Research On Quadruped Robot Motion Planning In Rough Terrain

With the increasing complexity of application scenarios and the continuous improvement of functional requirements,mobile robots often have to carry out transportation and survey tasks in rough terrain.Legged robotics has potential advantages in terms of mobility and versability as compared to wheeled vehicles.However,there is still a big difficulty on its autonomous motion planning.In order to solve this problem,a high-performance quadruped robot is designed,and the autonomous motion planning method is studied and verified by the experiments.This thesis analyzes the autonomous motion requirements of the quadruped robot in rough terrain,and proposes the motheod of mapping in the complex dynamic environment.Pointcloud achieved by the camera is used for mapping,according to the location and height.Based on the height of the terrain,the characteristic information such as gradient and roughness is calculated.Compared with the 2D navigation map,the elevation map has more information about terrain characteristic and provides visual information for the path planning,optimal foothold selection and swing leg trajectory planning of quadruped robot.In order to realize the autonomous motion of the quadruped robot,a good method is proposed based on the terrain characteristic and RRT* algorithm.RRT*algorithm randomly samples in the robot state space and evaluates the traversability of the robot.Finally,the bast path is obtained.This path planning method is completely different from that of the wheeled robot,and it is suitable for complex dynamic environment.According to different types of road conditions,different decision-making behaviors are selected,such as obstacle avoidance,obstacle crossing,stair climbing,etc.After obtaining the body path based on the above method,the motion planning of the quadruped robot body is further carried out,including the trajectory planning of the body posture,and the trajectory planning of the swing leg.In order to highlight the excellent motion ability of quadruped robot,trajectory optimization method is used to plan the body motion trajectory.The body motion trajectory and swing leg trajectory are taken as the optimization objectives,and the kinematics,dynamics model and foot contact model of the quadruped robot are taken as constraints.The continuous time optimization problem is constructed and solved based on interior point method.The kinematics and dynamics models of the quadruped robot are considered in the planning method,and the principle of foot force friction cone is used as the stability criterion,so the planning result is more systematic,accurate and stable.The design and construction of quadruped robot is completed,and relevant experiments are carried out on this basis.The experimental results show that the algorithm can make the quadruped robot have the ability of autonomous motion in complex dynamic environment,and have the ability of climbing,climbing stairs,walking stone path and so on.