| Robot localization problem in intelligent behavior of mobile robot has a very important significance, it is the key problem to achieve autonomous navigation, to achieve the complex intelligence mission in unknown environment, it also embodies the robot's perception and intelligence. In this paper, the family environment act as the research background of a mobile service robot , it stores a lot of things in the home environment and places disorder, it is difficult for service robot to achieve self-localization quickly and the establishment of environmental map, the combination of smart space and the mobile robot technology can solve this problem.Firstly, we analyse the intelligent space technology and the Research status of SLAM and the typical methods, we propose the use of intelligent space technology solution based on the general model and the technical difficulties of SLAM algorithm in the current, We give the implement of intelligence services under the smart space technology, and give the realization SLAM ideas and details the realization of the technical difficulties in the process and the key issues.Secondly, in order to enhance under conditions of uncertainty in the positioning of the anti-jamming performance of mobile robot, the robot was kidnapped while addressing the problem, improved Monte Carlo method, the camera with the combination of the global smart space. The algorithm also uses a motion model and observation model as the importance function from which to sample, to avoid the sample set does not include the real situation of the position and orientation sampling, can effectively solve the global localization and the abduction issue. Combined with the smart space Monte Carlo method to improve the efficiency of the robot localization.Then, in view of the image information obtained by visual sensor as traditional sonar, laser range finder information obtained from the direct, three or four chapters give a visual point, line feature extraction method. Access point for the characteristics of the global smart space camera observations and the combination of the traditional SLAM algorithm is given mobile robot model, the global camera observation model, the robot walking control strategy and the noise parameters of the selected implementation. Characteristics of the points in the visual process, there are some limitations, the paper further shows the algorithm based on line features. In this paper, Pineer3-DX mobile robot platform for research, experimental verification of the feature points based on point algorithm and line features of the algorithm, experimental results show that the algorithm based on line features in a structured environment to achieve the convergence faster. |
PDF Full Text Request