Monitoring Method of Permanent Magnet Synchronous Motor Temperature Variation Signal Based on Model Prediction

Monitoring Method of Permanent Magnet Synchronous Motor Temperature Variation Signal Based on Model Prediction

연구 분야: Infrastructure

학회: International Conference on Advanced Hybrid Information Processing

In order to improve the monitoring accuracy and quality of permanent magnet synchronous motor (PMSM) temperature variation signal, and achieve the ideal effect of high-precision monitoring of PMSM temperature variation signal, model prediction is introduced, and the monitoring method of PMSM temperature variation signal based on model prediction is studied. The wireless sensor technology is used to collect the temperature signals of various parts of the motor, integrate, clean, replace and protocol the original data, establish a deep learning network model to predict the characteristics of the motor temperature variation, identify the motor temperature variation signal, combine the variation pruning and variation interval, and use the delayed reporting strategy to monitor the early warning of the motor temperature variation signal, complete the monitoring of temperature variation signal of permanent magnet synchronous motor based on model prediction. The experimental analysis results show that the recall rate and accuracy rate of the design method are above 90%, and maintain detection efficiency above 97%, the monitoring accuracy of the temperature variation signal of the permanent magnet synchronous motor is high.