| Linear Switched Reluctance Motor(LSRM)is a new motor structure evolved from the development of switched reluctance motor,which has the outstanding advantages of simple structure,low manufacturing cost,good robustness and high degree of adaptation to harsh operating environment.These advantages make them outstandingly superior in linear reciprocating motion and have become a research hotspot in the field of linear motor drives.Linear switched reluctance motors have been widely studied and applied in linear drive fields such as rail transportation,cordless lifts,deep mine hoist systems and wave power generation.As the thrust of linear switched reluctance motors fluctuates greatly,the control process is more complex and affects their operational performance,so the study of their control methods is extremely important.This paper proposes a model predictive control method based on a modular stator linear switched reluctance motor based on a model predictive control strategy,and improves on this by proposing a three-dimensional look-up table method and a BP neural network method,and carries out a simulation comparison with the more mature double closed-loop PID control method.Finally,experimental validation is carried out on motors of different structures to prove the superiority and generality of the control method proposed in this paper.The main work of this paper is as follows:(1)In this paper,the basic structures of modular secondary-side stator linear switched reluctance motors and modular primary-side stator linear switched reluctance motors are described in detail,and their phase change processes are explained separately.The basic operating principles of the two motor structures are analysed and the corresponding equations are derived to study their circuit equations,mechanical equations and electromechanical equations.The basic drive circuit of the motor is given and the different modes of operation of the circuit are explained in detail.Finally,a brief comparison is made between the motor structure covered in this paper and a conventional linear switched reluctance motor structure,highlighting the uniqueness and superiority of the motor to which the control method of this paper applies.(2)This paper explains in detail the control strategy of a linear switched reluctance motor based on model predictive control.A brief introduction to the idea of model prediction is given first.Based on the operating principle and structural characteristics of the modular stator linear switched reluctance motor,a control framework with PID control in the outer loop and model prediction control in the inner loop is built in conjunction with the corresponding control ideas,and the control principles and control processes of the two control loops are described in detail.As the accuracy of the prediction model is closely related to the inductance value of the motor,which is based on a three-dimensional function of motor position and winding current,the inductance unsaturated curve is fitted linearly and non-linearly respectively,and the best-fitting fifth-order Fourier function is selected after comparison,and the inductance proportionality coefficient at different currents is fitted with a third-order Fourier function to jointly construct the fitted inductance function.This is followed by a description of the way in which the switching signals of the motor are combined for the phase-selected operation of the control system.Based on the control block diagram,the control circuit of the motor is built on SIMULINK,the drive circuit on Simplorer and the motor model on Maxwell,and joint simulations are carried out.The simulation results verify the good control effect of the model prediction control method.Finally,the control principle of the double closed-loop PID control method is introduced and the joint simulation is carried out.The simulation results are compared with the model prediction control proposed in this paper,and the comparison results show that the thrust fluctuation and speed fluctuation of the model prediction control during phase change and normal operation are significantly reduced compared with the double closed-loop PID control,which verifies the superiority of this control method.(3)The control method has been improved by proposing the three-dimensional look-up table method and the BP neural network method.The control process of the three-dimensional look-up table method is first introduced,followed by the corresponding joint simulation results for the desired speed of 0.1m/s,0.2m/s and 0.3m/s.Compared to the control results of the model predictive control method,the thrust fluctuations and speed fluctuations of the motor during phase change are reduced by more than half,which is an excellent control effect.Afterwards,the principle of neural networks is introduced and a joint simulation system based on BP neural networks is constructed,whose simulation results are almost identical to those of the three-dimensional look-up table method.Finally,the simulation results of the above four control methods are compared and contrasted to verify the superiority of the proposed control method and the improved method in thrust fluctuation and speed fluctuation suppression.(4)In this paper,a corresponding experimental platform is built and experimentally verified based on the proposed control method and simulation system.Firstly,the experimental platform is built based on a modular secondary-side linear switched reluctance motor,and the hardware part for converting the control signal to electrical signal is introduced in detail.The superiority of the speed fluctuation of the control method proposed in this paper is verified.Since the motor current is too small to reach the saturation zone,the experimental results of the three-dimensional look-up table method and the BP neural network method are not significantly better than the model prediction control method.In order to verify the generality of the control method proposed in this paper,an experimental platform is built based on a modular primary-side stator linear switched reluctance motor,and the selection of each control interface and the control method of each control board are described in detail.The neural network method with good control effect and simple construction was then selected for experiments with a rated speed of 0.12m/s on the modular primary-side linear switched reluctance motor experimental platform to verify the generality of the control method proposed in this paper. |
PDF Full Text Request