Research On Adaptive Speed Observer Of Induction Motor Based On Sliding Mode Theory

The induction motors are widely used in variable frequency speed regulation systems because of their characteristics of simple structure,strong durability,high reliability and low cost.The high-precision speed regulation of induction motor relies on the closed-loop control of the speed.In practical engineering applications,speed sensors such as encoders or resolvers are usually used to obtain the motor speed,but this increases the cost of the speed control system on the one hand and reduces the reliability of the system on the other hand.The speed-sensorless control can abandon the expensive speed sensor and help simplify the control structure,so it has been extensively studied.With the application of speed-sensorless control in practical engineering,it is required not only to achieve smooth operation under ordinary working conditions,but also to be robust to external disturbances and internal parameter perturbations.The sliding mode control is naturally robust.The sliding mode control is naturally robust.Therefore,this article will combine the sliding mode control theory to study the speed adaptive observer in the induction motor speed-sensorless control system to improve its robust performance.This ensures the stable operation of the induction motor drive system when subjected to external disturbances and internal parameter perturbations.The specific research contents are as follows:By applying the coordinate transformation to the induction motor model in three-phase coordinate system,the state equation of induction motor in synchronous rotating coordinate system is derived to realize the decoupling control of excitation and torque,so as to achieve the same control structure as DC motor.And then,the vector control system based on rotor field orientation is constructed.This article analyzes the causes and solutions of the sliding mode control chattering phenomenon,and gives a key introduction to the reaching law and high-order sliding mode methods.To improve the robustness of the speed observer,a first-order sliding mode observer similar to the adaptive full-order observer is designed according to the sliding mode control theory.Based on the Lyapunov stability theorem,the stability proof and parameter selection range of the observer are given.To supress the chattering of the observer,by adaptively adjusting the gain according to the stator current error,the double power reaching law is introduced to the sliding mode observer as a feedback term.Through convergence time analysis and discrete analysis,the second-order sliding mode characteristic of double power reaching law is proved.The simulation and experimental results show that the speed adaptive observer based on double power reaching law is robust and effective to weaken the chattering.However,the high-frequency switching term which causes the chattering still appears in the speed adaptive law.In order to solve the problem metioned above,a speed adaptive observer based on the second-order Super-Twisting algorithm is proposed.The state equation of the induction motor is reconstructed by defining the new state variables,and then combined with the second-order ST algorithm to supress the chattering,a rotor flux observer is designed.Then based on the MRAS principle,the ST sliding mode observer is considered as the reference model with the current model of the rotor flux linkage selected as the adjustable model,and then the speed adaptive law is designed to realize the speed observation.This solves the problem of high-frequency switching item appearing in the speed adaptive law.And in view of the traditional forward Euler discretization suffers from the problem that the discretization error increases with the speed.This paper points out the contradiction between the discretization complexity and the error in the traditional discretization method through theoretical analysis and then proposes an improved Euler discretization method.The proposed method can balance the discrete complexity and discrete error,and is very suitable for digital control systems.Simulations and experiments verify the effectiveness and superiority of the proposed observer and the improved Euler discretization method.