Abstract:As important equipment for natural gas transmission stations, electric drive centrifugal compressor units are generally in continuous high-speed operation. Long-term high-load operation is easy to induce drive motor failure, resulting in huge economic losses and casualties. Compressor group condition monitoring, maintenance and repair are all closely related to motor fault diagnosis, making compressor group fault diagnosis very important. Therefore, the study of motor abnormal fault warning technology is of great significance for the prevention of centrifugal compressor accidents. In this paper, the motor fault warning model is established based on the multi-sensor data and second-order Volterra series. The obtained prediction data is compared with the fusion data collected by the sensor to obtain a range under the normal operation of the motor, that is, the numerical set [66, 67]. Predicting the development trend of motor running based on motor monitoring data. The model is used to predict the simulation data and motor data, and the results verify the correctness of the model.

Research on Motor Fault Warning Technology Based on Second-Order Volterra Series