Research On Fault Diagnosis And Tolerance Technology In Automotive Engine

In recent years, with the vigorous development of the automobile industry, the auto structure improved constantly, especially the technical performance. How to use the scientific and effective engine test method instead of the backward one and ensure the auto running in safe and reliable state, which become the hotspot problems in today's technology research field. The sensors is very important unit in automobile engine fault diagnosis system, and the study on the sensors'data acquisition and fault diagnosis and the fault-tolerant technology, which also has meaningful significance. To sum up, this paper studies the auto engine fault diagnosis and fault-tolerant control techniques both comprehensively.Firstly, this paper illustrates the fault diagnosis and the fault-tolerant control structure of the multi-sensor system in automobile engine, meanwhile selects and analyzes characteristics and working principles of intake manifold pressure sensor. Then, based on the comprehensive study of fault diagnosis and the fault-tolerant technology, this paper studies both the federal Kalman filtering algorithm and artificial neural network algorithm in-depth. According to these two algorithm's superiority and feasibility of the link theory with practice, applies them to the design of fault diagnosis about intake manifold pressure sensor of auto engine. Meanwhile, this paper designs the overall frame of fault-tolerant control system and carries out Simulation Experiments at last.The research object of this simulation experiment is the common piezoresistive intake manifold pressure sensor in most automobile. Through the accurate parameter design and algorithm control, eventually achieve the desired goal——the signal's reconstruction and correction processing. Thus, the engine could work more efficiently and reliably, furthermore, to ensure the stability of the whole engine system.