Secondary Localization of Mobile Robot Based on Improved Least Square Method
In order to solve the problem of floating map and large positioning error of mobile robot in com-plex scene,a quadratic positioning algorithm of robot based on reflector is proposed.By fitting point cloud data,combining multi-frame fusion,point cloud contour filtering and line fitting technology,LM algorithm is used to effectively deal with redundant parameters based on least square method,and reduce the probabil-ity of cost function falling into local minimum value.The relative pose relationship between the mobile ro-bot and the reflector is obtained and high precision docking positioning is realized.A specific reflector with high reflection intensity was installed under the docking platform.With a latent mobile robot as the experi-mental carrier,the robot was set to cycle the fixed route 50 times under two different ground materials.The two-dimensional code coordinate system was calculated and the secondary positioning of the robot was real-ized based on the recognition of the reflector.According to the test results of the prototype,the maximum lateral deviation without reflector is+4.55 mm,the maximum longitudinal deviation is+4.91 mm,and the Angle deviation is 1.89°～-2.08°.When the reflector positioning is added,the transverse position de-viation value is 0.3 mm,the longitudinal deviation value is 0.5 mm,and the Angle deviation value is 0.2°.The improved algorithm has high fitting accuracy and can realize the accurate matching of the reflector.
万方数据
维普
