Research On High-order Sliding Mode Control Strategy Of Speed Sensorless Induction Motor

Induction motors(IMs)have been widely used in practice because of its characteristics of simple structure,durability,and low cost.The induction motor speed control system based on the vector control strategy can already achieve the same performance as the DC speed control system.With the higher requirements of industrial production on system volume,cost,reliability,etc.,speed sensorless control technology has gradually become the focus of academic research.This paper is aimed at two key parts of current controller and speed observer in speed sensorless induction motor vector control system.Focusing on the complex conditions in the system's internal parameter changes and external load distur bances,enhance control precision and robustness of the system by introducing high-order sliding mode control theory.The specific content of the paper is as follows:Firstly,the dynamic performance of traditional PI current controller is quantitatively analyzed by linear control theory.The motor equation is used to derive the influence of the change of the rotor rotor time constant on the motor flux linkage,output torque and stator voltage,and the importance of enhancing the robustness of the current loop parameters is proved in a graphical way.Then a first-order sliding mode current controller is designed and the stability of the controller is proved.In order to suppression the chattering,the paper proposes a high-order sliding mode algorithm inspired by second-order super-twisting algorithms and gives a mathematical proof of stability and convergence time.On this basis,a high-order sliding mode current controller is designed.The contrast experiments show that the high-order sliding mode algorithm can effectively suppress chattering and exhibit better robustness under the conditions of speed variation,load disturbance and motor parameter perturbation.In order to improve the accuracy and robustness of the rotational speed and flux linkage observation,this paper theoretically analyzes the defects of the traditional first-order sliding mode observer in terms of state quantity convergence and control law chattering level.Considering the dynamic quality of the state quantity reaching the sliding surface in the theory of approaching law,the stator current estimation information is introduced as a constant-speed term gain to design an improved exponential approach law,which enables the state variable to follow the system estimation state to adjust the convergence speed adaptively,forcing the system's chattering to decay.Further,the stator current and the rotor flux observer is constructed.Based on Lyapunov stability theorem,the stability of the observer is proved,and the speed adaptive law is derived.Then the effectiveness of the algorithm is verified by comparison experiments.Aiming to improve the convergence of the system,a fast terminal sliding mode surface is proposed to ensure that the state quantity and its first derivative converge to zero in a finite time.The decoupling of the ?? axis is achieved by defining the coupling term between the rotational speed and the flux linkage in the motor model as a new intermediate variable.The high-order sliding mode control law is obtained by integrating the switching term as a control quantity output.Then the paper establish the stator current observer and obtain the observation equation of the rotor flux linkage.Using the designed high-order sliding mode observer as the reference model,and the rotor flux current model is transformed into an adjustable model with the adjustable speed as the adjustable variable.Then the MRAS speed identification structure is built to realize closed-loop control of estimated speed.Moreover,the influence of current harmonics on the estimated speed is analyzed.Finally,the effectiveness of the high-order sliding mode algorithm is proved by simulation and experiment.