Research On Mobile Robot Of Pose Estimation In Structure Environment

Pose estimation made of location estimation and angle estimation is hotspot of mobile robot currently and the key technology achieving autonomous mobile robot localization. The structured environment have obvious characteristics of static objects, such as linear, angle, etc. Thus the structured environment is usually called indoor environment. The data that obtained form 2-D laser range finder is high accuracy and strong anti-interference. The 2-D laser range finder has been widely used in pose estimation of mobile robot. This paper will centre on the pose estimation of mobile robot base on the laser range finder in structured environment, The main work is as follows:The laser range finder made from the SICK named LMS200 is described, the parameters influence on scanning data acquisition is analyzed, that is data rate, resolution angle, location installation, scanning range. The data representation of the polar coordinates in accordance with the polar axis as the horizontal axis, polar radius as the vertical axis is proposed, the median filter window size and the three principles of data segmentation are given.Pose estimation model is introduced, the pose estimation formula in Cartesian coordinates is analyzed. Polar scan matching (PSM) of the core algorithm is described. The formulas about location estimation and angle estimation are given, meanwhile the program code is made. A structured environment in the pose estimation is proposed that is fusion the PSM and Kalman filtering, updated PSM angle estimation using Kalman filtering in certain angle, enhanced the pose estimation robustly.Three experiments are designed, including under the same angle in pose estimation, different rotation angles in the pose estimation, and Kalman filtering fusion PSM in pose estimation. The range residuals in location estimation and the size of the moving windows in angle estimation are determined ,and the results verify the PSM pose estimation effectively. At the same time, through the integration Kalman filter pose estimation with PSM experiments, the variance of observations for Kalman filter and the estimated the value of the angle range are given, the results verify the integration Kalman filter pose estimation with PSM feasibly.