The Research Of Method In The AC Motor Fault Diagnosis Based On Data Fusion

In order to solve some uncertain problems in the process of AC motor fault diagnosis. We developed a Data fusion method and introduced it's conception and principle. Meanwhile, we analyzed this method by an example of AC motor fault diagnosis. The diagnosis result shows that this method has it's advantages in the solve uncertain problems.This article is mainly about AC motor fault diagnostic method. We summarized the internal and external article in this field and discussed the BP net method and D-S theory method in the AC fault diagnosis. Besides, the main method—the joint of BP net and D-S theory was given. As those methods, we all summarized their characters and advantages.Firstly, the feature and the principle of the AC motor fault are given. Besides, we use the electric current sensor and the shocking sensor to measure the AC motor. In the end, we introduced the model of Multi-sensor data fusion to handle these problems of fault diagnosis.Secondly, we discussed the BP structure and the BP character and developed an appending momentum method which quicken the gathering speed. By an example of AC fault diagnosis, it shows that BP net has it's advantages in the some field.In order to solve the uncertain problem in the fault diagnosis, we discussed the application of the D-S theory in the fault diagnosis. By an example of AC fault diagnosis, it shows that D-S theory has it's advantages in this field.Finally, The synthetic method based on BP neural network and D-S theory is developed. This method contain both advantages in the BP and D-S. Based on several parameters which are gotten by multi-sensor on the AC motor, fusion diagnosis for AC motor is realized by several subnets. The diagnosis result show that this method offers good veracity and reliability to diagnosis.