Study On Fault Mechanism And Diagnosis Method For Automobile Wiper

Currently, vehicle noise pollution is increasingly serious, car manufacturers are increasing emphasis on automotive NVH characteristics, higher requirements are put forward for studies and fault diagnosis of the failure mechanism of the car motor noise and vibration. This topic originated from an automobile motor manufacturers noise reduction projects. The company produces a automobile wiper has the phenomena of excessive vibration, noise, and abnormal sound. At present, the traditional diagnosis method of artificial touch and ears is difficult to guarantee the accuracy of the wiper fault diagnosis, the study of failure mechanism have been done is not thorough and comprehensive enough, the diagnostic software has developed is not fast and accurate enough for online fault diagnosis. To this end, the failure mechanism and diagnostic methods of the automobile wiper is thoroughly studied and a wiper intelligent diagnostic software has been developed.Firstly, the wiper dynamic characteristics analysis has been done, and the failure mechanism for wiper is interpretated. The finite element model of the wiper parts are established by the method of comparing the result of modal experiment with the result of modal analysis and fixing the finite element parameters and structure Then by adding degree of freedom constraint on the finite element model to simulate the situation of the parts in the actual assembly conditions, by using the finite element analysis software to analysis the dynamic characteristics of various components of the wiper, to mastered the natural frequencies and vibration modes of the various components, as well as the modal stress of the rotor and harmonic characteristics of speed reducer shell and cover. Simulation and comparison of different bolt connection alternative to the effect of bolted joints dynamic characteristics is done by using different finite element connection method. Four fault types has been put forwadr according to the structural features and mechanism of the wiper. According to components dynamics analysis results and mechanic failure theory, explained wiper failure mechanism, explained the failure transfer pathway. Mainly explained the generation and transfer mechanism of worm abnormal sound and rolling bearing abnormal sound, as well as the frequency modulation phenomena of fundamental frequency, electromagnetic frequency and the meshing frequency, and occurrence of multi-failure phenomenon. Proposed the characteristic parameters of different failures, provided a theoretical basis for fault diagnosis by measuring the wiper external noise and vibration.Secondly, the fault diagnosis method has been studied. According to the principles and requirements of fault diagnosis based on neural network, selected the characteristic parameters and organize the training sample set, established a wiper fault standard sample database, as for the need of neural network training and the study of wiper failure. By studying the effect of the set up of BP neural network structure and training parameters on the accuracy of fault diagnosis, get the experience of setting the neural network structure and selecting the training method.Finally, a set of wiper intelligent diagnosis system based on BP neural network has been developed, which can achieve the fault diagnosis of a variety of wiper models. The system consists of two main parts, online detection and neural network training. Online testing program interface can achieve the fault diagnosis of a variety of wiper models. The neural network training program interface can train the neural network with the existing standard fault sample database, it can be used independently, and can train neural network for other identification purposes. The system has the performance of reliable, versatile, fast, easy to operate, it is suitable for the requirements of the wiper online detection.