Research On Gait Planning And Friction Compensation Control Of Quadruped Inspection Robot

The quadruped inspection robot has excellent ground adaptability and flexibility due to the discrete contact points between the legged structure and the ground.It is currently a research hotspot in the field of inspection robot.This project focuses on the two key technologies of stable walking and stable accuracy for quadruped inspection robots.This article conducts research on the following aspects from the perspective of gait planning and friction compensation control.This topic introduces the overall architecture of a quadruped inspection robot with series legs and full elbow,and takes the right front leg of the robot as an example to establish an equivalent link model.It briefly describes the analysis methods of posture and homogeneous transformation.Based on the improved DH parameter method,the forward and inverse kinematics model of a single leg of a quadruped inspection robot is constructed,the Jacobian matrix and statics model are derived.The Newton-Euler method is used to get the dynamic model of the single leg of the robot,which lays a theoretical foundation for the subsequent gait planning and friction compensation control.Aiming at the stability problem of quadruped inspection robot caused by complex working conditions,the workspace of the robot is drawn by Monte Carlo numerical method.On the basis of the known reach range of the foot end,the trajectory of the foot end is planned by piecewise polynomial,and the gait planning method using zero moment stability criterion is designed.The virtual simulation environment of the robot is built in MATLAB,and the diagonal trot gait simulation of the robot is designed.The results indicate that the gait planning method can achieve stable motion of the quadruped inspection robot during diagonal trot gait.A method for friction parameter identification and compensation control method is proposed to address the impact of nonlinear friction on the motion accuracy of quadruped inspection robot.An improved Lu Gre friction model with adaptive coefficient is used to model the joint friction torque,and an offline identification method for static and dynamic friction parameters is designed.The friction parameter values are identified through joint friction torque testing experiments.A backstepping adaptive compensation controller is designed based on the Lyapunov asymptotic stability criterion.The simulation results of friction compensation show that the tracking performance of robot joint using adaptive friction compensation control is significantly improved.Finally,experimental verification is conducted based on the physical prototype of the quadruped inspection robot.The mechanical structure and control system of the prototype are introduced,and the gait generation and friction compensation experiments of the quadruped inspection robot are conducted using the same control parameters as the virtual prototype simulation.The experimental data and simulation results are basically consistent,providing solution support for the engineering application of quadruped inspection robot.