Laser Range Finder-based Mobile Robot Simultaneous Localization And Map To Create,

With the continuous development of intelligent mobile robotics, the mobile robots which can independently accomplish task will be got more and more expectation. The mobile robots can independently accomplish task, one of the most important conditions is the environment map. The simultaneous localization and mapping of mobile robots under unknown environment will have important significance to the development of intelligent mobile robotics. The thesis mainly used laser scanner to get data information, then achieved the purpose that the simultaneous localization and mapping of mobile robot under unknown environment.The thesis expounded on the background, research significance and major research content, and then discussed the status and research methods of the current international and domestic, and mainly introduced the system hardware platform and the method of choosing sensor and the describing method of environment map.A process of local map building-robot localization-global map building is introduced in the thesis. In order to get the local map, used laser scanner to acquire environment data and used the inner sensor of mobile robot to get the location of robot. Compared the twice data from laser scanner and used extend kalman filter algorithm to get more precise location of the robot. Then updated the local map, finally fused the data of the local map and the current global map, updated the global map and achieved the purpose of creating the global map.An improved "merge-divided- merge" means was adopted in the local map building, using line segments to denote the environment information. The means can be divided into three processes: area divided, line segment distill and line segment fitting. In order to get more accurate local maps, a method named dynamic threshold was introduced in the thesis and used in the process of area divided and line segment distill. The method can filter some noise data and get an accurate local map in a less time. The thesis used the least squares method to get the parameter of line segments.Because the actual environment belonged to the nonlinear system, used the kalman filter algorithm is difficult for finding the strict filter formula. The thesis used extend kalman filter algorithm to realize the localization of mobile robot. The location that obtained by using the robot internal sensor regarded as the observation value of the forecast location, the line segments in the local map regarded as the observation value of the relative line segments in the current global map. So the parameters of using extend kalman filter algorithm are provided. A method called "line segments range" was introduced to reduce the time that find the relative line segments in the local map and the current global map.After the robot location, the accurate local map was obtained. Through compared the relation of line segments in the local map and the current global map, the line segments in the local map can be divided into two classes. One is the line segments had relation with the line segments in the current global map and the other is independent. The local map data can be fused into the current global map adopted different processing way, and then realized the global map building and updating.