Research On Motion Planning Method Of Tracked Robot In Unknown Environment

Tracked robots play an important role in the fields of logistics consignment,medical disinfection,etc.The motion planning of the robot,as the core of intelligent navigation,directly affects the efficiency and stability of operation execution.Aiming at the motion planning requirements of tracked robots in the environment of unknown obstacles,this thesis explores the path planning algorithm and trajectory planning algorithm of tracked robots based on the rapidly exploring random tree algorithm(RRT)and the dynamic window approach(DWA),and uses the robot operating system(ROS)as the software environment.The main works of this thesis are summarized as follows:(1)The hardware platform of tracked robot with sensing,decision-making and motion control modules is constructed.The ROS software architecture and navigation software package are introduced.The ROS tracked robot software platform is constructed,which provides a framework for the deployment of hybrid robot path and trajectory planning algorithm and provides a basis for the verification of the algorithm in this thesis.(2)The double-layer RRT*algorithm(DL-RRT*)is proposed to solve the problem of dynamic path planning for unknown obstacles and optimal target bias strategy.Firstly,the initial path is quickly explored through the first layer RRT*,and a sampling strategy with target bias is generated based on the probabilistic feedback information of the existing state tree nodes to reduce the sampling area of the new nodes.In order to avoid falling into the local optimal solution,the global strategy of greedy algorithm is adopted to optimize and eliminate the redundant nodes in the initial path.Then,in the second layer,the shortest path of target bias is found by using the feedback information of adjacent points on the initial path.For the environment of unknown obstacles,DLRRT*regenerates the state node tree for the dynamic target to quickly locate the new shortest target path and avoid collision with unknown obstacles in real time.(3)The extended dynamic window approach(EDWA)is proposed to solve the problem of low stability of tracked robot due to the distance limit between the predicted trajectory and obstacles.The sampling window expansion strategy is used to expand the velocity sampling space and predict the optimal trajectory.Considering the evaluation factors of the current position of the robot and the distance between the target point,the incremental evaluation factor of the target distance is introduced to constrain the speed and position of the robot,optimize the choice of the trajectory,and improve the smoothness of the planned trajectory and the smoothness of the robot’s running.(4)The double-layer RRT*algorithm and the extended dynamic window approach proposed in this thesis are applied to the ROS-based tracked robot platform to realize the hybrid planning navigation with path planning and trajectory planning.Through simulation and practical application experiments,the effectiveness of the two algorithms in the environment of unknown obstacles is verified.