Based on Engine Fault Diagnosis of noise Abstract With an increasing amount of car ownership, the engine failure rate is relatively high, so how to improve the efficiency of automobile engine fault diagnosis maintenance e an important topic. To the noise generated by an exception should be paid special attention while in engine fault diagnosis maintenance in certain parts .Generally speaking, from the noise characteristics, we can find the unusual parts quickly. Usually, the work is done by an experienced mechanic according to the sound to determine the cause of the malfunction, but find fault or not mostly depand on the level of personal experience ,and so it has an large affect on the efficiency of repair. In today's society because of the rapid development puter technology, we can use instruments to monitor the sound way, and then puter software to build prehensive and efficient fault diagnosis system which can help the master to find the causes, so that we can substantially increase Engine fault detection speed and accuracy, improve maintenance efficiency. In this paper, a microphone was used to record the fault noise in engine operation ,and then the spectrum analyzer was used to analysize frequency domain and the sound characteristic frequency vector was extracted, the BP work training was entered ,along puter software, it would ely identify the type of engine failure.
