Method for identifying misfire events of an internal combustion engine

• US 5,732,382 A
• Filed: 11/06/1996
• Issued: 03/24/1998
• Est. Priority Date: 11/06/1996
• Status: Expired due to Fees

• Current Assignee: Ford Global Technologies, LLC (Ford Motor Company)
• Original Assignee: Ford Global Technologies, LLC (Ford Motor Company)
• Inventors: Feldkamp, Timothy Mark, Puskorius, Gintaras Vincent, Feldkamp, Lee Albert, James, John Victor, Marko, Kenneth Andrew
• Primary Examiner(s): Dombroske, George M.
• Application Number:
  US08/744,258
• Time in Patent Office: 503 Days
• Field of Search: 73/116, 73/117.2, 73/117.3, 123/419, 123/436, 364/431.03, 364/431.07, 364/431.08, 701/101, 701/110, 701/111
• US Class Current: 701/110
• CPC Class Codes: F02D 41/1405 Neural network control
• CPC Class Codes: G01M 15/11 by detecting misfire

Abstract

A method for identifying engine combustion failure of an internal combustion engine having a plurality of cylinders, a crankshaft and a crankshaft position sensor includes the steps of operating the internal combustion engine to rotate the crankshaft, measuring rotational quantities of the crankshaft corresponding to events created by each of the plurality of cylinders during operation of the internal combustion engine, correcting the rotational quantities measured to remove periodic position irregularities to generate a corrected temporal signal, generating an acceleration signal of the crankshaft using the corrected temporal signals, and identifying combustion failures as a function of the acceleration signal. A time-lagged recurrent neural network utilizes the acceleration signal, along with other engine parameters to identify the cylinder-specific misfire events.

Claims

1. A method for identifying misfire events in an internal combustion engine having a plurality of cylinders, a crankshaft and a crankshaft position sensor, the method comprising the steps of:

operating the internal combustion engine to rotate the crankshaft;

measuring rotational quantities of the crankshaft corresponding to events created by each of the plurality of cylinders during operation of the internal combustion engine;

correcting the rotational quantities measured to remove periodic position irregularities to generate a corrected temporal signal;

generating an acceleration signal of the crankshaft using the corrected temporal signal;

applying a recurrent neural network to the acceleration signal; and

identifying misfire events as a function of the acceleration signal.

2. A method as set forth in claim 1 wherein the recurrent neural network is a time-lagged recurrent neural network.

3. A method as set forth in claim 2 including the step of training the time-lagged recurrent neural network with training data points of engine combustions.

4. A method as set forth in claim 3 wherein the training data points of engine combustions includes random patterns of misfire events.

5. A method as set forth in claim 4 including the step of using rotating speed of the crankshaft for an input to the time-lagged recurrent neural network.

6. A method as set forth in claim 4 including the step of using engine load for an input to the time-lagged recurrent neural network.

7. A method as set forth in claim 2 wherein the step of identifying misfire events includes the step of comparing the acceleration signal to a threshold signal.

8. A method as set forth in claim 7 including the step of generating the threshold signal based on the acceleration signal.

9. A method as set forth in claim 6 including the step of training the recurrent neural network with training data points of engine combustions.

10. A method as set forth in claim 9 wherein the training date points of engine combustions includes random patterns of misfire events.

11. A method as set forth in claim 10 including the step of using rotating speed of the crankshaft for an input to the recurrent neural network.

12. A method as set forth in claim 10 including the step of using engine load for an input to the recurrent neural network.

13. A method for identifying misfire events of an internal combustion engine having a plurality of cylinders, a crankshaft and a crankshaft position sensor, the method comprising the steps of:

operating the internal combustion engine to combust an air/fuel mixture in each of the plurality of cylinders;

receiving a timing signal identifying when each of the plurality of cylinders combusts the air/fuel mixture;

determining an acceleration signal of the crankshaft using the timing signal;

applying a recurrent neural network to the acceleration signal; and

identifying misfire events as a function of the acceleration signal.