Design And Research Of CPG-based Adaptive Control System

With the gradual expansion of the research scope of human nature,more and more robots are applied to some extreme,harsh,complex and even very dangerous special occasions.As a result,there is a growing demand for autonomous path-planning robots that can adapt to complex and even unpredictable environments.The four-legged android has become a research hotspot both at home and abroad because of its good load-bearing capacity,high stability and strong environmental adaptability.This paper aims to study the detection and adaptive walking of quadruped robot in complex environment.Select the appropriate elliptical nonlinear oscillator model and filter,set up the simulation model parameters in Simulink,and analyze the planned trajectory to verify the rationality of the adaptive gait trajectory.The measurement and control system is designed to analyze the force on the foot in the process of movement through the current foot force experiment,so as to replace the foot force sensor.Using Webots simulation platform and physical prototype to carry out walking experiments in different environments,to verify the rationality of adaptive walking strategy and control system.The main contents of this paper are as follows:Firstly,the control mechanism of gait trajectory,kinematics and central mode generator of quadruped robot is analyzed,and the mathematical model of elliptic nonlinear oscillator and filter is determined;the simulation model is established in Simulink to study the parameters of CPG,and the parameters are adjusted;the planned adaptive gait trajectory is analyzed,and the combination of the adaptive gait trajectory is demonstrated Rationality and accuracy.Secondly,the selection of hardware and the design of measurement and control system are carried out.For the detection system,The signals of the robot foot end force sensor are monitored by using the software of Ni Max,and the measured foot end force is collected.In the control system,based on the adaptive gait trajectory and kinematics solution,the upper computer software C#is programmed,the motor is controlled by CAN communication and the working current of the joint motor is collected,and the foot end force is calculated by the current,realize closed-loop control.Then,the linear relationship between the current and the joint torque is obtained through the motor Torque test,and the foot end force is calculated through the force Jacobian Matrix,the results show the rationality of the experiment and the accuracy of the force Jacobian Matrix and determinant.The motion of the robot in different environment is simulated by using Webots,the rationality of the adaptive gait walking strategy and control system is verified.Finally,through the physical prototype experiment,through the designed control platform,in the vacated,flat,Rectangular Block,pit environment for walking test analysis,the experimental results further demonstrate the rationality of the Central pattern generator based adaptive control system and walking strategy.