Active Disturbance Rejection Control For An Induction Motor Drive System

Since an induction motor is essentially a complex object with nonlinear,multivariable,strong coupling,time-varying parameters and large disturbances,it is important to improve the performance of induction motor drive system.However,external disturbances,parameter changes,fluctuations and other unforeseen factors affect the control performance of the induction motor,especially the robustness performance and the dynamic anti-disturbance characteristic need more optimization control.The Active Disturbannce Rejection Control(ADRC)has less reliance on the parameters of the controlled object model,and its inner oberser can realize the unforeseen disturbance elimination,so that many scholars focus on it.In this thesis,ADRC and induction motor control are combined to improve the robustness and dynamic disturbance rejection of the induction motor drive system.The specific research contents are listed as follows:Firstly,in order to solve the problem that the ADRC method for the induction motor control system has many difficult tuned parameters,so that ADRC for induction motor control system based on the Ant Colony Optimization(ACO)algorithm is proposed.ACO is integrated into the design of ADRC parameter.According to the feedback information from the induction motor,the self-learning ability of ACO works as the optimization mechanism,and the ADRC controller parameters are obtained by the self-learning ability of ACO after iterative calculation,so that the reliance on its control parameters is reduced under the ADRC control,and the performance of the controller is enhanced.This thesis focuses on the establishment of objective function and parameter optimization,comparison with the traditional methods in terms of effectiveness,convergence and optimization efficiency in the experiment.It can be conclude d that the ADRC robustness based on ant colony algorithm for parameter optimization is better than the traditional ADRC.Secondly,after the further research,it is found that the fixed pheromone volatilization coefficient of ACO is the main reason that affects its optimization performance.This is because the pheromone volatilization coefficient of ACO is constant and contrary to the actual local precise search,and the historical information is also contrary to the fast convergence of the algorithm.Only the optimization algorithm with dynamic update ability can further improve the optimization performance.Therefore,choose the Particle Swarm Optimization(PSO)algorithm with the ability to linearly update as the research object,and associate the dynamic updating process of the swarm algorithm with the actual optimization process.Based on the actual optimization process,the particle aggregation degree and the evolution speed are introduced to update the inertia weight adaptively.The induction motor system based on ADRC is used as the test object.ACO and APSO are compared on the effectiveness,convergence and efficiency,and the experimental results show that APSO has higher efficiency and convergence.Internal model control(IMC)is a robust control strategy,has the advantages of less adjustable parameters and strong robustness.However,there are two problems in the traditional current loop controller.Firstly,IMC has only one adjustable time constant of the filter.It is a one-degree-of-freedom controller and cannot achieve separation adjustment between tracking performance and anti-interference performance;Secondly,IMC cannot effectively eliminate the uncertain disturbances such as the model mismatch,parameter perturbation,and step load variation.This thesis selects the internal model control(IMC)for induction motor as a research case.Introduced by the thought of ADRC to solve the problems above,therefore IMC for the induction motor based on ADRC is proposed.State observer is planted into IMC to estimate the disturbance,and the system uncertainty and the external disturbance are treated as the total system disturbances.System state errors between the measured variables and the observed variables are treated as the system feedback.The appropriate feedback gain is analyzed and designed through the Jury stability judge,and the method of disturbance compensation is combined with the current loop of IMC.IMC for induction motor based on ADRC is verified in the aspects of the step load torque,parameter variation,and input disturbance under the experimental conditions.It is verified that IMC introduced by the thought of ADRC can improve the anti-disturbance performance of current loop.Finally,speed identification is a key part of the induction motor speed sensorless control.In order to suppress the influence of external disturbance on the identification result,such as measuring noise,step load torque and parameter variation.Learn from the basic thought of ADRC,it is introduced into the speed sensorless control of the induction motor.In this thesis,a speed estimation method based on active disturbance rejection observer(ADRO)is proposed.The system model disturbance is observed and compensated by state observer to eliminate the influence caused by rotor parameter variation on the observation accuracy.This method also avoids the influence of integral part and improves the observation accuracy of active disturbance rejection observer.Under the Lyapunov stability provement,the speed estimation method based on ADRO is designed and verified.As to two kinds of disturbance about the input and the observation,the anti-disturbance and the stability performance are analyzed.In the experiment,the speed estimation method based on ADRO is verified in the aspects of step load torque,parameter variation and input disturbance.In this thesis,it uses a 2.2kW induction motor as the control object and builds a simulation platform based on the Matlab/Simulink and an experiment platform based on the TMS320F28335 chip from Texus Instrument(TI)Company.The above-mentioned ADRC method based on ant colony algorithm,ADRC method based on adaptive particle swarm algorithm,internal model control method based on ADRC and sensorless control method based on active disturbance rejection observer are simulated and verified by experiments.The results show that the performance of the induction motor drive system in the aspects of robustness and dynamic disturbance rejection has been effectively improved.