Virtual sensor based engine control system and method

• US 20080201054A1
• Filed: 09/29/2006
• Published: 08/21/2008
• Est. Priority Date: 09/29/2006
• Status: Active Grant

• Current Assignee: Caterpillar Incorporated
• Original Assignee: Caterpillar Incorporated
• Inventors: Jayachandran, Amit, Grichnik, Anthony J., Seskin, Michael, Jacobson, Evan E.
• Granted Patent: US 8,478,506 B2
• Time in Patent Office: Days
• US Class Current: 701/102
• CPC Class Codes: B60W 2050/0039 of the propulsion unit
• CPC Class Codes: B60W 2050/0057 Frequency analysis, spectral techniques or transforms
• CPC Class Codes: B60W 40/00 Estimation or calculation of non-directly measurable driving parameters for road vehicle drive control systems not related to the control of a particular sub unit, e.g. by using mathematical models
• CPC Class Codes: B60W 50/00 Details of control systems for road vehicle drive control not related to the control of a particular sub-unit , e.g. process diagnostic or vehicle driver interfaces
• CPC Class Codes: F02D 2041/1433 using a model or simulation of the system
• 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 41/1462 with determination means using an estimation
• CPC Class Codes: F02D 41/1467 with determination means using an estimation
• CPC Class Codes: F02D 41/2425 Particular ways of programming the data

Abstract

A method is provided for a virtual sensor system. The method may include obtaining data records associated with a plurality of input parameters and at least one output parameter and adjusting values of the input parameters based on autocorrelation of respective input parameters. The method may also include reconfiguring the input parameters based on cross-correlation of respective input parameters relative to the output parameter and establishing a first virtual sensor process model indicative of interrelationships between the adjusted and reconfigured input parameters and the output parameter.

Claims

1. A method for a virtual sensor system, comprising:

obtaining data records associated with a plurality of input parameters and at least one output parameter;

adjusting values of the input parameters based on results from analyzing respective input parameters;

reconfiguring the input parameters based on cross-correlation of respective input parameters relative to the output parameter; and

establishing a first virtual sensor process model indicative of interrelationships between the adjusted and reconfigured input parameters and the output parameter.

2. The method according to claim 1, wherein the adjusting the values of the input parameters is based on autocorrelation of the respective input parameters;

3. The method according to claim 2, wherein the adjusting includes:

inputting values of an individual input parameter to an autocorrelation function;

calculating a series of values of the autocorrelation function based on the values of the individual input parameter;

determining whether a calculated autocorrelation value exceeds a predetermined autocorrelation threshold; and

if the calculated autocorrelation value exceeds the threshold, performing an averaging operation on the values of the individual input parameter such that the calculated autocorrelation value is reduced below the threshold.

4. The method according to claim 1, wherein the reconfiguring includes:

inputting values of an individual input parameter and values of the output parameter to a cross-correlation function;

calculating a series of values of the cross-correlation function based on the values of the individual input parameter and the values of the output parameter;

determining whether a calculated cross-correlation value exceeds a predetermined cross-correlation threshold; and

if the calculated cross-correlation value exceeds the threshold, adding an additional input parameter that is the individual input parameter with a delay corresponding to the calculated cross-correlation value.

5. The method according to claim 4, further including:

determining whether a plurality of calculated cross-correlation values exceeds the predetermined cross-correlation threshold; and

if the plurality of calculated cross-correlation values exceed the threshold, adding a plurality of additional input parameters each representing the individual input parameter with a separate delay corresponding to a calculated cross-correlation value.

6. The method according to claim 4, further including:

determining whether a plurality of calculated cross-correlation values exceeds the predetermined cross-correlation threshold;

if the plurality of calculated cross-correlation values exceed the threshold, determining a largest calculated cross-correlation value among the plurality of calculated cross-correlation values; and

adding an additional input parameter that is the individual input parameter with a delay corresponding to the largest calculated cross-correlation value.

7. The method according to claim 1, wherein the at least one output parameter includes at least one regulated emission level of an engine.

8. The method according to claim 7, wherein the at least one output parameter includes one of a NO_x emission level, a soot emission level, and an HC emission level.

9. The method according to claim 1, wherein the plurality input parameters include engine speed, fuel rate, injection timing, intake manifold temperature, intake manifold pressure, inlet valve actuation end of current, and injection pressure.

10. The method according to claim 5, wherein the plurality input parameters include at least one delayed representation of one of the plurality input parameters.

11. The method according to claim 1, wherein the adjusting the values of the input parameters is based on autoregression of the respective input parameters.

12. The method according to claim 1, further including:

providing the output parameter as a feedback to the input parameters for establishing the virtual sensor process model.

13. The method according to claim 12, further including:

establishing a second virtual sensor process model indicative of interrelationships between the adjusted and reconfigured input parameters and another output parameter different from the at least one output parameter; and

coupling the first virtual sensor process model and the second virtual sensor process model simultaneously to provide engine control parameters.

14. The method according to claim 13, wherein the another output parameter includes one of soot oxidation rate, soot passive regeneration rate, exhaust manifold temperature, air system pressure and temperature estimations, gas-to-brick temperature offset estimation, and auxiliary regeneration flame detection temperature.

15. The method according to claim 1, wherein the establishing includes:

creating a neural network computational model indicative of the interrelationships between the input parameters and the output parameter;

training the neural network computational model using the data records;

validating the neural network computation model using the data records;

determining desired statistical distributions of the plurality of input parameters of the computational model; and

recalibrating the plurality of input parameters based on the desired statistical distributions to define a desired input space.

16. The method according to claim 15, wherein determining the desired statistical distributions further includes:

determining a candidate set of the input parameters with a maximum zeta statistic using a genetic algorithm; and

determining the desired distributions of the input parameters based on the candidate set,

wherein the zeta statistic ζ

is represented by;

$\mathrm{\zeta <br><br>}=\underset{1}{\overset{j}{\sum <br><br>}}<br><br>\underset{1}{\overset{i}{\sum <br><br>}}<br><br><br><br>S_{\mathrm{ij}}<br><br><br><br>\left(\frac{{\mathrm{\sigma <br><br>}}_i}{<br><br>{\stackrel{\_}{x}}_i}\right)<br><br>\left(\frac{<br><br>{\stackrel{\_}{x}}_j}{{\mathrm{\sigma <br><br>}}_j}\right),$

provided that x_i represents a mean of an ith input;

x_j represents a mean of a jth output;

σ

_i represents a standard deviation of the ith input;

σ

_j represents a standard deviation of the jth output; and

|S_ij| represents sensitivity of the jth output to the ith input of the computational model.

17. A method for a virtual sensor based engine control system, comprising:

obtaining data records associated with a plurality of engine control parameters and a plurality of sensing parameters including a first sensing parameter and a second sensing parameter;

establishing a first virtual sensor process model indicative of interrelationships between the plurality of engine parameters and the first sensing parameter, wherein the plurality of engine parameters are both input parameters and output parameters of the first virtual sensor process model; and

establishing a second virtual sensor process model indicative of interrelationships between the plurality of engine parameters and the second sensing parameter, wherein the plurality of engine parameters are both input parameters and output parameters of the first virtual sensor process model.

18. The method according to claim 17, further including:

combining the output parameters of the first virtual sensor process model and the output parameter of the second virtual sensor process model to generate a combined output parameters; and

providing control functions to an engine based on the combined output parameters.

19. The method according to claim 17, wherein the engine parameters include engine speed, fuel rate, injection timing, intake manifold temperature, intake manifold pressure, inlet valve actuation end of current, and injection pressure.

20. The method according to claim 17, wherein the plurality of sensing parameters include NO_x emission level, soot emission level, HC emission level, soot oxidation rate, soot passive regeneration rate, exhaust manifold temperature, air system pressure and temperature estimations, gas-to-brick temperature offset estimation, and auxiliary regeneration flame detection temperature.

21. A computer system for establishing a virtual sensor system, comprising:

a database configured to store information relevant to the virtual sensor system; and

a processor configured to;

obtain data records associated with a plurality of input parameters and at least one output parameter;

adjust values of the input parameters based on autocorrelation of respective input parameters;

reconfigure the input parameters based on cross-correlation of respective input parameters relative to the output parameter; and

establish a virtual sensor process model indicative of interrelationships between the adjusted and reconfigured input parameters and the output parameter.

22. The computer system according to claim 21, wherein, to adjust the values of the input parameters, the processor is further configured to:

input values of an individual input parameter to an autocorrelation function;

calculate a series of values of the autocorrelation function based on the values of the individual input parameter;

determine whether a calculated autocorrelation value exceeds a predetermined autocorrelation threshold; and

if the calculated autocorrelation value exceeds the threshold, perform an averaging operation on the values of the individual input parameter such that the calculated autocorrelation value is reduced below the threshold.

23. The computer system according to claim 21, wherein, to reconfigure the input parameters, the processor is further configured to:

input values of an individual input parameter and values of the output parameter to a cross-correlation function;

calculate a series of values of the cross-correlation function based on the values of the individual input parameter and the values of the output parameter;

determine whether a calculated cross-correlation value exceeds a predetermined cross-correlation threshold; and

if the calculated cross-correlation value exceeds the threshold, add an additional input parameter that is the individual input parameter with a delay corresponding to the calculated cross-correlation value.

24. The computer system according to claim 21, wherein the processor is further configured to:

determine whether a plurality of calculated cross-correlation values exceeds the predetermined cross-correlation threshold; and

if the plurality of calculated cross-correlation values exceed the threshold, add a plurality of additional input parameters each representing the individual input parameter with a separate delay corresponding to a calculated cross-correlation value.

25. The computer system according to claim 21, wherein the at least one output parameter includes at least one of regulated emission level of a NO_x emission level, a soot emission level, and an HC emission level of an engine.

26. The computer system according to claim 24, wherein the plurality input parameters include fuel rate, fuel rate with a first delay, and fuel rate with a second delay.

27. The computer system according to claim 24, wherein the plurality input parameters include injection timing, injection timing with a first delay, and injection timing with a second delay.

28. The computer system according to claim 21, wherein, to establish the virtual sensor process model, the processor is further configured to:

create a neural network computational model indicative of the interrelationships between the input parameters and the output parameter;

train the neural network computational model using the data records;

validate the neural network computation model using the data records;

determine desired statistical distributions of the plurality of input parameters of the computational model; and

recalibrate the plurality of input parameters based on the desired statistical distributions to define a desired input space.

29. The computer system according to claim 28, wherein, to determine the desired statistical distributions, the processor is further configured to:

determine a candidate set of the input parameters with a maximum zeta statistic using a genetic algorithm; and

determine the desired distributions of the input parameters based on the candidate set,

wherein the zeta statistic ζ

is represented by;

$\mathrm{\zeta <br><br>}=\underset{1}{\overset{j}{\sum <br><br>}}<br><br>\underset{1}{\overset{i}{\sum <br><br>}}<br><br><br><br>S_{\mathrm{ij}}<br><br><br><br>\left(\frac{{\mathrm{\sigma <br><br>}}_i}{<br><br>{\stackrel{\_}{x}}_i}\right)<br><br>\left(\frac{<br><br>{\stackrel{\_}{x}}_j}{{\mathrm{\sigma <br><br>}}_j}\right),$

provided that x_i represents a mean of an ith input;

x_j represents a mean of a jth output;

σ

_i represents a standard deviation of the ith input;

σ

_j represents a standard deviation of the jth output; and

|S_ij| represents sensitivity of the jth output to the ith input of the computational model.

30. The computer system according to claim 21, wherein the processor is implemented using a field programmable gate array (FPGA) device and may be reconfigured to perform the functions of obtaining, adjusting, reconfiguring, and establishing.