Model-free Adaptive Control Based On Algebraic Estimation And Its Application In Knee Exoskeletons

With the development of science and technology,the objects and applications in control systems are becoming more and more complex.Mathematical models based on mechanism modeling cannot fully effectively describe the dynamic characteristics of the system.As an emerging method without relying too much on mathematical model information,the model-free control method has gradually become a research hotspot.Currently,there are many model-free control methods.With the system’s non-linearity,uncertainty and disturbance becoming more and more complicated,the model-free control method of intelligent PID structure also needs to further optimize the internal structure and improve the control accuracy and self-adaptation.Based on the intelligent PID control of French Fliess,this paper studies model-free control method in depth,and proposes a variable model-free inversion control method based on algebraic estimation.This method reduces the order of high-order systems through extremely local modeling,estimates the dynamic characteristics of the system using an algebraic method and compensates them to the inversion controller,designs a variable control channel gain,and optimizes the system structure.It also uses genetic particle swarm optimization to tune the controller.parameter.The research content of this article is mainly as follows:First,in order to quickly and accurately estimate the dynamic characteristics of the system,a non-progressive astringent algebraic estimation method is proposed.First,the research background and theoretical basis of the algebraic method are introduced;secondly,an estimation expression based on the algebraic method is derived for the knee exoskeleton system;then the accuracy of the estimation algorithm is verified with Matlab simulation;finally,the non-progressive convergent algebraic estimation method is explained Features and applications.Secondly,in order to further improve the accuracy of intelligent PID control(iPID),a model-free backstepping control method(MFBC)based on algebraic estimation is proposed.First introduce the design steps of the intelligent PID control method;then derive the inversion control expression to replace the PID controller and prove its stability;finally,combine Matlab simulation to compare the control effects of iPID and MFBC,highlighting the effectiveness and superiority of MFBC.Then,for the parameter setting in MFBC,the genetic algorithm particle swarm optimization(GAPSO)algorithm with particle cross-update is introduced,and a model-based inversion control method based on algebraic estimation optimized by the genetic particle swarm optimization(GAPSO-MFBC)is proposed.First introduce the basic process of particle swarm optimization(PSO);then use the crossover and mutation operations in the genetic algorithm for the update part of the particle swarm algorithm;then use the GAPSO algorithm to search for the better parameters of the controller;and finally verify the GAPSO by Matlab simulation The effectiveness of MFBC and the rationality of GAPSO algorithm.Finally,in order to optimize the internal structure of MFBC and further improve its adaptability,an α adaptive model-free backstepping control method based on algebraic estimation(α-AMFBC)is proposed.First explain the need for adaptive parameters in model-free control;then use gradient projection algorithm and recursive forgetting factor least squares algorithm to estimate the gain of the control channel gain α,and finally compare the effects of the four methods through Matlab simulation to verify the effectiveness of the algorithm Performance and superiority,it promotes model-free control to model-free adaptive control.