Research On Key Technologies Of Sliding Mode Observers For Permanent Magnet Synchronous Motor Control System

With the increasing application field of permanent magnet synchronous motor(PMSM),to meet the high efficiency and high reliability operation of the PMSM in a complex environment,it is necessary to solve the stability problems caused by the change of internal parameters and external load disturbance of PMSM.The method of sliding mode observers compensates and controls the system by observing variables,which effectively suppresses the influence of the change of internal parameters and external disturbance of the system,so as to improve the performance of the system.Therefore,this dissertation takes the speed loop control algorithms,current loop control algorithms and sensorless control algorithms of the control system for PMSM as the research objects,and studies the research of speed loop sliding mode control technology,current loop decoupling control technology,sensorless control technology and data storage technology of the control system for PMSM based on sliding mode observer,which has great significance in theory and application.Firstly,PMSM model is built and the concept of sliding mode surface in sliding mode control is analyzed to finally select a global integral sliding mode surface.The speed loop to maintain a continuous control law and designs the sliding mode motion to run through the sliding surface of the whole system,causes the problem that the conventional integral sliding mode surface produces a integral saturation effect under the condition of large initial errors and disturbance,which leads to the overshoot of the driving motor and even the instability of the whole system.A speed sliding mode composite control strategy for PMSM based on load torque observer is proposed.The observer estimates the disturbance and the observation feedforward value to the output of the speed sliding mode controller so that the small switching gain can meet the requirements of resisting disturbance.The chattering of the system is weakened and the response characteristics of speed tracking are improved,then the contradiction between the adjusting ability of the switching gain of sliding mode control to the load disturbance and the inherent chattering is solved.At the same time,the theories of the three observers is analyzed,and the sliding mode observer with absolute advantages in convergence speed and parameter design is selected as the load torque observer.The sliding mode observer is added to enhance the immunity of the speed ring sliding mode control and the ability to suppress chattering.The decoupling effect of voltage feedforward decoupling control,internal mode decoupling control and current deviation decoupling control under the condition of motor parameters change is not ideal,and the phenomenon of current coupling is analyzed in this dissertation.In response to these problem,a compound decoupling control strategy based on sliding mode observer is proposed and uses a sliding mode observer to compensate for current coupling.The tracking characteristics of the current sliding mode observer,the estimated value of the stator current as the state variable,and the state variable is fed back to the system input terminal to compensate the current error value.This method not only achieves better compensation control for the current loop,completely decoupling the d and q axis currents,but also ensures the dynamic response speed of the system and improves the robustness of the system to the disturbance of motor inductance parameters.In order to improve the reliability of the control system and make the system run stably under the condition of in sensorless control system of PMSM,the conventional sliding mode position and speed observer is analyzed.The approach law used in this method is the conventional exponential approach law,and the switching function produces large chattering phenomenon in the back EMF estimation,which can filter out the chattering through the low pass filter,but the large observation error is caused.On the basis of combining Super-twisting algorithm higher order approach law and error sliding mode surface,a Super-twisting algorithm sliding mode observer method is proposed to effectively suppress sliding mode chattering while eliminating low pass filtering to improve observation accuracy.After analyzing the influence of PMSM parameters on the amplitude and the phase of the observed results,an adaptive Super-twisting algorithm sliding mode observer method for motor parameters is proposed.The adaptive law of motor parameters is designed by using Leonov's stability law,so that the observer parameters are updated in real time.The rotor position and the speed estimation are not affected by the change of motor parameters,and the strong robustness is owned.In order to facilitate the experimental debugging progress of sliding mode observer algorithm,effectively analyze the causes of fault,reduce the difficulty of fault detection and improve the efficiency of fault processing,it is very important to store debugging parameters and fault data efficiently.An EEPROM memory is usually connected to the peripheral circuit of the main control chip,and the data is stored through I~2C bus,which is convenient for analysis.However,the conventional I~2C communication method of EEPROM data storage has the problem of occupying the main cycle time,which affects the important state identification and counter processing in the process of test debugging.An EEPROM data storage method which does not occupy the main cycle time is proposed.The I~2C function of this method does not follow the hardware action to single step execution,but performs other tasks to release the wait time.Without occupying the execution time of the main loop,the method can store and read the data under any memory address and any transmission data length,make efficient use of EEPROM storage,and improve the utilization of DSP.The data storage method can provide data support for the experimental debugging of sliding mode observer algorithm efficiently.Developed a permanent magnet synchronous motor drive controller to complete the hardware designed,software designed and system experiment verification of the drive controller.