Research On Predictive Control Strategies Of Brushless Doubly-fed Machines

The Brushless Doubly-Fed Machine(BDFM)has two sets of stator windings.The power windings are supplied by power grid,and the control windings are supplied by the inverter.BDFM can be controlled with slip power adjusted by the control windings.This new type of motor,which both operates as a motor or generator,greatly reduces the requirements for the converter capacity.And it cancels the brush and slip ring,so that the maintainability and reliability are enhanced.Besides,it has multiple operation modes and the ability to adjust power factor.All advantages show BDFM has broad application prospects in the field of large capacity speed regulation and power generation.Therefore,research on the control strategy of BDFM is of great significance to the practical application of engineering.In order to realize flux and torque decoupling,and solve the problem of the complexity of control system caused by so many parameters,in this article,the steady-state operation characteristics of the motor are analyzed.According to the working principle of BDFM,the mathematical model in the synchronous and static reference frame are established,and the state space models of BDFM are also given.On the basis,two control strategies are studied.The main contents of this paper are as follows:1)The operation principle of BDFM is analyzed,and the steady-state operation characteristics of BDFM are studied by simulation.Then,the mathematical model in the rotor reference frame is deduced.By taking negative conjugate transformation,state space models in the synchronous and static reference frame are derived,which are the basis of analyzing the dynamic and static characteristics of the motor and further studying control strategies.2)For existing literatures cannot realize the decoupling control of flux and torque of BDFM,a decoupling control strategy based on feedback linearization theory is studied.By using the state space model in the synchronous reference frame,the feedback control law is obtained,and the linear decoupling between output variables and input variables is realized.The design scheme of the control system and necessary conditions for the feedback linearization are obtained,then the control of flux and torque is realized.By utilizing simulation research and comparative analysis,it is proved that the feedback linearization strategy can realize precise decoupling of flux and torque,and has good regulation performance.3)For the problem that the feedback linearization control is highly dependent on motor parameters and its system structure is complex,a predictive flux angle control strategy is proposed.Firstly,the direct torque control and predictive control methods are analyzed and studied.Then,by using the state space model in the static reference frame,a flux angle(the angle between the power motor and the stator flux of the control motor)is selected for predictive control,instead of torque or current in conventional predictive control strategies,the dependence on parameters is reduced.Parameters required are only stator resistance,so the system structure is simple,and its stability is improved.The simulation results show that the strategy can effectively reduce torque ripples and improve current waveforms.It also has good dynamic and static performances,and the control effect is not affected when parameters change.