Misfire detector with torsional oscillation filtering

• US 5,752,213 A
• Filed: 02/29/1996
• Issued: 05/12/1998
• Est. Priority Date: 02/29/1996
• Status: Expired due to Fees

• Current Assignee: Ford Global Technologies, LLC (Ford Motor Company)
• Original Assignee: Ford Global Technologies, LLC (Ford Motor Company)
• Inventors: Krishnaswami, Vasanth, Marko, Kenneth Andrew, Bryant, Bruce David
• Primary Examiner(s): LOUIS JACQUES, JACQUES H
• Application Number:
  US08/609,028
• Time in Patent Office: 803 Days
• Field of Search: 364/431.08, 364/431.07, 364/431.01, 364/431.03, 364/431.04, 73/117.3, 73/116, 73/118.1, 73/112, 123/419, 123/436, 123/425, 123/435, 123/494, 123/617
• US Class Current: 701/111
• CPC Class Codes: F02B 77/08 Safety, indicating, or supervising devices thermal insulation F02B77/11; monitoring or diagnostic devices for exhaust-gas treatment apparatus F01N11/00
• CPC Class Codes: G01M 15/11 by detecting misfire

Abstract

Irregularities in crankshaft velocity introduced when measuring crankshaft rotation at a section of a crankshaft in an internal combustion engine that is less damped to torsional oscillations than is another more accessible crankshaft section are corrected using a NARMAX-generated model to predict rotation measurements that would have been obtained at the inaccessible section based on data actually collected at the accessible crankshaft section. Thus, the effects of torsional oscillations in the crankshaft are substantially filtered away, resulting in crankshaft acceleration values that form the basis of a misfire detector having nearly maximum signal-to-noise performance.

Claims

1. A misfire detection system for, an internal combustion engine having a plurality of cylinders coupled to a crankshaft, said crankshaft having a flywheel portion to which a flywheel is mounted, each cylinder producing respective ignition firings and misfires during operation of said engine, said engine including a plurality of devices for measuring engine operating conditions at respective event times, said engine operating conditions including one or more of crankshaft rotation interval, crankshaft speed, crankshaft acceleration, mass air flow, and fuel pulse width, said misfire detection system comprising:

modeling means coupled to said plurality of devices and including a first model interrelating said engine operating conditions and crankshaft acceleration for predicting a crankshaft acceleration value of said crankshaft near said flywheel portion without directly sensing motion of said flywheel portion, and including a second model interrelating said engine operating conditions and an acceleration that would be present near said flywheel portion assuming

     1) that respective engine operating conditions are present as determined at a selected event time, and

     2) that an individual misfire has occurred in a respective cylinder of said engine at said selected event time for predicting an acceleration deficit; and

a comparator outputting a misfire indication in response to a comparison of said predicted crankshaft acceleration value and said predicted acceleration deficit.

2. The vehicle of claim 1 further comprising a threshold adjuster coupled to said comparator for adjusting sensitivity of said misfire indication.

3. The vehicle of claim 1 wherein said modeling means is derived as a non-linear autoregressive moving average with exogenous inputs model based upon training data obtained from a test engine.

4. A method of detecting individual misfires in an internal combustion engine having a plurality of cylinders coupled to a crankshaft, said crankshaft having a flywheel portion to which a flywheel is mounted, each cylinder producing respective ignition firings and misfires during operation of said engine, said engine including a plurality of devices for measuring engine operating conditions at respective event times, said engine operating conditions including one or more of crankshaft rotation interval, crankshaft speed, crankshaft acceleration, mass air flow, and fuel pulse width, said method comprising the steps of:

determining said engine operating conditions at a selected event time;

predicting a crankshaft acceleration value of said crankshaft near said flywheel portion without directly sensing motion of said flywheel portion by inputting into a first model respective engine operating conditions including at least one of said crankshaft rotation interval, crankshaft speed or crankshaft acceleration;

predicting an acceleration deficit by inputting respective engine operating conditions into a second model having a configuration corresponding to said engine and interrelating said engine operating conditions with and acceleration that would be present near said flywheel portion assuming

     1) that respective engine operating conditions are present as determined at said selected event time, and

     2) that an individual misfire has occurred in a respective cylinder of said engine at said selected event time; and

comparing said predicted crankshaft acceleration value and said predicted acceleration deficit to produce a misfire indication.

5. The method of claim 4 further comprising the step of scaling either of said predicted crankshaft acceleration value or said predicted acceleration deficit by a predetermined scale factor.

6. The method of claim 4 wherein said predicting steps are based on at least one non-linear autoregressive moving average with exogenous inputs model of said engine, said non-linear autoregressive moving average with exogenous inputs model being derived from training data obtained from a test engine.