Research And Application Of Active Disturbance Rejection Control On Active Prosthesis

With the development of intelligent control,sensor and micro electronic technology,the design and research of active prosthesis is paid more and more attention to both in China and abroad.Compared with the conventional passive prosthesis,active prosthesis can not only realize real-time environmental perception and autonomous control,but also can supply power to amputees in upstairs,downstairs,upslope and downslope.In order to realize the goal of real-time terrain detection and coordinate control,the research and design of the active prosthesis is carried out in three aspects,including information acquisition,terrain recognition and motion control:Firstly,a terrain recognition method based on laser sensor is proposed through analysis of human lower limb movement,and a hardware data collection system with laser distance sensor and inertial measurement unit is designed to acquire the laser distance and terrain height informationSecondly,in order to solve the problem of the terrain recognition rate and the recognition delay,the laser distance information and terrain height information are processed by wavelet de-noising algorithm,the feature parameters are extracted,then choose the probabilistic neural network for terrain recognition.The result shows that the system can effectively recognize the five kinds of terrains and improve the identification accuracy.Finally,the model of active prosthesis is established by Lagrange method,and an Active Disturbance Rejection Control method for active prosthesis is proposed aiming at the problem of parameter uncertainty in the process of modeling.The function of Tracking Differential,Extended State Observer and disturbance,NonLinear Error Feedback are simulated and verified,based on this,applied the ADRC algorithm to the control of active prostheses,an experimental simulation model is built in the MATLB/Simulink to do the active disturbance rejection control simulation experiment for active prosthesis.The result shows that the controller can quickly follow the system input and achieve good control performance.