Research On Autonomous Navigation And Obstacle Avoidance Technology Of Indoor Omnidirectional Mobile Robo

Autonomous navigation technology is a key technology for mobile robot to realize autonomy and intelligence.With the advancement of artificial intelligence and sensor technology,the autonomous navigation technology of mobile robots has made great progress.Aiming to solve the problem of poor flexibility and adaptability of the mobile navigation robot in the complex indoor environments,this thesis designs an omni-directional mobile robot based on mcnamu wheel to further improve the flexibility of mobile robot in narrow indoor spaces.The main contributions of this thesis as organized as follows:(1)We designed a novel mobile robot system framework based on the mcnamu wheel to model the Omni-directional mobile robot autonomous navigation system and established the corresponding kinematics model,track estimation model,and lidar model of mcnamu wheel robot.(2)Relying on SLAM technology for environmental perception,we completed the positioning of the mobile robot and the construction of the map.To address the puzzle of low particle diversity and low mapping accuracy that exists in traditional RBPF algorithms,we further proposed a method to modify the suggestion distribution and adaptive resampling to improve the accuracy of simultaneous positioning and mapping in the complex indoor environments.(3)We studied the global planning algorithms in static environments and the local planning algorithms in dynamic environments.To address the problem of the ant colony algorithm having a slow convergence speed and easy falling into the local optimal solution,we proposed an adaptive volatile factor and further improved the heuristic function to optimize the global planning ability of the mobile robot under static obstacles.Meanwhile,the artificial potential field method is introduced into the dynamic obstacle environment,and the virtual traction force is introduced to improve the problems of unreachable targets and the local minimum in the traditional artificial potential field method.The autonomous navigation and dynamic obstacle avoidance of the omnidirectional mobile robot are accordingly realized.(4)Based on the ROS navigation framework,an omnidirectional mobile robot experiment platform was built to verify the proposed SLAM algorithm and autonomous navigation algorithm.The omni-directional mobile robot autonomous navigation and obstacle avoidance experiments were carried out in three different indoor environments to verify the effectiveness of the proposed methods.