Adapting a Gain Factor of an Acceleration Controller for a Motor Vehicle

Abstract

Methods and devices for adapting a gain factor of an acceleration controller for a motor vehicle are provided. An acceleration controller specifies an acceleration setpoint for the motor vehicle in a time increment. The acceleration setpoint is specified as a function a speed setpoint of the motor vehicle, an actual speed of the motor vehicle, and the gain factor. The device stores the speed setpoint, the actual speed, and the acceleration setpoint specified as information for at least two time increments, select a first subset of the information, and train a model as a function of the first subset. The model predicts an actual speed in a later time increment from at least one stored actual speed and at least one stored acceleration setpoint, select a second subset of the information, and adapt the gain factor as a function of the second subset, the model and the acceleration controller.

Claims

1.-8. (canceled)

9. A device for adapting a gain factor of an acceleration controller for a motor vehicle, comprising: the acceleration controller configured to specify an acceleration setpoint for the motor vehicle in a time increment as a function of: a speed setpoint of the motor vehicle, an actual speed of the motor vehicle, and the gain factor, wherein the device is configured to: store the speed setpoint, the actual speed, and the acceleration setpoint specified as a function thereof in each case as information for at least two time increments, select a first subset of the information, train a model as a function of the first subset, wherein the model is configured to predict an actual speed in a later time increment from at least one stored actual speed and at least one stored acceleration setpoint, select a second subset of the information, and adapt the gain factor as a function of the second subset, the model and the acceleration controller.

10. The device according to claim 9, wherein the acceleration controller is further configured to determine the acceleration setpoint from the product of the gain factor and the difference between the speed setpoint and the actual speed.

11. The device according claim 9, wherein the device is further configured to store the information in a ring buffer, and a capacity of the ring buffer is limited to storing the information from at most 5000 time increments.

12. The device according to claim 9, wherein the device is further configured to: train the model by a first weighting factor and a second weighting factor optimized in such a way that a prediction error of the model is minimized, wherein the first weighting factor specifies an influence of the at least one stored actual speed on the prediction, and the second weighting factor specifies an influence of the at least one stored acceleration setpoint on the prediction.

13. The device according to claim 9, wherein the device is further configured to: adapt the gain factor by the device being set up; predict a state of the motor vehicle as a function of the second subset, the model and the acceleration controller; and adapt the gain factor such that a measure of controller quality based on the state of the motor vehicle is minimized.

14. The device according to claim 13, wherein the state of the motor vehicle comprises the actual speed of the motor vehicle and/or the acceleration setpoint of the motor vehicle in at least one time increment.

15. The device according to claim 9, wherein the device is further configured to: store the information in a ring buffer, wherein a capacity of the ring buffer is limited to storing the information from at most 5000 time increments; train the model by a first weighting factor and a second weighting factor being optimized with a Levenberg-Marquardt algorithm such that a prediction error of the model is minimized, wherein the first weighting factor specifies an influence of the at least one stored actual speed on the prediction, and the second weighting factor specifies an influence of the at least one stored acceleration setpoint on the prediction; adapt the gain factor by predicting a state of the motor vehicle as a function of the second subset, the model, and the acceleration controller; and optimize the gain factor with a Levenberg-Marquardt algorithm such that a measure of controller quality based on the state of the motor vehicle is minimized.

16. A method for adapting a gain factor of an acceleration controller for a motor vehicle, comprising: specifying, by the acceleration controller an acceleration setpoint for the motor vehicle in a time increment as a function of a speed setpoint of the motor vehicle, an actual speed of the motor vehicle and the gain factor; storing the speed setpoint, the actual speed and the acceleration setpoint specified as a function thereof as information for at least two time increments; selecting a first subset of the information; training a model as a function of the first subset, wherein the model is configured to predict an actual speed in a later time increment from at least one stored actual speed and at least one stored acceleration setpoint, selecting a second subset of the information; and adapting the gain factor as a function of the second subset, the model, and the acceleration controller.

Description

BRIEF DESCRIPTION OF THE DRAWING

[0046] FIG. 1 shows a device according to the present subject matter for adapting a gain factor VF of an acceleration controller BR for a motor vehicle.

DETAILED DESCRIPTION

[0047] The acceleration controller BR is set up to specify an acceleration setpoint SB for the motor vehicle in a time increment as a function of a speed setpoint SG of the motor vehicle, an actual speed IG of the motor vehicle, and the gain factor VF.

[0048] In addition, the acceleration controller BR is set up to determine the acceleration setpoint SB from the product of the gain factor VF and the difference between the speed setpoint SG and the actual speed IG.

[0049] The device is set up to store the speed setpoint SG, the actual speed IG, and the acceleration setpoint SB specified as a function thereof in each case as information for at least two time increments.

[0050] The device is set up to store the information in a ring buffer RS, wherein a capacity of the ring buffer RS is limited to storing the information from at most 5000 time increments.

[0051] In addition, the device is set up to select a first subset ET of the information and to train a model MU as a function of the first subset ET, wherein the model MU is set up to predict an actual speed IG in a later time increment from at least one stored actual speed IG and at least one stored acceleration setpoint SB.

[0052] The device is set up to train the model MU by a first weighting factor and a second weighting factor being optimized such that a prediction error of the model MU is minimized, wherein the first weighting factor specifies an influence of the at least one stored actual speed IG on the prediction, and wherein the second weighting factor specifies an influence of the at least one stored acceleration setpoint SB on the prediction.

[0053] In addition, the device is set up to select a second subset ZT of the information and to adapt the gain factor VF as a function of the second subset ZT, the model MU, and the acceleration controller BR, for example by using an optimization means CU.

[0054] The device is set up to adapt the gain factor VF by the device being set up to predict a state of the motor vehicle as a function of the second subset ZT, the model MU, and the acceleration controller BR and to adapt the gain factor VF such that a measure of controller quality based on the state of the motor vehicle is minimized.

[0055] The state of the motor vehicle comprises at least one actual speed IG of the motor vehicle and/or at least one acceleration setpoint SB of the motor vehicle in a time increment.