Research On Design And Motion Control Simulation Of Quadruped Robot

Quadruped robot is currently becoming a major mainstream research direction for foot-type mobile robots.This article is mainly aimed at solving the two problems in the current research of quadruped robots,namely the position and pose control problem and the trajectory tracking optimization problem during the motion of the quadruped robot.Using the methods of virtual model control and model predictive control,relevant researches on dynamic modeling and control problems were carried out,mainly including the following aspects:Based on solving the problems faced in the current research of quadruped robots,the latest research results and application prospects of quadruped robots are fully investigated.After analyzing and summarizing the advantages and disadvantages of existing solutions,the starting point and necessity of the quadruped robot configuration design scheme and its control algorithm research are clarified.Based on the reference to the mainstream configuration of quadruped robots at home and abroad,combined with the requirements of the subject and the actual situation,a full-elbow quadruped robot configuration is designed,using 12 joint servo motors and 18 degrees of freedom(including 6 body freedoms degrees and 12 joint degrees of freedom).The main structure is made of high-strength aluminum alloy and carbon fiber material,using multiple visual cameras installed on the top and front of the body to build a vision-based motion measurement system for the robot.A six-dimensional force sensor is installed on the sole of the foot to receive the force of the sole of the ground signal.According to the characteristics of the four-legged robot dynamics model,a simplified principle of the four-legged robot dynamics model is proposed to decouple the four-legged robot base body and the leg dynamics model,and the kinematics of the base body and the leg mechanism are derived respectively.And dynamics models,as well as the contact force model of the four-footed robot's sole and the ground.Finally,in the dynamic simulation software MBDyn,a complete virtual prototype model of the quadruped robot is established.Starting from the common gait types of quadruped robots,this paper focuses on analyzing the characteristics of two typical gaits of crawling Crawl and diagonal trot.Then combined with related methods to calculate the expected foot drop point of the swing foot,and select the classic cycloid function to plan the swing path of the foot end.Finally,based on the foot-end ground contact signal to design and plan the leg movement timing,the crawling and diagonal jogging gait logic timing and judgment process are respectively planned,which lays the foundation for the related simulation and experimental research in several other research contents of this paper.Aiming at the pose control of quadruped robot,an improved virtual model control method of fusion pose is proposed.By constructing virtual components(springs and dampers)on the base of quadruped robot,the virtual model control of quadruped robot is obtained.Rate,and input the base speed as desired to the leg mechanism controller.For the support phase and the swing phase,impedance control and virtual model control are used,combined with the derived leg dynamics model,and the calculated moments of each leg joint are calculated.Simulated in the dynamics simulation software MBDyn,the concept of motion efficiency of the quadruped robot was proposed,a variety of simulation conditions were designed,the effectiveness of the designed controller was tested,and the motion of the quadruped robot on different road characteristics was tested.The effect of efficiency.Considering the complex movement route and multi-constraint situation of the quadruped robot in the actual environment,combined with the characteristics of multiconstraint and online optimization of model predictive control,a trajectory tracking optimization controller for quadruped robot based on model predictive control is designed,including The setting of discrete linearization error state space expression,objective function and motion constraints of the kinematics model of the foot robot.Different simulation conditions were set,and the optimal tracking control performance of the trajectory tracking controller was tested under different speeds and different control parameter settings.