Research On Model Predictive Current Control Of Induction Motor

Compared with traditional high-performance AC speed regulation strategies,Model Predictive Control(MPC)has the advantages of simple working principle,flexible control and strong adaptability to deal with nonlinear constraints.In recent years,with the rapid development of digital processors and the improvement of predictive control theories,MPC has attracted wide attention in the field of power electronics,especially in motor drives.The basic research of Model Predictive Current Control(MPCC),which is an important branch of MPC,has been relatively complete.The research focus of MPCC has shifted to taking the advantages of multi-objective,multi-constraint and non-linearity,so as to refine the predictive model,optimize the control structure,improve the control performance and the robustness of MPCC system.In this paper,aiming at the application of MPCC in induction motor drives,the following research is carried out.(1)The mathematical models of induction motor in three-phase stationary coordinate system,αβ coordinate system and dq coordinate system are established.The structure and working principle of MPC are introduced and analyzed,which lays a theoretical foundation for the design of MPCC system.(2)A MPCC system based on the current-mode-based flux observer and switch constraint is studied.Firstly,the current-mode-based flux observer discretized by trapezoidal method is used to observe the rotor flux;Secondly,on the premise that the inner loop of MPCC is equivalent to the first-order inertia loop,the speed controller and flux controller are designed by engineering design method.Then,while constructing the discrete current predictive model,the cost function of MPCC is analyzed,and a cost function with switch constraint is designed on this basis.Finally,simulation and experiment verify the control performance of MPCC algorithm and the effectiveness of switch constraints.(3)An improved MPCC system based on the full-order flux observer and delay compensation is researched.The discrete full-order flux observer with feedback gain matrix is designed by analyzing its working principle and deducing its mathematical model.The pole placement method applied in the design process is expounded as well.At the same time,the cause of the delay is analyzed.Then,the two-step prediction of stator current and the future reference value of stator current calculated based on extrapolation method are used to compensate the delay.Drawing support from simulation and experimental platform,the corresponding comparative simulation and experiment are completed.The observation effect of the full-order flux observer and the effectiveness of the delay compensation strategy are verified.(4)The rotor resistance online identification strategy based on instantaneous reactive power Model Reference Adaptive System(MRAS)is introduced into the above two MPCC algorithms to improve the system.In order to study the influence of motor parameter change on the control effect of the system,the parameter sensitivity of induction motor is analyzed systematically and theoretically.Taking the rotor resistance,which is sensitive to the change of parameters,as the main research object,the voltage model and current model of instantaneous reactive power of the system are selected as the reference model and the adjustable model in MRAS,and the estimation value of rotor resistance is obtained based on Popov’s hyperstability criterion.The simulation and experimental results show that the rotor resistance online identification strategy based on instantaneous reactive power MRAS can effectively improve the control performance and robustness of the two MPCC systems above.