can comprise interacting sub-models, such as a tire model, a drivetrain model, a driver model, a road model, a model of the drive component, etc. For the test run P, the route 12 is specified, for example in the form of incline, decline, geometry (straight line, curve), traffic, other road users, road signs, traffic lights, etc. Stochastically generated events, such as the random behavior of other road users, can also be taken into account.The driving route 12 serves as input for the driving model 11, and the driving model 11 determines a temporal sequence (in predetermined periods) of setpoint values ​​SWA for the operation of the drive train 6, such as a throttle valve position or injection quantity for an internal combustion engine, or a torque and a speed for an electric motor. The setpoint values ​​SWA are set on the drive train 6 of the vehicle 2 so that the vehicle 2 is operated on the vehicle test bench 1 according to the specifications of the test run P. The driving model 11 also determines a temporal sequence (in predetermined periods, which may be the same as or different from the periods of the setpoint values ​​SWA for the operation of the drive train 6) of setpoint values ​​SWB for the at least one load machine 8, such as a load torque or a load speed, which are set using the load machine 8.The target values ​​SWA and SWB are typically generated and specified in time series in the range of microseconds to milliseconds. In this way, a drive with vehicle 2 is simulated on vehicle test bench 1 with test run P, and vehicle 2 on vehicle test bench 1 experiences essentially the same conditions as during a real drive along route 12.

[0020] Instead of simulating the test run P in the test bench automation unit 10, a temporal sequence of setpoint values ​​SWA for the operation of the drive train 6 and a temporal sequence of setpoint values ​​SWB for the at least one load machine 8, which represent a certain travel distance 12, can of course also be stored as a test run P, which are called up to carry out a test run P and set on the vehicle test bench 1.

[0021] According to the invention, a vehicle system, such as a control or driver assistance system of the vehicle 2, which requires a position of the vehicle 2 for its function, is to be tested on the vehicle test bench 1. For this purpose, a control unit 20 is provided on the vehicle 2 (usually in the form of a network of individual control units), in which the function of the vehicle system to be tested is implemented, usually in the form of control software executed on control hardware (microprocessor-based hardware) of the control unit 20.

[0022] Furthermore, a position sensor 21 is provided on the vehicle 2, which receives a GNSS signal GS and generates a position signal PS therefrom, which is transmitted to the control unit 20. The position signal PS can already be a position, or a signal from which the control unit 20 can determine the position. The vehicle 2, specifically the control unit 20 of the vehicle 2, thus has knowledge of the vehicle's own position, provided it receives a GNSS signal GS. The own position on the vehicle test bench 1 is, of course, a virtual position, which results from the specifications for the test run P, such as the route 12.

[0023] Since the vehicle 2 is stationary on the vehicle test bench 1, a real GNSS signal would provide a position that does not match the test run P. Essentially, a constant position would be recorded. Typically, however, no real GNSS signal GS can be received at all on the vehicle test bench 1 because the vehicle test bench 1 is located in a test hall in which a GNSS signal GS is often not receivable. This would result in a discrepancy between the position recorded by the position sensor 21 and the (virtual) position that the vehicle 2 should have according to the test run P, for example, when virtually driving the virtual route 12. To prevent this, a GNSS signal generator 22 is provided on the vehicle test bench 1, which generates and transmits a GNSS signal GS according to the specifications of the test run P. Based on the test run P, a virtual temporal progression of the position of the vehicle 2 results (e.g.in GNSS coordinates), which are generated and transmitted at the vehicle test bench 1 using the GNSS signal generator 22 in the form of a GNSS signal GS. The vehicle 2 thus receives a GNSS signal GS at the vehicle test bench 1, which should match the test run P. Thus, the known position derived in the vehicle 2 from the GNSS signal GS should also match the test run P.

[0024] For this purpose, the test bench automation unit 10 generates a temporal sequence (in predetermined periods) of target values ​​SWS, for example target positions of the vehicle 2, for the GNSS signal generator 22 when carrying out the test run P, which are converted by the GNSS signal generator 22 into a temporal sequence of GNSS signals GS.

[0025] Based on the received GNSS signals GS at a specific time (in a specific time period), vehicle 2 determines an own position EP of vehicle 2 based on the respective position signal PS of position sensor 21 when performing the test run P. This position EP can be used by the vehicle system under test, such as a control or driver assistance system. When performing the test run P, a chronological sequence (in predefined time periods) of the own position EP of vehicle 2 is then obtained.

[0026] Since in reality, a GNSS signal GS cannot always be received by vehicle 2, or the GNSS signal GS is not always reliable, for example, in tunnels, or only a weak GNSS signal GS can be received, for example, in narrow valleys, a plausibility check is provided in the control unit 20 of vehicle 2 to detect and filter out incorrect or unreliable self-positions EP determined from a GNSS signal GS. For the plausibility check, the control unit 20 of vehicle 2 determines a comparison position VP based on sensor signals SD supplied by at least one further vehicle sensor 23 (different from position sensor 21). The at least one further vehicle sensor 23 is a driving environment sensor, such as a radar, lidar, camera sensor, or driving state sensor, such as an acceleration sensor, speed sensor, wheel speed sensor, etc., of vehicle 2.For the plausibility check, the own position EP determined from the GNSS signal GS is compared with the comparison position VP in order to detect inconsistencies. The comparison position VP can be determined, for example, based on a last reliable own position EP and the virtual movement of the vehicle 2 along the route 12 detected in the vehicle 2 using the vehicle sensor 23. For example, a wheel speed sensor could be used to determine a distance traveled or a vehicle speed. The movement of the vehicle 2, such as a speed, originates, for example, from driving condition sensors such as a wheel speed sensor and can be assumed to be known in the vehicle 2 during operation of the vehicle 2. An inconsistency occurs, for example, if the own position EP deviates too greatly (according to a predetermined maximum permissible deviation) from the comparison position VP.In case of an inconsistency, vehicle 2 can take a predetermined action, for example activate an emergency program.

[0027] An inconsistency can also arise if the vehicle also derives a position from other signals that then does not match the determined self-position EP. For example, some vehicles use information from a mobile phone network to determine their position. Therefore, for certain vehicles, it may be advantageous to shield mobile phone network signals on the vehicle test bench 1 to prevent potential negative influences on the test run P.

[0028] A plausibility check as described above is of course also carried out on a vehicle 2 on the vehicle test bench 1. When performing a test run P on the vehicle test bench 1, it is therefore important to specify the GNSS signals GS generated with the GNSS signal generator 22 in such a way as to avoid possible inconsistencies between the own position EP determined from the GNSS signal GS and a comparison position VP determined internally in the vehicle 2 from other sensor signals SD.

[0029] To ensure this, the invention proceeds as follows.

[0030] Vehicle 2 transmits the comparison position VP determined at a specific time (in a specific period) via a transmitting unit 24 to vehicle test bench 1, specifically to the test bench automation unit 10 of vehicle test bench 1. Transmission occurs, for example, via short-range communication, such as ITS-G5 or WLAN, which is often implemented in vehicles 2. If necessary, the transmitting unit 24 could be retrofitted to vehicle 2 for implementation on a vehicle test bench 1. The transmitting unit 24 receives the comparison position VP from the control unit 20. Transmission can occur using a standardized communication protocol or a proprietary communication protocol.

[0031] Based on the specifications of the test run P, in particular based on the dynamics (speed, acceleration) of the vehicle movement simulated during the test run, a position change AP of the vehicle 2, which the vehicle 2 travels until the next journal based on the specified route 12, can be determined in the test bench automation unit 10 at each journal. Using the received comparison position VP and the position change AP, a new target value SWS for the GNSS signal generator 22 can then be determined. According to the invention, the target value SWS for the GNSS signal generator 22 is therefore no longer determined directly from the specifications of the test run P (such as the route 12), but from the comparison position VP supplied by the vehicle 2 and a position change AP derived from the specifications of the test run P. The inventive procedure is also shown in Fig. 2 in the form of a flow chart.

[0032] In the simplest case, the target value SWS for the GNSS signal generator 22 results from the sum, in particular as a vector sum, of the received comparison position VP and the position change AP. In order to avoid excessively large position jumps, for example because the position change AP becomes too large in a particular time step, the determined target value SWS could also be looped in via a filter, for example as an average of the target values ​​SWS over at least two consecutive time steps. If the difference between the own position EP and the comparison position VP becomes too large, which can be controlled by a predetermined limit value, this could also be reported back to the test bench automation unit 10 that the vehicle 2 is "not keeping up". In this case, for example, the test run P, such as the route 12, could also be adapted.

[0033] The procedure according to the invention ensures that, when performing the test run P, the own position EP determined from the GNSS signal GS (which is now derived from the comparison position VP of vehicle 2) and the comparison position VP are always kept within an acceptance interval, so that no inconsistencies can arise between the own position EP and the comparison position VP in vehicle 2. This also ensures that, for vehicles 2 with a plausibility check as explained above, the plausibility check does not lead to any inconsistencies when performing a test run P on a vehicle test bench 1.

Claims

Patent claims 1. A method for carrying out a predetermined test run (P) with a vehicle (2) on a vehicle test bench (1), wherein the vehicle is arranged in a stationary manner on the vehicle test bench (1) and a drive train (6) of the vehicle (2) is connected to at least one load machine (8) on the vehicle test bench (1), and when carrying out the test run (P) according to the specifications of the test run (P), a temporal sequence of at least one setpoint value (SWA) for the operation of the drive train (6) is determined, which is set by the drive train (6), and a temporal sequence of at least one setpoint value (SWB) for the load machine (8) is determined, which is set by the load machine (8), and a temporal sequence of at least one setpoint value (SWS) for a GNSS signal generator (22) is determined, from which the GNSS signal generator (22) generates a temporal sequence of GNSS signals (GS) and on the vehicle test bench (1) sends,wherein a position sensor (21) is used on the vehicle (2), which receives the transmitted GNSS signal (GS) in order to determine an own position of the vehicle (2) from the received GNSS signal (GS), characterized in that the vehicle (2) determines a comparison position (VP) of the vehicle (2) from sensor data (SD) of at least one further vehicle sensor (23) during the execution of the test run (P), and the vehicle (2) sends the determined comparison position (VP) to the vehicle test bench (1), that a position change (AP) of the vehicle (2) is determined from the predetermined test run (P), and that the target value (SWS) for the GNSS signal generator (22) is determined from the comparison position (VP) and the position change (AP).

2. Method according to claim 1, characterized in that a time sequence of at least one setpoint value for a drive component (9) of the drive train (6) is determined as the time sequence of the at least one setpoint value (SWA) of the drive train (6), which setpoint value is set by the drive component (9) of the drive train (6).

3. Vehicle test bench for carrying out a predetermined test run (P) with a vehicle (2), wherein the vehicle (2) is arranged in a stationary manner on the vehicle test bench (1) and a drive train (6) of the vehicle (2) is connected to at least one load machine (8) on the vehicle test bench (1), wherein a test bench automation unit (10) is provided on the vehicle test bench (1), which, when carrying out the test run (P), determines a time sequence of at least one setpoint value (SWA) for the operation of the drive train (6) in accordance with the specifications of the test run (P), and determines a time sequence of at least one setpoint value (SWB) for the load machine (8), which sets the load machine (8), and determines a time sequence of at least one setpoint value (SWS) for a GNSS signal generator (22), from which the GNSS signal generator (22) generates a temporal sequence of GNSS signals (GS) and transmits them to the vehicle test bench (1), wherein a position sensor (21) is provided on the vehicle (2), which receives the GNSS signal (GS) and generates a position signal (PS), and a control unit (20) of the vehicle (2) is provided to determine an own position of the vehicle (2) based on the position signal (PS), characterized in that the control unit (20) of the vehicle (2) receives sensor data (SD) from at least one further vehicle sensor (23) of the vehicle (2) during the execution of the test run (P), and the control unit (20) determines a comparison position (VP) of the vehicle (2) from the sensor data (SD), and the vehicle (2) is configured to transmit the determined comparison position (VP) to the test bench automation unit (10) of the vehicle test bench (1), that the test bench automation unit (10) determines a change in position (AP) of the vehicle (2) from the specified test run (P),and that the test bench automation unit (10) determines the target value (SWS) for the GNSS signal generator (22) from the obtained comparison position (VP) and the determined position change (AP).

Images