PLANT CONTROL APPARATUS

• US 20110172897A1
• Filed: 01/04/2011
• Published: 07/14/2011
• Est. Priority Date: 01/14/2010
• Status: Active Grant

• Current Assignee: Honda Motor Co., Ltd. (Honda Motor Company)
• Original Assignee: Honda Motor Co., Ltd. (Honda Motor Company)
• Inventors: TSUZUKI, Koichiro, Yasui, Yuji
• Granted Patent: US 8,725,393 B2
• Time in Patent Office: Days
• US Class Current: 701/103
• CPC Class Codes: F02D 2041/1412 using a predictive controller
• CPC Class Codes: F02D 2041/1433 using a model or simulation of the system
• CPC Class Codes: F02D 35/023 by determining the cylinder pressure
• CPC Class Codes: F02D 35/024 using an estimation
• CPC Class Codes: F02D 41/1401 characterised by the control or regulation method F02D41/1473, F02D41/1477 take precedence
• CPC Class Codes: F02D 41/1405 Neural network control
• CPC Class Codes: F02D 41/1406 with use of a optimisation method, e.g. iteration
• CPC Class Codes: F02D 41/222 relating to the failure of sensors or parameter detection devices
• CPC Class Codes: F02D 41/40 with means for controlling injection timing or duration
• CPC Class Codes: F02D 41/401 Controlling injection timing F02D41/402 takes precedence
• CPC Class Codes: F02D 41/402 Multiple injections
• CPC Class Codes: Y02T 10/40 Engine management systems

Abstract

A plant control apparatus includes a plant model and a control input determining device. The plant model estimates a control variable of a plant based on an input including a control parameter to adjust a first control variable of the plant. The control input determining device is configured to correct a temporary value of the control parameter so that a first control variable model value satisfies a predetermined constraint. The control input determining device is configured to determine a control input for the plant based on the temporary value of the control parameter that has been corrected. The first control variable model value is defined as an estimated value of the first control variable calculated by the plant model when the temporary value of the control parameter is input to the plant model.

Claims

1. A plant control apparatus comprising:

a plant model to estimate a control variable of a plant based on an input including a control parameter to adjust a first control variable of the plant; and

a control input determining device configured to correct a temporary value of the control parameter so that a first control variable model value satisfies a predetermined constraint, and configured to determine a control input for the plant based on the temporary value of the control parameter that has been corrected, the first control variable model value being defined as an estimated value of the first control variable calculated by the plant model when the temporary value of the control parameter is input to the plant model.

2. The plant control apparatus according to claim 1,

wherein a second control variable model value is defined as an estimated value of a second control variable that is calculated by the plant model when a temporary value of the control parameter is input to the plant model, and

wherein the control input determining device includes

a plurality of optimizers configured to correct the temporary value of the control parameter under different optimization conditions so that the first control variable model value satisfies the predetermined constraint, and

an optimum value selector configured to select a temporary value of the control parameter, the temporary value making the second control variable model value best fit a predetermined priority condition, from corrected temporary values of the control parameter that have been corrected by the plurality of optimizers, and configured to determine a selected temporary value of the control parameter as the control input for the plant.

3. The plant control apparatus according to claim 1,

wherein the plant model includes a neural network including a plurality of neurons connected to each other, each of the neurons outputting a value in accordance with a predetermined function.

4. The plant control apparatus according to claim 1, further comprising:

a first control variable detector configured to detect the first control variable; and

a model corrector configured to adaptively correct the plant model so as to reduce a deviation of a detection value of the first control variable from the estimated value of the first control variable, the estimated value being calculated by the plant model.

5. The plant control apparatus according to claim 3, further comprising:

a first control variable detector configured to detect the first control variable; and

a model corrector configured to calculate a correction coefficient so as to reduce a deviation of a detection value of the first control variable from the estimated value of the first control variable, the estimated value being calculated by the plant model,

wherein the correction coefficient is included in an input for the neural network, the correction coefficient being used for adaptively correcting a correlation between the input and an output.

6. The plant control apparatus according to claim 3, further comprising:

a first control variable detector configured to detect the first control variable; and

a model corrector configured to calculate a correction coefficient so as to reduce the deviation of a detection value of the first control variable from the estimated value of the first control variable, the estimated value being calculated by the plant model,

wherein the plant model is configured to calculate the first control variable based on a product or a sum of an output of the neural network and the correction coefficient.

7. The plant control apparatus according to claim 5,

wherein a reference parameter including at least one of inputs for the plant model is defined, and

wherein the model corrector includes

a weight function setting device to define a plurality of regions that overlap each other in a space having the reference parameter as a basis, and to set a plurality of weight functions in respective regions, the weight functions being normalized and having nonzero values in the regions,

a correction value calculator configured to calculate a correction value for each of the regions so as to minimize the product of a value of the weight function and the deviation, and

a determination device configured to determine the correction coefficient based on a sum of products of values of the weight functions and the correlation coefficient over all of the regions.

8. The plant control apparatus according to claim 1,

wherein the plant comprises an internal combustion engine,

wherein the first control variable comprises a maximum cylinder pressure for each combustion cycle of the internal combustion engine,

wherein the control parameter includes a fuel injection quantity for each combustion cycle of the internal combustion engine and a fuel injection timing, and

wherein the predetermined constraint is that the first control variable model value is equal to or smaller than a predetermined upper limit.

9. The plant control apparatus according to claim 1,

wherein the plant comprises an internal combustion engine,

wherein the first control variable comprises a maximum cylinder pressure for each combustion cycle of the internal combustion engine,

wherein the control parameter includes a fuel injection quantity for each combustion cycle of the internal combustion engine and a split injection ratio for main fuel injection, and

wherein the predetermined constraint is that the first control variable model value is equal to or smaller than a predetermined upper limit.

10. The plant control apparatus according to claim 1,

wherein the plant comprises an internal combustion engine,

wherein the first control variable comprises an amount of NOx exhausted from the internal combustion engine,

wherein the control parameter includes a fuel injection timing for each combustion cycle of the internal combustion engine and a split injection ratio for main fuel injection, and

wherein the predetermined constraint is that the first control variable model value is equal to or smaller than an upper limit set in accordance with an operation state of the internal combustion engine.

11. The plant control apparatus according to claim 1,

wherein the plant comprises an internal combustion engine including a catalytic converter disposed in an exhaust passage thereof, the catalytic converter purifying exhaust gas,

wherein the first control variable comprises a temperature of the internal combustion engine,

wherein the control parameter includes a fuel injection timing for each combustion cycle of the internal combustion engine and a split injection ratio for a main fuel injection, and

wherein the predetermined constraint is that the first control variable model value is close to a target temperature set in accordance with an activation temperature of the exhaust purifying catalyst.

12. The plant control apparatus according to claim 1,

wherein the plant comprises an internal combustion engine including a variable compression mechanism capable of changing a compression ratio in a combustion chamber,

wherein the first control variable comprises a maximum cylinder pressure for each combustion cycle of the internal combustion engine,

wherein the control parameter includes an ignition timing for each combustion cycle of the internal combustion engine and a compression ratio command to the variable compression mechanism, and

wherein the predetermined constraint is that the first control variable model value is equal to or smaller than a predetermined upper limit.

13. The plant control apparatus according to claim 1,

wherein the plant comprises an internal combustion engine including a variable compression mechanism capable of changing a compression ratio in a combustion chamber,

wherein the first control variable is a knock intensity of the internal combustion engine,

wherein the control parameter includes an ignition timing of the internal combustion engine and a compression ratio command to the variable compression mechanism, and

wherein the predetermined constraint is that the first control variable model value is equal to or smaller than a predetermined upper limit.

14. The plant control apparatus according to claim 2,

wherein the plant model includes a neural network including a plurality of neurons connected to each other, each of the neurons outputting a value in accordance with a predetermined function.

15. The plant control apparatus according to claim 2, further comprising:

a first control variable detector configured to detect the first control variable; and

a model corrector configured to adaptively correct the plant model so as to reduce a deviation of a detection value of the first control variable from the estimated value of the first control variable, the estimated value being calculated by the plant model.

16. The plant control apparatus according to claim 3, further comprising:

a first control variable detector configured to detect the first control variable; and

a model corrector configured to adaptively correct the plant model so as to reduce a deviation of a detection value of the first control variable from the estimated value of the first control variable, the estimated value being calculated by the plant model.

17. The plant control apparatus according to claim 14, further comprising:

a first control variable detector configured to detect the first control variable; and

a model corrector configured to adaptively correct the plant model so as to reduce a deviation of a detection value of the first control variable from the estimated value of the first control variable, the estimated value being calculated by the plant model.

18. The plant control apparatus according to claim 14, further comprising:

a first control variable detector configured to detect the first control variable; and

a model corrector configured to calculate a correction coefficient so as to reduce a deviation of a detection value of the first control variable from the estimated value of the first control variable, the estimated value being calculated by the plant model,

wherein the correction coefficient is included in an input for the neural network, the correction coefficient being used for adaptively correcting a correlation between the input and an output.

19. The plant control apparatus according to claim 14, further comprising:

a first control variable detector configured to detect the first control variable; and

a model corrector configured to calculate a correction coefficient so as to reduce the deviation of a detection value of the first control variable from the estimated value of the first control variable, the estimated value being calculated by the plant model,

wherein the plant model is configured to calculate the first control variable based on a product or a sum of an output of the neural network and the correction coefficient.

20. A plant control apparatus comprising:

plant model means for estimating a control variable of a plant based on an input including a control parameter to adjust a first control variable of the plant; and

control input determining means for correcting a temporary value of the control parameter so that a first control variable model value satisfies a predetermined constraint, and configured to determine a control input for the plant based on the temporary value of the control parameter that has been corrected, the first control variable model value being defined as an estimated value of the first control variable calculated by the plant model when the temporary value of the control parameter is input to the plant model.