Research On Motion Planning Approach Of Mobile Robot In Unstructured Environment

In recent years,mobile robots have been widely used in industrial,commercial and human living environments.Mobile robots can complete dangerous tasks such as nuclear power plant inspections and post-earthquake rescues,as well as people’s livelihood services such as logistics distribution and guided tours.In order to realize that mobile robots can assist or even replace human work in some fields,the problem of motion planning of mobile robots in unstructured environments has always been an urgent challenge to be overcome in the field of robotics.In this paper,the problem of target point navigation motion planning in known environments and autonomous exploration motion planning in unknown environments are studied respectively.The main research contents are as follows:(1)Design of mobile robot system.This paper firstly designs and establishes the mobile robot system.According to the application requirements,the motion form of the mobile robot is determined,the kinematic model of the mobile robot is established,the URDF model is built based on the ROS system,the main functions of each sensor module are realized,the real robot platform is built,and the mobile robot system framework is proposed.(2)Research on target point navigation motion planning method in known unstructured environment.Aiming at the problem that the motion state of real-time obstacle avoidance planning is unstable when a mobile robot moves in a known unstructured environment,a hybrid motion planning method(TEB-PF)based on TEB and APF is proposed,which can realize path planning,speed planning and real-time obstacle avoidance.The virtual potential energy of the APF superimposed potential field and the virtual kinetic energy conversion function of the robot is establish,plan the real-time speed of the mobile robot,and improve the motion comfort;The optimized TEB approach plans the local optimal path and solves the problems related to the local minimum region and non-reachable targets;The safety area of the dynamic obstacle is constructed to realize turning or emergency stop obstacle avoidance,thereby effectively ensuring the safety of the car-like robot in emergency situations.The simulation experiments show that the proposed approach has superior kinematic characteristics and satisfactory obstacle avoidance planning effects and can improve the motion comfort and safety of the mobile robot.In the practical test,the mobile robot moves stably in a dynamic scenario,and the proposed approach satisfies the actual application requirements.(3)Research on autonomous exploration motion planning method in unknown unstructured environment.Aiming at the efficiency and safety of mobile robots exploring unknown unstructured environments,a hierarchical framework for autonomous exploration motion planning(HF-AEM)based on boundary point exploration is proposed,which includes frontier point exploration level and exploration motion level.The frontier point exploration level finds the optimal frontier in the global scope and attracts the robot to move to the unknown area;the exploration motion level analyses local environment information in real time,realizes dynamic motion planning and completes obstacle avoidance control.Based on the forward filtering angle and cost function to screen the frontier point to obtain the optimal frontier point.The existence of virtual target points enables the robot to achieve target towing at any moment.Frontier point potential field established to make the robot move towards the unknown area.The environment around the robot is analyzed,a safe obstacle avoidance desired point is obtained for tracking,a scenario-speed conversion mechanism is established,and the target point and obstacle information are weighed to complete the motion speed planning.The setting of the detection range of dynamic obstacles ensures that the robot can avoid obstacles quickly during exploration.Experiments in different simulation scenarios and real environments designed to verify the effectiveness of the method.Results show that our method is more efficient and safer than existing methods.