Dual-vector Model Predictive Control Of Brushless DC Moto

Brushless DC motor has the characteristics of high power density,high efficiency and reliable operation.It is widely used in aerospace,automotive,electric tools,medical devices and many other fields.Compared with the traditional vector control.Model predictive control can satisfy the system constraints.Achieve faster response speed and higher steady-state performance.To further improve the reliability of the control system.In this paper,the sensorless technology and dual vector model predictive control are combined to carry out indepth research.The main contents of this paper are as follows:Firstly,the traditional vector control strategy of Brushless DC motor is studied.Vector control can suppress torque ripple effectively.It has the characteristics of small torque fluctuation,high operation efficiency and no noise pollution.However,the system structure is complex.Difficult parameter adjustment and other defects.In this paper,the surface mounted brushless DC motor is studied.The control strategy of direct axis current id=0 is adopted.Based on DSP28335 controller,an experimental platform is built for experiments.It lays a theoretical and experimental foundation for the follow-up study of control strategy.Secondly,for the traditional vector control,multiple PI controllers are cascaded.It is difficult to ensure the selection range of parameters under wide speed regulation and wide load conditions.Difficult parameter adjustment.The model predictive control of Brushless DC motor is deeply studied.Because in the traditional single vector model predictive control.There is only one effective voltage vector in any sampling period.Although the control is simple and the dynamic response is fast.However,there are still large control errors.To further improve the control accuracy of the system.This paper focuses on the star dual vector model predictive control and orthogonal decomposition dual vector model predictive control.The utility model is characterized in that the action time and amplitude of the voltage vector are adjustable.The selection range of voltage vector is effectively increased.This series of methods not only have the ability of fast dynamic response.And its control accuracy is comparable to that of vector control.Simulation and experimental results verify the effectiveness of this series of methods.Finally,in order to further improve the reliability of the control system.A sensorless dual vector model predictive control algorithm based on sliding mode observer is studied.Considering the inherent chattering in the sliding mode observer.The switching function of sliding mode observer is studied.The approach law and convergence effect of four switching functions(symbolic function,saturation function,sigmoid function and new switching function)are analyzed and compared.The results show that the sigmoid function with variable reaching law and the new switching function are better than the sign function and saturation function in suppressing chattering.The experimental results verify the feasibility of this method.