基于支持向量机的汽车发动机故障诊断研究

he car engine as the power source, the running state of a direct influence on the vehicle's work conditions. The engine trouble all the fault of car of a high proportion, the high cost of maintenance, spend time long. Not in the disintegration of the engine, if can timely and accurate to its current technology conditions make a judgment, marshaling failure positions, points out that the cause of the problem and elimination method, will improve the car use of reliability and security. This paper studies the content is based on support vector machine (SVM) car engine fault diagnosis research, the main research idea is: to easy to test of vibration signal for fault diagnosis of the signal source, the comprehensive utilization of based on signal processing and energy of the variation of the failure feature extraction method, set up different fault type of training set, based on support vector machine method to realize the failure pattern recognition. Research results show that the method of automobile engine fault recognition has important engineering significance.