The System Design And Implementation In Autonomous Mobile Robot Based On ROS

Intelligent service robot is a current research hot spots in the field of robotics,and its related research in some electronics manufacturers push gradually into people's daily lives.As an important feature of smart,intelligent autonomous mobile service robots is the key to ensure practical.In this paper,work carried out as follows in autonomous mobile robot design and implementation with ROS(Robot Operating System,an open source robot operating system).First,for the past robot development requires a lot of repetitive code,practicability and poor scalability issues,this paper designed to autonomous mobile robots through introducing ROS.System architecture comprising:A chassis and two differential drive bracket;A system of environment watching about laser radar and kinect camera;A ROS-based intelligent navigation software.Second,in the traditional targeting methods for a single sensor is positioned low accuracy,we use multi-sensor data fusion positioning with Kalman filter and adaptive weighted average combining to improve the measurement accuracy;For the extended Kalman filter anti-jamming capability weak,assuming the noise issue obey a Gaussian probability distribution,using particle filter to improve the positioning accuracy and proposing an improved algorithm Fast SLAM to reduce computational complexity.Experimental results show the effectiveness of the work.Third,in the path planning,the paper had a analysis experiments about global path planning algorithm with Dijkstra and A* algorithm and local path planning algorithm about Trajectory Rollout and DWA algorithm.Taking into account the navigation process may encounter unexpected situations,with adaptive Monte Carlo localization and tracking mobile robot pose to navigate unknown environment,with external force verify emergency situations by experiments,and ultimately verify it feasibility.Fourth,we tested several experiments in the practical application scenarios of this paper,in different situations to test the efficiency of robots to performtasks and their systematic analysis,creating more accurate maps,completing path planning and obstacle avoidance.