Research And Implementation Of Local Obstacle Planning Strategies For Robots With Dynamic Obstacles

As robot technology and artificial intelligence develop,mobile robots have been widely used in industrial production,logistics and warehousing,and public services.Au-tonomous mobile robot as a mobile intelligent reflect the most important,route planning is one of the key supporting technologies.Faced with a complex and changing real environ-ment,how to effectively plan a reasonable path to make the robot quickly and safely cross various obstacles,successfully reach the target position and complete the prescribed task is to consider whether the path planning of the mobile robot is scientific and reasonable,avoiding It is an important reference index for whether the barrier performance is superior.At present,the path planning effect of mobile robots on the market in a static working en-vironment is general,and there is room for optimization,and the navigation effect is worse in a complex and dynamic environment,such as being unable to cross obstacles or easily falling into a local minimum.Therefore,this paper focuses on the laboratory's AGV trans-port robot as the specific research object for the complex and changeable indoor uncertain environment.The research and implementation of mobile robot autonomous navigation technology are carried out.By referring to the latest bibliographies in robot research,this paper summarized 3key technologies-map construction,positioning navigation and path planning,introduced some common path planning algorithms both in global and local views,and finally con-structed a mobile robot on differential drive model.The solution in this paper first opti-mized the heuristic function(based on the traditional A~* algorithm)by using the compro-mise value of Euler distance and Manhattan distance as the estimated value,and increased the offset cost of pre-selected node towards the targets when neighborhood nodes share the same cost.Thus robots can calculate the optimal global path with better usage of path points.Secondly,by applying the traditional TEB trajectory planning algorithm on multi-objective optimization,the objective function was optimized? by increasing the con-straints between speed and acceleration,the system constraints were more reasonable? and by introducing TEB parallel trajectories based on topology,applying local topology map,homologous path filtering and optimal path selection,the problem of local minimum was also avoided,which made the local path planning of mobile robots safer and more reli-able.Finally,this thesis proposed a real-time obstacle avoidance strategy combining A~*algorithm and TEB trajectory planning algorithm in complex indoor environment,and a self-recovery strategy for robot when obstacle avoidance strategy fails.The results of this paper were verified in theory and simulated in ROS RVIZ sim-ulation environment and ROS Stage simulation platform.And by testing the laboratory AGV transport robot in complex environment,the robot had shown the ability in reaching the designated location quickly and smoothly,meanwhile avoiding both still and mov-ing targets safely in real-time.Thus the real-time obstacle avoidance strategy was proven effective.