Research On Identification Method Of Load Torque And Moment Of Inertia For Asynchronous Motor

Three-phase asynchronous motor is widely used in industrial electric vehicles due to their advantages of durability and low cost.Because of different operating conditions,the load torque and moment of inertia of the motor system change frequently,which cause the fluctuation of motor speed.In order to improve the anti-disturbance performance of the system,it is necessary to identify the above parameters online and adjust the controller dynamically.For load torque,most existing methods are based on motion equation to construct observer,but they rely on high-precision encoder and not suit for industrial electric vehicles with low-precision encoder.And the load torque observers based on rotor position depend on the moment of inertia.For the moment of inertia,mostly existing identification methods have the problem that the convergence time and accuracy cannot be satisfied at the same time.To solve the above problems,this paper proposes a super-twisting sliding mode load torque observation method based on low-precision encoder.Meanwhile,an improved model reference adaptive moment of inertia identification method is proposed.The main research work and innovations are as follows:In this paper,a load torque observer based on super-twisted sliding mode algorithm for low precision encoder is proposed.The observer starts with the rotor position measured by the low-precision encoder and fit its data,and then the sliding mode observer was input.The sliding mode surface was constructed with the rotor position observation error and its first derivative,and the super-twisting control law was used to ensure the finite time convergence and reduce the chattering of the system.After proving the robustness and stability of the system,the simulation proves that the proposed method can quickly and accurately track the variation of load torque,and effectively suppress the speed fluctuation by feeding forward the observed value of load torque.To solve the problem that the fixed adaptive gain can not satisfy the speed and accuracy of convergence at the same time for the identification method of inertia based on the model reference adaptive algorithm,an improved adaptive gain design method was proposed.The method adjust the adaptive gain based on the difference of identification results at different time.The adaptive gain fast increase gain in order to improve the convergence speed when the moment of inertia changes,and gradually decreases in the identification stability stage to ensure the convergence accuracy.The method makes the adaptive gain change smoothly in the form of iteration,and avoids the chattering of identification results.Simulation results show that the improved model reference adaptive moment of inertia identification method can guarantee the convergence time and accuracy when the moment of inertia changes.Finally,the proposed method is verified on three-phase asynchronous motor experimental platform.The experimental results of load torque observation show that the proposed load torque observer can quickly and effectively observe the changes of load torque,and the speed fluctuation caused by load torque mutation can be greatly reduced by feeding forward the load torque observation value,and the anti-disturbance performance of the system can be improved.The identification experimental results of moment of inertia show that the improved model reference adaptive moment of inertia identification method can ensure the convergence accuracy while improving the convergence time.