Control device and method for calculating an output parameter for a controller

Patent metadata

Field	Value
Publication number	US-10013658-B2
Application number	US-201113643464-A
Country	US
Kind code	B2
Filing date	Apr 6, 2011
Priority date	Apr 27, 2010
Publication date	Jul 3, 2018
Grant date	Jul 3, 2018

Abstract

A control device in a vehicle includes a unit for calculating, during operation of the vehicle, on the basis of at least one input variable ascertained during operation, at least one output variable for a control system of functions of the vehicle. The control device performs the calculation of the output variables using a Bayesian regression of training values ascertained, before operation, for the output variable and the input variable.

First claim

What is claimed is:

1. A control device in a vehicle, comprising:
a calculation unit configured to calculate, during operation of the vehicle and on the basis of at least one input variable ascertained during operation of the vehicle, at least one output variable for a control system of a function of the vehicle, wherein the calculation unit is configured to perform the calculation of the at least one output variable using a Bayesian regression over training values ascertained, before operation, for the at least one output variable and the at least one input variable;
a computation unit configured to carry out the Bayesian regression at least partly; and
a logic circuit that is associated with the computation unit of the control device, wherein the logic circuit is optimized for the calculation of specific computation steps of the Bayesian regression;
wherein the at least one input variable ascertained during the operation of the vehicle represents at least one of temperature signals, rotation speed signals, and quantity signals, the at least one of temperature signals, rotation speed signals, and quantity signals being received by the control device from at least one of: (i) at least one sensor of the vehicle, (ii) at least one other control device of the vehicle, and (iii) at least one module of the vehicle outside the control device;
wherein the at least one output variable includes at least one combustion engine-relevant parameter; and
wherein the control device controls a combustion engine of the vehicle using the at least one combustion engine-relevant parameter during operation of the vehicle.

2. The control device as recited in claim 1 , wherein the calculation unit implements the Bayesian regression as a kriging or a sparse Gaussian process.

3. The control device as recited in claim 2 , wherein the calculation unit determines a variance, ascertained from the Bayesian regression, of the output variable.

4. The control device as recited in claim 3 , further comprising:
means for initiating a corrective action if the ascertained variance exceeds a predetermined threshold value.

5. The control device as recited in claim 3 , wherein for the calculation of the at least one output variable, a Bayesian regression dynamically modeled by autoregression is carried out over the training values ascertained before operation.

6. The control device as recited in claim 1 , wherein the logic circuit associated with the computation unit is configured to calculate exponential functions.

7. The control device as recited in claim 1 , wherein the calculation unit implements the Bayesian regression as a Gaussian process.

8. The control device as recited in claim 7 , wherein the calculation unit determines a variance, ascertained from the Bayesian regression, of the output variable.

9. The control device as recited in claim 8 , further comprising:
means for initiating a corrective action if the ascertained variance exceeds a predetermined threshold value.

10. The control device as recited in claim 8 , wherein the logic circuit associated with the computation element is configured to calculate exponential functions.

11. The control device as recited in claim 1 , wherein the calculation unit implements the Bayesian regression as a kriging process.

12. The control device as recited in claim 1 , wherein the calculation unit implements the Bayesian regression as a sparse Gaussian process.

13. The control device as recited in claim 1 , wherein the calculation unit is further configured to:
determine a variance, ascertained from the Bayesian regression, for the at least one output variable;
wherein the control device is configured to control the vehicle according to a first control scheme, and when the ascertained variance exceeds a predetermined threshold value that indicates that a model used for the Bayesian regression is not reliable, switch from the first control scheme to a second fail-safe control scheme and control the vehicle according to the second fail-safe control scheme which prevents damage to the vehicle.

14. A method for calculating by a control device in a vehicle, during operation of the vehicle and on the basis of at least one input variable ascertained during operation, at least one output variable for a control system of a function of the vehicle, comprising:
ascertaining training values for the at least one output variable and the at least one input variable before operation of the vehicle;
during operation of the vehicle, receiving, by the control device, the at least one input variable ascertained during the operation of the vehicle including at least one of (i) temperature signals, (ii) rotation speed signals, and (iii) quantity signals, from at least one of: (i) at least one sensor, (ii) at least one other control device, and (iii) at least one module outside the control device, the at least one of the temperature signals, rotational signals, and quantity signals being ascertained during the operation of the vehicle;
during operation of the vehicle, performing the calculation of the at least one output variable using a Bayesian regression over the training values;
during operation of the vehicle, carrying out the Bayesian regression at least partly with a computation unit of the control device; and
optimizing a logic circuit that is associated with the computation unit of the control device for the calculation of specific computation steps of the Bayesian regression;
wherein the at least one output variable includes at least one combustion engine-relevant parameter; and
wherein the method further comprises controlling, by the control device, a combustion engine of the vehicle using the at least one combustion engine-relevant parameter during the operation of the vehicle.

15. The method as recited in claim 14 , wherein a variance, ascertained from the Bayesian regression, of the output variable is also taken into consideration for the control system.

16. The method as recited in claim 15 , wherein the Bayesian regression is implemented as a kriging or a sparse Gauss process and is dynamically modeled by autoregression carried out over the training values ascertained before operation.

17. The method as recited in claim 15 , wherein the Bayesian regression is implemented as a kriging process.

18. The method as recited in claim 15 , wherein the Bayesian regression is implemented as a sparse Gauss process.

19. The method as recited in claim 15 , wherein the Bayesian regression is implemented as a Gauss process.

20. The method as recited in claim 14 , further comprising:
determining a variance, ascertained from the Bayesian regression, for the at least one output variable;
controlling the vehicle according to a first control scheme; and
when the ascertained variance exceeds a predetermined threshold value that indicates that a model used for the Bayesian regression is not reliable, switching from the first control scheme to a second fail-safe control scheme and controlling the vehicle according to the second fail-safe control scheme which prevents damage to the vehicle.

Assignees

• Streichert Felix
• Lang Tobias
• Markert Heiner
• Aue Axel
• Kruse Thomas
• Imhof Volker
• Richardsen Thomas
• Schulmeister Ulrich
• Bannow Nico
• Diener Rene
• Kloppenburg Ernst
• Saetzler Michael
• Ulmer Holger
• Bosch Gmbh Robert

Inventors

• Streichert Felix
• Lang Tobias
• Markert Heiner
• Aue Axel
• Kruse Thomas
• Imhof Volker
• Richardsen Thomas
• Schulmeister Ulrich
• Bannow Nico
• Diener Rene
• Kloppenburg Ernst
• Saetzler Michael
• Ulmer Holger

Classifications

• G06N7/01

  Probabilistic graphical models, e.g. probabilistic networks · CPC title

• F02D2041/1433

  using a model or simulation of the system · CPC title

• G05B17/02

  electric · CPC title

• F02D41/1401

  characterised by the control or regulation method (F02D41/1473, F02D41/1477 take precedence) · CPC title

• G06N7/005

  Physics · mapped topic