Data processing: artificial intelligence – Neural network – Structure
Reexamination Certificate
1999-04-21
2002-08-13
Davis, George B. (Department: 2121)
Data processing: artificial intelligence
Neural network
Structure
C706S043000
Reexamination Certificate
active
06434541
ABSTRACT:
TECHNICAL FIELD
This invention relates generally to detection of misfires in automotive internal combustion engines and, more particularly, to a bit-serial based recurrent neuroprocessor for processing data from an internal combustion engine in order to diagnose misfires in automotive engines in real-time.
BACKGROUND ART
Misfire diagnostics in internal combustion engines requires the detection and identification of improper combustion events in each firing cylinder. In order to utilize information from sensors now in production use, the diagnostic is based upon analysis of crankshaft dynamics. Misfire detection relies on computing the derivatives of crankshaft position sensor signals to determine short term average crankshaft accelerations several times each revolution. These accelerations are computed over rotational segments spanning the intervals between engine cylinder firings.
While analysis of crankshaft dynamics permits detection of engine misfire in many circumstances, the signal signatures of these events are obscured by complicated dynamics arising from torsional oscillations occurring in the crankshaft itself as a result of the excitation of its intrinsic normal modes. Since detection of misfires is based upon the principle that engine misfires cause a torque impulse to be absent, with a consequent torque or acceleration deficit, the detection of the deficit forms the basis of the misfire diagnostic. The diagnostic algorithms must detect the acceleration deficit in such a manner as to reliably detect the engine misfires, while recognizing the more frequently occurring normal combustion events. Analysis of the dynamics must result in failure detecting capabilities in excess of 95% of all failure events, with simultaneous identification of normal events as normal, with accuracies approaching 99.9%.
SUMMARY OF THE INVENTION
In accordance with the present invention, an engine diagnostic system is provided that includes a neuroprocessor capable of efficiently performing the required computations for detecting engine misfire and/or performing other diagnostic functions. The architecture and hardware is sufficiently flexible to be able to perform the misfire diagnostic task and still have the capability of performing other diagnostics or control functions in automotive systems such as idle speed control and air/fuel ratio control. This flexibility is achieved through the use of a high speed hardware realization of basic neural network blocks or units and time-multiplexing these blocks to form the specific neural architecture required for any designated task.
More specifically, the neuroprocessor achieves its compactness and cost effectiveness by employing a combination of bit-serial and bit-parallel techniques in the implementation of the neurons of the neuroprocessor and reduces the number of neurons required to perform the task by time multiplexing groups of neurons from a fixed pool of neurons to achieve the successive hidden layers of the recurrent network topology. For most recurrent neural network vehicular applications such as misfire detection, a candidate pool of sixteen silicon neurons is deemed to be sufficient. By time multiplexing, the sixteen neurons can be re-utilized on successive layers. This time-multiplexing of layers radically streamlines the VLSI architecture by significantly increasing hardware utilization through reuse of available resources.
REFERENCES:
patent: 5943659 (1999-08-01), Giles et al.
Aranki Nazeeh
Feldkamp Lee A.
James John V.
Marko Kenneth A.
Puskorius Gintaras V.
Davis George B.
Ford Global Technologies Inc.
Lippa Allan J.
LandOfFree
Automotive engine misfire detection system including a... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Automotive engine misfire detection system including a..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Automotive engine misfire detection system including a... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2921398