Research And Implementation Of Online Inertia Identification For PMSM Servo System

With the continuous development of "industry 4.0",the requirements of industrial automation equipment are also improving.As an indispensable part of the industrial control field,the servo system affects the control precision and work efficiency of the actuator.The performance improvement of servo system is particularly important,and the on-line inertia identification technology is an important technology to improve the performance of servo system.The moment of inertia is the design parameter of the speed controller.Some variable working conditions will cause the moment of inertia mutation,resulting in parameter mismatch,and then affect the response and stability of the servo system.Therefore,this thesis focuses on the research of online inertia identification technology under variable working conditions.The main research contents are as follows:First of all,this thesis introduces the mathematical model of permanent magnet synchronous motor(PMSM)in various coordinate systems and the relationship between coordinate transformation,and the PMSM vector control strategy principle and simulation model are introduced and established in detail.In addition,according to the current,speed and position controller of the three-loop control structure of the servo system,the reasonable controller parameters are designed according to the actual physical significance,and the influence of the bandwidth of the control loop on its performance is analyzed emphatically,and the specific application of inertia identification is analyzed correspondingly.Then,considering the strong coupling relationship between the moment of inertia and the load torque of the servo system,an online inertia identification algorithm based on gradient correction method and extended Kalman filter is designed.The inertia identification algorithm is designed based on the gradient correction method and the torque observation algorithm is designed based on the extended Kalman filter.The combination of the two algorithms makes the online inertia identification algorithm can effectively identify the inertia in the case of variable torque.The effectiveness of the algorithm is verified on the simulation platform,and it has certain advantages compared with the traditional gradient correction method.Next,this thesis designs an online inertia identification algorithm based on Landau algorithm and disturbance observer.The algorithm takes Landau algorithm as the inertia identification module,disturbance observer as the torque observation module,torque observation value as the input of inertia identification module,and inertia identification value as the input of torque observation module.Through the simulation platform,the feasibility of inertia identification was verified under the conditions of different speeds and different torques,and compared with the traditional Landau algorithm,the identification effect of the proposed algorithm was better.Finally,the online inertia identification algorithm mentioned in this thesis is verified on the servo experimental platform,and the feasibility of the algorithm is verified on the built servo platform of various working conditions.The influence of variable parameters in the algorithm on the identification results is also verified on the experimental platform.Moreover,it is verified that the proposed algorithm has stronger adaptability than the traditional algorithm on the experimental platform.