Apparatus and method for diagnosing an engine using computer based models in combination with a neural network

• US 6,240,343 B1
• Filed: 12/28/1998
• Issued: 05/29/2001
• Est. Priority Date: 12/28/1998
• Status: Expired due to Term

• Current Assignee: Caterpillar Incorporated
• Original Assignee: Caterpillar Incorporated
• Inventors: Schricker, David R., Sarangapani, Jagannathan
• Primary Examiner(s): DOMBROSKE, GEORGE M
• Application Number:
  US09/221,924
• Time in Patent Office: 883 Days
• Field of Search: 73/116, 73/117.2, 73/117.3, 73/118.1, 701/29, 701/31, 701/34, 701/101, 701/102, 701/110, 701/111
• US Class Current: 701/29
• CPC Class Codes: F02D 2041/1433 using a model or simulation of the system
• CPC Class Codes: F02D 41/1405 Neural network control
• CPC Class Codes: F02D 41/22 Safety or indicating devices for abnormal conditions in air/fuel ratio feedback systems F02D41/1495, in electric control linkage F02D11/107, in purge control systems F02M25/0809
• CPC Class Codes: G01M 15/05 by combined monitoring of two or more different engine parameters
• CPC Class Codes: Y02T 10/40 Engine management systems

Abstract

An apparatus and method for diagnosing an engine using computer based models in combination with a neural network which reduces the neural network training and updating required is disclosed. The method and apparatus determine modeled values for a plurality of engine parameters as a function of sensed values in a plurality of initial values; sense a plurality of actual values for the engine parameters; and determine a difference between the actual values and the modeled values for the respective engine parameters. The differences are inputted into a neural network which generates an output pattern as a function of the differences and a plurality of weight values. The weight values and the initial values are then updated as a function of a comparison between the output patterns and a desired pattern, and the engine is responsively diagnosed as a function of the output pattern.

Claims

1. A method for diagnosing an engine, including the steps of:

(a) determining modeled values for a plurality of engine parameters as a function of sensed values and a plurality of initial values;

(b) sensing a plurality of actual values for the engine parameters;

(c) determining a difference between the actual values and the modeled values for the respective engine parameters;

(d) inputting the differences determined in step (c) into a neural network to generate an output pattern as a function of the differences and a plurality of weight values;

(e) updating the weight values and the initial values as a function of a comparison between the output patterns generated in step (d) and a desired pattern; and

(f) responsively diagnosing the engine as a function of the output pattern.

2. The method, as set forth in claim 1, where in step (a) the initial values are determined for the engine when running under predetermined normal conditions.

3. The method, as set forth in claim 1, wherein the engine parameters comprise exhaust temperature, oil pressure and power output.

4. The method, as set forth in claim 3, wherein the engine parameters further comprises boost pressure for a turbocharger of the engine.

5. The method, as set forth in claim 4, wherein the power output is a function of power at a brake and power at a torque converter.

6. The method, as set forth in claim 3, wherein the engine has at least two cylinders, and the engine parameters further comprise a difference between exhaust temperatures for the cylinders.

7. A method for diagnosing an engine, comprising the steps of:

(a) determining a plurality of initial values for a selected plurality of engine parameters under a predetermined normal engine operating condition;

(b) sensing the plurality of engine parameters and determining modeled engine parameter values as a function of the sensed engine parameters and the initial values;

(c) determining actual engine parameters;

(d) comparing the actual engine parameters to the modeled engine parameters and responsively determining corresponding difference values;

(e) providing an input pattern to a neural network to generate an output pattern, the input pattern being a function of the difference values;

(f) updating the neural network and the initial values as a function of a comparison of the output pattern with a desired pattern; and

(g) diagnosing an engine problem as a function of the output pattern.

8. The method, as set forth in claim 7, wherein the modeled engine parameter values comprise a modeled value for exhaust temperature determined from a sensed engine speed, a sensed fuel rate, and an air fuel ratio for the engine.

9. The method, as set forth in claim 7, wherein the modeled engine parameter values comprise an exhaust temperature modeled value;

an oil pressure modeled value and a power modeled value.

10. Apparatus for diagnosing an engine, comprising:

(a) means for determining a plurality of initial values for a selected plurality of engine parameters under a predetermined normal engine operating condition;

(b) means for sensing the plurality of engine parameters;

(c) means for determining modeled engine parameter values as a function of the sensed engine parameters and the initial values;

(d) means for determining actual engine parameters;

(e) means for comparing the actual engine parameters to the modeled engine parameters and responsively determining corresponding difference values;

(f) means for providing an input pattern to a neural network to generate an output pattern, the input pattern being a function of the difference values;

(g) means for updating the neural network and the initial values as a function of a comparison of the output pattern with a desired pattern; and

(h) means for diagnosing an engine problem as a function of the output pattern.

11. Apparatus, as set forth in claim 10, wherein the means for determining actual engine parameters comprise a diagnostic module operatively connected to an engine speed sensor, a fuel rate sensor, at least one exhaust temperature sensor, an oil temperature sensor, and an oil pressure sensor.

12. Apparatus, as set forth in claim 11, wherein the means for determining actual engine parameters further comprise at least one brake temperature sensor, a boost pressure sensor and a torque converter sensor.

13. Apparatus, as set forth in claim 10, wherein the means for determining the modeled engine parameter values, the means for comparing the actual engine parameters to the modeled engine parameters and responsively determining corresponding difference values comprise a diagnostic module.