Research On The Model Predictive Control Of Brushless DC Motor

In recent years,the brushless DC motor(BLDCM)has quickly gained wide attention due to its reliable operation and wide speed range.However,due to technical problems such as suppression of torque ripple and sensorless control in BLDCM,it is rarely used in high precision and high reliability applications.Model predictive control(MPC)has the advantages of good robustness and strong control effect,which can effectively suppress the nonlinearity,parallelism and inaccuracy of the process,and can be used for each of the controlled variables and manipulated variables in the system process.Constraints are processed.Applying MPC to BLDCM control has practical significance for improving system performance and suppressing BLDCM torque ripple.Firstly,this dissertation introduces the current development status of BLDCM,summarizes and summarizes the research results of low torque ripple and nodel predictive control and analyzes the basic structure,working principle,mathematical model,driving method and current sampling of BLDCM.In the way of analyzing the reason of commutation torque ripple and its suppression strategy,the model prediction method is proposed to reduce the torque ripple of BLDCM.The model prediction strategy of continuous control set is selected to design the current loop.The current model is divided into two parts:non-commutation period and commutation period.The adaptive prediction model is selected,and a reasonable cost function is selected to constrain and switch.The current prediction model is a way to reduce the torque ripple caused by current commutation.The BLDCM double closed-loop control system based on model predictive control is designed and simulated.The simulation waveform has fast response speed and short adjustment time.The simulation results verify the effectiveness of the MPC algorithm.Finally,use DSP as the core,the BLDCM control system experimental platform was established,and the performance of the MPC algorithm was verified by writing the software program.Through comparative analysis,compared with the traditional PI control algorithm,the model prediction can effectively suppress the commutation torque ripple,improve the static performance and dynamic performance of the system,and improve the stability of the system operation.