Electronic Fuel Control System Fault Diagnosis Approach Based On State Observer

With the rapid development of economy in our country, the total number of vehicle runs up, vehicle exhaust emissions already have become one of the main air pollution. So the Third Country Standard Regualition for Vehicle Pollutant Discharger has come true and On-Board Diagnosis (OBD) systems has been forced to set. In the future, exhaust emission regualition will be stricter and making cycle will be shorter.In order to meet the requirements of reducing energy consumption and exhaust emissions, the engine management system is becoming more functional and need to be supplied with more sensors and actuators, so engine electronic control system will become more complicated and needs more reliable sensors and actuators. In the situation, on-board fault diagnosis technology for engine sensors and actuators has become a focus at home and abroad.This paper mainly studies the application of the fault diagnosis method based on observers in the engine electronic control system. At first, we use the Beard Fault Detection Filter(BFDF) to isolate and diagnosis the failures of the sensors and actuators, and give robust designing for Beard Fault Detection Filter, then build the state-space model of the throttle-speed response and identificate all parameters of the model by I.D.Landau's identifications technique. At last, we bring in unknown input observor and use this technology to design a diagnosis strategy for the three key sensors(CPS, TPS and MAP) in the electronically controlled internal combustion engine. When any failure of the three sensors happens, such as constant gain varition, constant gain varition, zero-death, random failures and so no, the residual of the unknown input observor will obvious deviate from zero. Judging from whether the residual exceed the range permitted, we can know which sensor is at fault. Model simulation and the experiment results show that the method is effective.