WAVELET-BASED ARTIFICIAL NEURAL NET COMBUSITION SENSING

• US 20040084024A1
• Filed: 10/31/2002
• Published: 05/06/2004
• Est. Priority Date: 10/31/2002
• Status: Active Grant

• Current Assignee: Delphi Technologies, Inc. (BorgWarner Incorporated)
• Original Assignee: Delphi Technologies, Inc. (BorgWarner Incorporated)
• Inventors: Malaczynski, Gerard Wladyslaw, Baker, Michael Edward
• Granted Patent: US 6,805,099 B2
• Time in Patent Office: Days
• US Class Current: 123/435
• CPC Class Codes: F02B 1/12 with compression ignition with fuel-air charge ignited by compression ignition of an additional fuel F02B7/00
• CPC Class Codes: F02D 35/021 using an ionic current sensor
• CPC Class Codes: F02D 41/1405 Neural network control
• CPC Class Codes: G01L 23/22 for detecting or indicating knocks in internal-combustion engines; Units comprising pressure-sensitive members combined with ignitors for firing internal-combustion engines

Abstract

A method and apparatus for real-time measurement of combustion characteristics of each combustion event in each individual cylinder coupled with an ability to control the engine based upon the combustion characteristics are shown. The invention includes using selective sampling techniques and wavelet transforms to extract a critical signal feature from an ionization signal that is generated by an in-cylinder ion sensor, and then feeds that critical signal feature into an artificial neural network to determine a desired combustion characteristic of the combustion event. The desired combustion characteristic of the combustion event includes a location of peak pressure, an air/fuel ratio, or a percentage of mass-fraction burned, among others. The control system of the engine is then operable to control the engine based upon the combustion characteristic.

Claims

1. A method for controlling an internal combustion engine, comprising

determining a combustion characteristic of an individual combustion event of the engine by:

concurrently sampling an ionization signal and determining at least one engine operating parameter during a combustion event of the internal combustion engine,

transforming the ionization signal with a wavelet filter,

extracting at least one critical signal feature from the ionization signal that has been transformed by the wavelet filter; and

processing the at least one critical signal feature and the at least one engine operating parameter through an artificial neural network; and

controlling the internal combustion engine based upon the combustion characteristic.

2. The method of claim 1, wherein determining a combustion characteristic of an individual combustion event of the engine occurs each combustion event.

3. The method of claim 1, wherein transforming the ionization signal with a wavelet filter comprises:

tuning the wavelet filter to the at least one critical signal feature;

decomposing the ionization signal using the tuned wavelet filter; and

reconstructing the decomposed ionization signal.

4. The method of claim 3, wherein the at least one critical signal feature comprises a location of a second peak ionization value after an ignition event relative to the rotation of the crankshaft of the engine.

5. The method of claim 4, wherein extracting at least one critical signal feature from the ionization signal that has been transformed by the wavelet filter comprises determining a specific location of the second peak ionization value after the ignition event relative to the rotation of the crankshaft of the engine.

6. The method of claim 5, wherein determining a combustion characteristic of the engine comprises using the artificial neural network to determine a magnitude of engine torque based upon the specific location of the second peak ionization value after the ignition event relative to the rotation of the crankshaft of the engine.

7. The method of claim 6, wherein controlling the internal combustion engine based upon the combustion characteristic comprises controlling at least one output device of the engine based upon the magnitude of engine torque.

8. The method of claim 1, wherein concurrently sampling an ionization signal and at least one engine parameter during a combustion event of the internal combustion engine comprises sampling during a critical period of time after an ignition event.

9. The method of claim 1, wherein concurrently sampling an ionization signal and at least one engine parameter during a combustion event comprises using a sampling and processing scheduler to activate an analog data sampler during a time when the critical signal feature is present in the ionization signal.

10. The method of claim 3, wherein the at least one critical signal feature comprises a curve containing a first peak ionization value after an ignition event.

11. The method of claim 10, wherein extracting at least one critical signal feature from the ionization signal that has been transformed by the wavelet filter comprises determining a time-integral of the curve containing the first peak ionization value of after the ignition event.

12. The method of claim 11, wherein determining a combustion characteristic of the engine comprises using the artificial neural network to determine an air/fuel ratio based upon the magnitude of the time-integral of the curve containing the first peak ionization value after the ignition event.

13. The method of claim 12, wherein controlling the internal combustion engine based upon the combustion characteristic comprises controlling at least one output device of the engine based upon the air/fuel ratio.

14. The method of claim 1, wherein determining at least one engine operating parameter during a combustion event of the internal combustion engine comprises determining an engine load, a spark advance, an amount of delivered fuel and a mass fraction of EGR delivered.

15. The method of claim 1, wherein processing the at least one critical signal feature and the at least one engine operating parameter through an artificial neural network comprises

gathering data to train the artificial neural network;

training the artificial neural network using the gathered data;

implementing the trained artificial neural network as at least one algorithm in the controller.

16. The method of claim 1, wherein processing the at least one critical signal feature and the at least one engine operating parameter through an artificial neural network comprises

gathering data from the ionization signal to train the artificial neural network;

compressing the gathered data for storage;

decompressing the gathered data that has been stored;

training the artificial neural network using the decompressed data;

implementing the trained artificial neural network as at least one algorithm in the controller.

17. A method for determining a combustion characteristic of an internal combustion engine, comprising

concurrently sampling an ionization signal and determining at least one engine operating parameter during a combustion event;

transforming the ionization signal with a wavelet filter;

extracting at least one critical signal feature from the ionization signal that has been transformed by the wavelet filter; and

processing the at least one critical signal feature and the at least one engine operating parameter through an artificial neural network.

18. A system to control an internal combustion engine, comprising a controller operably connected to:

at least one output device of the internal combustion engine,

at least one ion sensor operable to measure an ionization signal from a combustion event of the internal combustion engine, and

at least one engine sensor operable to measure at least one engine operating parameter;

wherein said controller is operable to;

concurrently sample an ionization signal using the at least one ion sensor and determine at least one engine operating parameter using the at least one engine sensor during a combustion event,

transform the ionization signal with a wavelet filter,

extract at least one critical signal feature from the ionization signal that has been transformed by the wavelet filter, and

process the at least one critical signal feature and the at least one engine operating parameter through an artificial neural network to determine at least one combustion characteristic of the engine; and

wherein the controller controls the at least one output device of the internal combustion engine based upon the at least one combustion characteristic of the engine.

19. The system of claim 18, wherein having the controller control the at least one output device of the internal combustion engine based upon the at least one combustion characteristic comprises controlling engine torque.

20. The system of claim 18, wherein the internal combustion engine comprises a compression-ignition engine.

21. The system of claim 18, wherein the internal combustion engine comprises a spark-ignition engine.

22. The system of claim 18, wherein the internal combustion engine comprises a homogeneous charge compression-ignition engine.

23. A system to determine a combustion characteristic of an internal combustion engine, comprising:

a controller operably connected to;

at least one output device of the internal combustion engine,

at least one ion sensor operable to measure an ionization signal from a combustion event of the internal combustion engine, and

at least one engine sensor operable to measure at least one engine operating parameter;

wherein said controller is operable to;

concurrently sample an ionization signal using the at least one ion sensor and determine at least one engine operating parameter using the at least one engine sensor during a combustion event,

transform the ionization signal with a wavelet filter,

extract at least one critical signal feature from the ionization signal that has been transformed by the wavelet filter, and

process the at least one critical signal feature and the at least one engine operating parameter through an artificial neural network.