Research On Simulation Localization And Map Building Algorithm For Demolition Robot

Demolition robot are widely used with the transformation of the city and the development of industry.Currently,the way to control demolition robot is mainly by manual remote without the ability to move independently,this cause the efficiency and accuracy of the operation are limited.To move independently Demolition robot requires the sensor information for spatial self-localization and environmental perception,this process is called simulation location and map building(SLAM).This technology involves sensor information processing and several mathematical models is the prerequisite of demolition robot wove independently.This paper first discuss the relationship between localization and mapping from the general problem in SLAM;Establish probabilistic model for SLAM;The realization of SLAM algorithm which based on the extended Kalman filter and particle filter algorithm is being discussed;Both SLAM algorithms are simulated and compared in the MATLAB,then choice the better performance of the algorithm as the theoretical basis of demolition robot SLAM.Secondly,Focus on the localization problem of the demolition robot SLAM:the problems and difficulties in the localization of the demolition robot SLAM are being discussed.The localization model which involved in two kinds of localization methods are analyzed and modeled.Besides the matching localization of the demolition robot is verified in the MATLAB.Based on the multi-sensor information fusion technology,the localization and matching of the odometry is matched with the localization of the scanning to solve the problem of localization accuracy which caused by the demolition robot track slip.On the basis of the theory we discussed above,the Fast SLAM algorithm improved by localization algorithm can be used as SLAM algorithm for the demolition robot.Finally,with the improved SLAM algorithm to ROS operating system open source package,establish a simulation environment in Gazebo simulator,comparison of odometer localization and odometry model fusion algorithm of map building.The hardware platform and software platform are built on the basis of the demolition robot platform,the improved algorithm is tested in the laboratory with high precision laboratory map.Finished the research on demolition robot slam algorithm.