ASCERTAINING AN INPUT VARIABLE OF A VEHICLE ACTUATOR USING A MODEL-BASED PREDICTIVE CONTROL

Abstract

The disclosure relates to the process of ascertaining an input variable of a vehicle actuator using a model-based predictive control. According to one exemplary arrangement, a processor unit is designed to access trajectory information and a state data set, which represents a state of surroundings of a vehicle and/or the state of the vehicle and/or a driving state of the vehicle, by an interface. The processor unit carries out a secondary condition algorithm in order to calculate a secondary condition and an MPC algorithm for a model-based predictive control. By carrying out the secondary condition algorithm, a secondary condition is ascertained for the MPC algorithm on the basis of the trajectory information and on the basis of the state data set. By carrying out the MPC algorithm, an input variable is ascertained for an actuator of the vehicle on the basis of the secondary condition. This is carried out in particular such that in a future predicted trajectory, the vehicle follows the specified trajectory with a specified degree of reliability.

Claims

1. A processor unit for ascertaining an input variable of a vehicle actuator model-based predictive control (“MPC”), the processor unit including an interface, the processor unit having been set up to: access, by the interface, trajectory information and a state data record that represents an ambient state of the vehicle and/or a state of the vehicle and/or a running state of the vehicle, execute an auxiliary-condition algorithm for calculating an auxiliary condition, execute an MPC algorithm for model-based predictive control, ascertain an auxiliary condition for the MPC algorithm by executing the auxiliary-condition algorithm as a function of the trajectory information and as a function of the state data record, ascertain an input variable for an actuator of the vehicle by executing the MPC algorithm as a function of the auxiliary condition, so that in a future predicted trajectory, the vehicle follows the predetermined trajectory with a predetermined safety.

2. The processor unit as claimed in claim 1, wherein the state data record comprises a friction-coefficient data record that represents a quality of a tire/roadway contact.

3. The processor unit as claimed in claim 2, the processor unit having been set up to receive the friction-coefficient data record from a sensor unit, in which connection to the sensor unit registers quality of the tire/roadway contact.

4. The processor unit as claimed in claim 2, the processor unit having been set up to ascertain the friction-coefficient data record statistically.

5. The processor unit as claimed in claim 1, wherein the processor unit is set up to cause the actuator to be acted upon by the ascertained input variable.

6. The processor unit as claimed in claim 1, wherein the processor unit is set up to ascertain the auxiliary condition as a function of an entirety of the predetermined trajectory by executing the auxiliary-condition algorithm.

7. A system for ascertaining an input variable of a vehicle actuator by model-based predictive control, the system including a processor unit as claimed in claim 1.

8. The system as claimed in claim 7, the system further including a sensor unit, the sensor unit being set up to register an ambient state of the vehicle to generate, on the basis thereof, the state data record, and to transmit said data record to the processor unit.

9. The system as claimed in claim 8, the sensor unit registering the quality of the tire/roadway contact by a sensor to generate, on the basis thereof, a friction-coefficient data record, and to transmit friction-coefficient data record to the processor unit.

10. (canceled)

11. A method for ascertaining an input variable of a vehicle actuator using model-based predictive control, the method comprising the following steps: accessing trajectory information and a state data record by an interface of a processor unit, the state data record representing one of an ambient state of the vehicle and a state of the vehicle and a running state of the vehicle, executing an auxiliary-condition algorithm for calculating an auxiliary condition by the processor unit, executing an MPC algorithm for model-based predictive control by the processor unit, ascertaining an auxiliary condition for the MPC algorithm by executing the auxiliary-condition algorithm as a function of the trajectory information and as a function of the state data record, and ascertaining an input variable for an actuator of the vehicle by executing the MPC algorithm as a function of the auxiliary condition, so that in a future predicted trajectory the vehicle follows the predetermined trajectory with a predetermined safety.

12. A computer-program for ascertaining an input variable of a vehicle actuator using model-based predictive control, the computer-program, when executed in a processor unit, instructing the processor unit to: access, via an interface, trajectory information and a state data record that represents at least one of an ambient state of the vehicle and a state of the vehicle and a running state of the vehicle, execute an auxiliary-condition algorithm for calculating an auxiliary, execute an MPC algorithm for model-based predictive control, ascertain an auxiliary condition for the MPC algorithm by executing the auxiliary-condition algorithm as a function of the trajectory information and as a function of the state data record, and ascertain an input variable for an actuator of the vehicle by executing the MPC algorithm as a function of the auxiliary condition, so that in a future predicted trajectory the vehicle (1) follows the predetermined trajectory with a predetermined safety.

13. The processor unit as claimed in claim 4, the processor unit having been set up to ascertain the friction-coefficient data record stochastically.

14. The system as claimed in claim 7, the system further including a sensor unit, the sensor unit being set up to register a state of the vehicle to generate, on the basis thereof, the state data record, and to transmit said data record to the processor unit.

15. The system as claimed in claim 7, the system further including a sensor unit, the sensor unit being set up to register a running state of the vehicle to generate, on the basis thereof, the state data record, and to transmit said data record to the processor unit.

16. The method of claim 11, wherein the processor instructs the actuator to perform an adjustment corresponding to the input variable.

Description

BRIEF DESCRIPTION OF DRAWINGS

[0028] Exemplary arrangements of the disclosure will be elucidated in more detail in the following with reference to the drawings which are schematic and not true to scale, wherein like or similar elements have been provided with the same reference symbol. Shown in the drawings are:

[0029] FIG. 1 a plan view of a vehicle with a system for controlling a vehicle by model-based predictive control.

[0030] FIG. 2 a plan view of a roadway on which the vehicle according to FIG. 1 is traveling autonomously or semi-autonomously, and

[0031] FIG. 3 modules for ascertaining an input variable of an actuator of the vehicle according to FIG. 1.

DETAILED DESCRIPTION

[0032] FIG. 1 shows a vehicle 1. The vehicle 1 includes a system 2 for ascertaining an input variable of a vehicle actuator and for controlling a vehicle by model-based predictive control. In the exemplary arrangement shown, the system 2 comprises a processor unit 3, a memory unit 4, a communications interface 5 and a sensor unit 6. In the exemplary arrangement shown, the sensor unit 6 includes a digital camera system which is designated in the following by “camera”. The camera 6 may be arranged on the vehicle in such a manner that a field of view of the camera 6 can capture an external surrounding field of the vehicle and optionally a roadway 7 (FIG. 2) on which the vehicle 1 is moving. The sensor unit 6 may include a number of further sensors 16 which, for instance, have been set up to ascertain the position of the vehicle, the speed, yaw-rate or lateral acceleration thereof, and to provide these measurements to the processor unit 3. The processor unit 3 can access this information in the form of a state data record 17, for instance via the communications interface 5.

[0033] The camera 6 records images successively and continuously for example, also of the roadway 7—as a result of which a video or an image sequence also of the roadway 7 is generated. Within the individual images of this video or of this image sequence, image details can be defined (regions of interest, ROI for short), within which a current coefficient of friction, for example, between the tires of the vehicle 1 and the road 7 can be detected by methods of image recognition.

[0034] Alternatively, the current coefficient of friction between the tires of the vehicle 1 and the roadway 7 can also be ascertained by stochastic methods or can be transmitted to the vehicle 1 from another vehicle, from a traffic-infrastructure device or from a traffic server, so that the processor unit 3 can likewise access it in the form of the state data record.

[0035] Moreover, the vehicle 1 includes several actuators. A first actuator 8, a second actuator 9 and a third actuator 10 are represented by way of example in FIG. 1. The first actuator 8 can control the steering of the vehicle 1 (steering actuator 8). The second actuator 9 can control the drive of the vehicle 1 (drive actuator 9). The third actuator 10 can control the brakes of the vehicle 1 (brake actuator 10).

[0036] A computer-program 11 may be stored in the memory unit 4. The computer-program 11 can be executed in the processor unit 3, for which purpose the processor unit 3 and the memory unit 4 have been connected to one another by the communications interface 5. When the computer-program product 11 is executed in the processor unit 3, it instructs the processor unit 3 to perform the functions described in connection with the drawing or, to be more precise, to execute steps of the method.

[0037] FIG. 2 shows the vehicle 1 on the roadway 7 of a road. In the example shown in FIG. 2, the vehicle 1 is traveling on the roadway 7 at a speed of 20 m/s. The roadway 7 has a first route segment 12 and a second route segment 13 which directly adjoins the first route segment 12. The first route segment 12 runs straight and is 50 m long. The second route segment 13 is a left-hand curve.

[0038] FIG. 3 shows schematically individual components and algorithms of the computer-program 11 for ascertaining an input variable 19 of at least one of the vehicle actuators 8, 9 and 10 by model-based predictive control. The computer-program 11 includes an auxiliary-condition algorithm 14 (feasible-states estimator), for calculating an auxiliary condition 18, and an MPC algorithm 15 for calculating an input variable 19, for instance for the brake actuator 10 of the vehicle 1. Moreover, the computer-program 11 includes a trajectory algorithm 20 for calculating an MPC reference trajectory P from trajectory information P.

[0039] The interface 5 of the processor unit 3 accesses the state data record 17 which contains, for instance, information about the current position and speed of the vehicle 1 and also about the coefficient of friction of the roadway 7 on the curve 13. Moreover, the interface 5 of the processor unit 3 accesses trajectory information P and a trajectory P, calculated by the trajectory algorithm, for the vehicle 1. In the example according to FIG. 2, the trajectory information P, which is supplied to the auxiliary-condition algorithm 14, implies that the designated course of the journey is firstly undertaken on the straight route segment 12 straight ahead for 50 m and then curved to the left in accordance with the progression of the curve 13. The trajectory P calculated by the trajectory algorithm is shorter than the trajectory information P and consequently contains only a fraction of the trajectory information P.

[0040] If the prediction horizon of the MPC algorithm 15 amounts to 1 s, for instance, an MPC control without auxiliary conditions for a safe range of states can only react to the approaching left-hand curve 13 if the latter lies within the prediction horizon. However, depending upon the current tire/roadway contact, this may already be too late in order to reduce the speed of the vehicle 1 to such an extent that a safe negotiating of the curve 13 is possible. With a view to solving this problem, the processor unit 3 executes the auxiliary-condition algorithm 14 and thereby ascertains at least one auxiliary condition 18 for the MPC algorithm 15 as a function of the state data record 17 and as a function of the trajectory information P. By virtue of the auxiliary-condition algorithm 14, auxiliary conditions 18 can be calculated that constitute an outer boundary of the state space that can be realized by the vehicle 1 on the trajectory P. The at least one auxiliary condition 18 for the MPC algorithm 15 is ascertained in such a manner that the MPC control is in a safe state or in an implementable state. Consequently a safe or possible auxiliary condition 18 of the state for the MPC control is ascertained. In other words, the at least one auxiliary condition 18 delimits a safe state space that can be provided to the MK algorithm 15 as auxiliary condition for the predetermined trajectory P.

[0041] The MPC algorithm 15 comprises a model, by which a trajectory of the vehicle 1 in the future is calculated on the basis of a calculated future input variable of at least one actuator 8, 9, 10. This calculated or predicted trajectory is compared with the predetermined trajectory P and is optimized, by adaptation of a future manipulated variable or input variable 19 of the actuator 8, 9 and/or 10 in question (for example, steering angle, braking torque, drive torque, etc.), in such a manner that the predicted trajectory comes as close as possible to the predetermined trajectory.

[0042] The processor unit 3 executes the MPC algorithm 15 and thereby ascertains an input variable 19, for instance for the brake actuator 10 of the vehicle 1, as a function of the auxiliary condition 18, so that in a future predicted trajectory the vehicle 1 follows the predetermined trajectory with a predetermined safety. The processor unit 3 furthermore instructs the actuator 10 to perform an adjustment corresponding to the input variable 19. If, for instance, the state data record 17 contains the information that it will be very slippery on the curve 13, the auxiliary condition 18 can then be ascertained by the auxiliary-condition algorithm 14 in such a manner that the input variable 19, ascertained by execution of the MPC algorithm 15, for the brake actuator 10 provides for braking the vehicle 1 sufficiently in order to guarantee a safe passage through the curve 13 and to prevent the vehicle 1 from entering the region of the curve at too high a speed.