Research On Key Technology Of Mobile Robot Based On 2D Laser Radar

Mobile robot autonomy and intelligence is an important realization of intelligent logistics technology.The precise self-localization and environmental identification of mobile robots are the basis of their autonomy and intelligence,Therefore,path planning and simultaneous localization and mapping(SLAM)are important topics in mobile robot research.In this thesis,used odometry and laser radar as the robot Robot-BX main sensor,around the path planning and simultaneous localization and mapping of two aspects of basic research.First,the path planning method has been studied.For the traditional D*algorithm in the practical application,the inflection point of the global path is too close to the inside of the obstacle inflation area,easy to make the robot into the map inflation area and even hit the obstacles and other issues,used the split-merge method to replan the global path at the inflection point,the original inflection point outside the path of a certain distance,the experiment proved that the improved D*algorithm significantly improved the robot at the inflection point of the autonomy barrier performance.Second,the simultaneous localization and mapping method has been studied.Established the mathematical model of SLAM problem,and analyzed and studied theoretically the SLAM problem solution based on probability estimation method.The simulation experiment of mapping and navigation in unknown environment is carried out on the gmapping algorithm.The influence of the resampling threshold factor selection and the resampling strategy on the positioning accuracy and computational complexity of the gmapping algorithm is studied.The threshold factor k and the algorithm are calculated the complexity of the algorithm is found to be the highest in the vicinity of 0.5,and the computational complexity of the algorithm increases with the increase of the threshold factor k;the system resampling strategy,whether it is in the algorithm localization accuracy or in the computational complexity,are better than the multinomial resampling strategy.Finally,the improved D*algorithm and the traditional D*algorithm are experimented and compared with the mobile robot Robot-BX in the laboratory.The result proved the practicability of the improved D*algorithm.Used gmapping algorithm carring out localization and mapping experiments on Robot-BX platform,the experimental results show that the proposed algorithm is practical,as well as the practical significance of the research content.