Research On Model-free Non-cascade Speed Control Of PMSM Drive System Under Voltage And Current Constraints

Permanent Magnet Synchronous Motor(PMSM)has been widely employed in the intelligent manufacturing,servo system,electric vehicles,home appliances and other industrial applications due to its advantages,such as high power density,high efficiency and easy maintenance,etc..To achieve excellent speed control performance,the cascade control structure is favored in the PMSM drive system because of its intuitive physical concept and decoupled loops.However,the dynamic performance of speed loop is unsatisfactory as a result of the limited control bandwidth of cascade control structure in the PMSM drive system.Recently,the direct speed control of non-cascade control structure is proposed to break the traditional structure of cascade control structure.With merging the outer speed loop into inner current loop and design single controller,the simultaneous speed and current control with different time scales can be realized.Therefore,the non-cascade control structure simplifies the system control structure and improves the speed dynamic performance.In fact,the uncertainties and unknown disturbances in PMSM drive system makes the practical operations complex and unpredictable,it is urgent to overcome the key technology of improving speed dynamics and current steady state performance simultaneously and achieve robust control under the dual constraints of inverter maximum output voltage and maximum armature current.Therefore,it is important in theoretical research and engineering application to improve the control quality and robustness of PMSM drive system undoubtedly.In order to realize the non-cascaded control of SMPMSM drive system,the state variable unifying speed and current with different time scales are first defined,then,the ultra-local model is established for real-time estimation of disturbances,such as uncertain parameters,unmodeled dynamics and unknown perturbations.Based on the ultra-local model,the non-cascaded model-free predictive speed control(NMF-PSC)of SMPMSM drive system is proposed.Generating the inverter voltage reference on the premise of stabilizing the Lypunov function,the NMF-PSC method can achieve stable operations of the system.Afterwards,the inverter voltage reference is modified by the graph analysis method to satisfy voltage constraint and current constraint in the linear modulation region.Therefore,in the case of eliminating reference current,the NMF-PSC method breaks through the key technology of generating inverter reference voltage that meets voltage and current constraints.Finally,the dynamic performance and robustness of SMPMSM drive system can be improved in this way.To further improve the steady-state control performance of SMPMSM drive system,the integral dynamic variable with adaptive coefficients is constructed based on speed error and q-axis current.Then,the designed adaptive integral dynamic variable and armature d–axis current are set as state variables to establish the ultra-local model of the drive system.Based on the ultra-local model,the non-cascade adaptive integral speed control(NAISC)are proposed,and the non-cascade adaptive integral speed controller is designed accordingly.In order to guarantee the system stability,the coefficients of controller are limited to certain ranges which are deduced by the Lypunov function.Meanwhile,the coefficients of controller are adaptively adjusted according to system operations,the further refinement of control action can be realized,the robustness and anti-interference performance of the system are improved as well.At last,the Lagrange optimal method is employed within the voltage vector hexagonal to obtain the optimal inverter reference voltage and improve the utilization of DC-link voltage under the dual constraints of inverter maximum output voltage and maximum armature current.With above-mentioned researches,based on the ultra-local model of SMPMSM drive system,the direct speed control laws deduced from traditional sliding mode surface and integral sliding mode surface are provided respectively.Then,the innovative design of dynamic weight factor is developed.According to the characteristics of high dynamic response of traditional sliding mode control and excellent steady-state performance of integral sliding mode control,the direct speed composite control(DSCC)of SMPMSM drive system is proposed by combining them with the dynamic weight factor.The determination basis for key control parameters of DSCC is obtained through Lyapunov function design and system stability proof.The dynamic weight factor can be designed in mutilple ways,Gaussian function with speed error is introduced as its implementation,which maps the running status of the drive system.Therefore,through the state perception of system operations by the dynamic weight factor,DSCC method can allocate the ratio of two sliding mode control actions under voltage and current constraints dynamically.Therefore,the smooth transition between dynamic state and steady state can be achieved to fully utilize the advantages of two sliding mode control methods,and the control performance and robustness of the drive system can be comprehensively improved in the complex and changeable operations.