Research On Hybrid Intelligent Fault Diagnosis Technology And Application In Fault Diagnosis Of Induction Motors

Asynchronous motor, as the most widely used dr iving device, plays an irreplaceable role in every field of people's life, its fault often brings unexpected disaster. So it is very important to study the fault diagnosis technology of asynchronous motor. The technology of asynchronous motor fault diagnos is by the previous pure rely on experience and use mathematical model of traditional methods and later using a combination of various sensors and artificial intelligence method, obtained the rapid development in the past few decades. However, various single artificial intelligence methods mainly used at present also has many deficiencies, especially for the diagnosis of complex system, the diagnosis often results in low accuracy. O n the basis of this, this paper proposes a new method based on hybrid intelligent fault diagnosis technology, which is based on the advantages of different artificial intelligence diagnosis methods and combined with different signal processing methods.The asynchronous motor fault diagnosis for research object, first from the background and significance of the research are introduced, and briefly discusses the development status and trends of the domestic and foreign asynchronous motor fault diagnosis technology, definition and brief introduction of hybrid intelligent fault diagnosis technology at the same time, and from the stator current signal and the vibration signals of a variety of common fault types of asynchronous motor is analyzed. Then the two steps of EMD decomposition and Hilbert spectrum analysis in Hilbert Huang transform are studied, and the method is used to deal with the fault signal, the characteristic signal of the stator current signal and the vibration signal is extracted, and the fault feature vector is constructed. In the fault type identification method, the current widely used BP neural network, through in-depth study of the network, to understand and grasp the advantages and disadvantages, and aiming at the shortcomings of the BP neural network, and puts forward the particle swarm algorithm to optimize the BP neural network hybrid artificial intelligence technology, and focuses on the analysis and study of the working principle and algorithm of particle swarm algorithm and how to optimize the BP neural network.Finally, the fault diagnosis system model of ind uction motor based on hybrid intelligent fault diagnosis technology is constructed, and take 4 kinds of working conditions of the induction motor is normal, the stator turn to turn short circuit, the rotor broken bar and the bearing fault as the diagnosis object, simulation and analysis of the system model. Through the simulation results, it is proved that the fault diagnosis system of asynchronous motor based on hybrid intelligent fault diagnosis technology is correct and feasible.