Reserch Of Motor Turn-to-turn Insulation Test Alorithm Based On Artificial Intelligence

Motor is electromechanical conversion equipment, which has a wide range of applications and generally used in construction, industry, metallurgy, transport, home appliances and the lifeline of national economy and security fields. The operation stability of motor runs a direct impact on the stability of the economy. Motor reliability largely depends on the winding insulation, after long-term operation the motor winding withstand electrical, thermal, environmental factors and occur damage phenomenon. The electrical and mechanical properties of winding decreased. Therefore, we should establish the motor turn-to-turn insulation test system and regularly detect motor winding insulation to avoid the motor winding insulation aging, which has become to ensure safe operation and production in every walk of life.In this paper, use the detecting coil method to detect the fault of interturn short circuit in rotor windings and analyze air-gap induction electromotive force. The amplitude of leakage flux will be changed when the winding has the inter-turn fault. Use the wavelet analysis, wavelet packet decomposition and neural network algorithm as a theory, comprehensive analyze the performance and parameters characteristics of the air-gap induction electromotive force in rotor windings, select MATLAB algorithm as a software development platform, eslabish the motor turn-to-turn insulation test algorithm based on artificial intelligence. Thus realize the function to determine the location of fault winding.Simulating the fault diagnosis process of motor winding insulation, a variety of detection methods has been used to detect motor turn-to-turn insulation fault. Comprehensive system of analysis and evaluating the diagnostic methods, design the motor turn-to-turn insulation detection system. The design content mainly include fourier transform of air-gap induction electromotive force, the wavelet analysis of denoising the main flux, wavelet packet to extract the fault feature information, establish and train the BP neural network. Experimental test showed that the diagnostic system is accurate and effective to detect air-gap induction electromotive force in rotor winding. It will be high quality production and safe operation. It provides a simple and practical means of turn-to-turn insulation fault diagnosis and a new artificial intelligence diagnostic system in the turn-to-turn insulation fault detection is proposed, so that diagnostic system will be more scientific. With the increasingly development of fault detection and diagnosis techniques, insulation diagnostic techniques of rotor winding has theoretical significance for further. The artificial intelligence techniques applied to motor rotor winding fault diagnosis, it is useful to accurately determine the location of fault winding and improve the motor diagnostic efficiency.