Model Predictive Control Of Bearingless Permanent Magnet Slice Motor And Its Displacement Sensorless Technology

With the continuous improvement of China’s manufacturing capacity,higher requirements have been put forward for the optimization and upgrading of the manufacturing industry,and the development strategy of "Made in China 2025" is clearly pointed out to improve the electrical drive equipment to promote the upgrading of the high-end manufacturing industry.In order to solve the common friction problem of traditional motors,The bearingless permanent magnetic slice motor(BPMSM)is taken as the research object in this dissertation,which has the advantages of no wear,small size,good sealing and high power density,and can be used as the drive core of the pump to realize the pollution-free transmission of liquid.There is great potential for development in special transmission fields such as life sciences,aerospace and semiconductor processing,which require high precision and cleanliness.1.The BPMSM is taken as the research object in this dissertation,the working principle,mathematical model,model predictive control and non-displacement sensor technology are studied,and the main work and results are as follows:2.The research value and background of bearingless permanent magnet slice motors are summarized and reviewed,and the research status of bearingless lamella motors,the research status of model predictive control and the research status of non-displacement sensors at home and abroad are summarized.After analyzing the basic structure and working principle of BPMSM,the mathematical model of BPMSM was derived.3.Aiming at the problems of signal delay in digital control systems caused by the traditional control strategy of BPMSM,the model predictive control method is proposed to control the BPMSM.In the torque control,the model predictive torque control(MPTC)is used for torque control and the three-vector model predictive suspension force control(TMPSFC)is used for suspension force control,cost function and action time are given in combination with the actual mathematical model of BPMSM,and the candidate voltage vector combination is optimized,and the second-order generalized integrator(SOGI)observation flux with higher observation accuracy is given.Finally,the BPMSM model predictive control system is constructed,and the simulation shows that the proposed method has superior performance in terms of dynamic response,control accuracy and anti-interference ability compared with traditional vector control.4.Aiming at the problems of cost increase and volume increase caused by the BPMSM displacement sensors in engineering applications,a non-displacement sensor technology based on the BP neural network is proposed to detect the rotor displacement.The BPMSM displacement subsystem was established,and on the basis of identifying the left reversible of the BPMSM displacement subsystem,the left inverse system was constructed with the BP neural network,and the BPMSM displacement detection was realized after series with the model predictive control.Finally,simulation is used to verify that the proposed displacement self-detection method has superior tracking ability.5.The digital control experimental platform with DSP as the core was developed,the hardware circuit design part and software control algorithm part of the motor control system were designed,the visual host computer interface based on Lab VIEM was developed,and the proposed model predictive control was accelerated,interfered with,floated and other experiments,and the results showed that compared with the traditional vector control strategy,the proposed control strategy has the advantages of good robustness,high control accuracy and fast dynamic response.