Research On Track Re-planning Technology In Autonomous Navigation Process In Semi-structured Environment

Autonomous mobile robot is an important branch of intelligent robot,and path planning technology is a key research hotspot in the field of mobile robot technology.At present,there are plenty of research results on navigation planning in a structured environment.Because lots of classical path planning algorithms have been proposed by scholars.However,in the semi-structured environment where local unknown obstacles exist,there are still many problems in the path planning of robots.In this thesis,a hybrid path planning algorithm based on improved A*algorithm and adaptive dynamic window method is proposed for the semi-structured environment.The principle and implementation flow of the global path planning algorithm A*algorithm was analyzed,and the cost of the grid map required for its global planning was improved.The traditional Boolean grid map was improved into a safety cost map according to the distance between the obstacle and the robot.At the same time,according to the road state energy consumption state on the map,the energy cost map was also improved.And finally,the two maps were merged to form a final cost map,the evaluation function of the A*algorithm was modified to increase the amount of information considered in path planning.Aiming at solving the problem that the traditional A*algorithm generates too many inflection points and thus with poor path smoothness,the weighting idea was introduced.The number of inflection points of the A*algorithm was reduced,which was verified by simulation experiments.The dynamic window method in the local path planning algorithm was investigated.A parameter adaptive method was proposed for the problem that the path speed and security cannot be considered.According to the distance between the robot and the obstacle and the target point,the sampling window size and the azimuth score weight ratio were automatically adjusted,though which the robot obtains higher speed and shorter path in the planning process.The algorithm principle was verified by simulation experiments.Finally,the improved two algorithms were combined and processed,and the implementation flow of the hybrid path planning algorithm was given.The robot operating system ROS and its three-dimensional physical simulation plug-in Gazebo were introduced.The hybrid path planning algorithm was implemented in the navigation system framework provided by ROS and verified in the simulation environment Gazebo.It can be concluded from the experimental results show that the hybrid path planning algorithm can be used in path planning and safe obstacle avoidance in a semi-structured environment with temporary obstacles.