Research On Vector Control Technology Of Interior Permanent Magnet Synchronous Motor Based On EMPC

Interior permanent magnet synchronous motor(IPMSM)is featured by small size,high power density and strong load capacity.Therefore,it is widely used in electric vehicles.In the processing of high-performance control application by IPMSM,vector control technology has played a leading role for a long time.By virtue of its excellent dynamic performance,model predictive control(MPC)has been extensively applied in industry.However,due to its difficulty in overcoming large computation quantity of algorithm,its application in motor driving has been restricted for long.The explicit model predictive control(EMPC)reduces computation quantity of algorithm by calculating optimization process offline.By virtue of its simple performance tuning and strong constraint handling capacity of controller,it is of great significance in improvement of IPMSM dynamic performance.With IPMSM as the research object,a study was made on EMPC algorithm in this thesis.First,the mathematic model of IPMSM was established and the vector control technology was briefly introduced.Aiming at non-linear problems of mathematic model,three coupling terms,?_ei_d,?_ei_q and i_di_q were converted into measurable disturbance,realizing the linearization of IPMSM mathematic model.With consideration to linearization of mathematic model and control requirements of vector control technology,appropriate state variables were chosen to establish EMPC predictive model.Second,for solving the large computation quantity in online optimization processing of MPC,the process to transform this problem into offline quadratic programming(QP)problem by EMPC were derived in details.With respect to voltage and current restriction of motor driving system,it was proposed to approximately replace the current and voltage limit circle with regular hexagon and convert it into linear constraint of QP problem by EMPC.In order to improve anti-load disturbance capacity of EMPC,based on nonlinear mathematic model,a TITO Luenberger torque observer was constructed.This thesis also provided the derivation method of EMPC closed loop transfer function and made detailed analysis on influenc e tendency of each parameter in parameter matrix on closed loop pole and the method to determine value.In the meanwhile,a simulation verification on IPMSM in all-speed range was made.Analysis was also made on the existence of direct-axis steady state current error in IPMSM constant power area.In order to solve this problem,it was suggested to replace the ?_e in ?_ei_d and ?_ei_q with the constant ?_i,establish external voltage compensation mechanism,derive constraint processing and parameter selection method,and make simulation analysis in constant power range of IPMSM.Finally,an experiment platform based on Infineon XMC4500 was erected and the software program was written.A tuning program for PI parameter was put forward,making contrast simulation between PI controller and EMPC controller.According to simulation results,EMPC can realize non-overshoot control,and follow any given order of speed well.While ensuring a good rapidity,it also can present a good stability,thus realizing a dynamic performance better than that of PI controller.