Research On "Kidnapping" Relocation Of Mobile Robot Based On Linear Features

With the rapid development of smart factories and modern logistics technology,intelligent mobile robots play an increasingly important role in production and life.Efficient and accurate real-time positioning in an indoor structured environment has become an urgent problem to be solved.This paper mainly studies the information fusion of lidar and odometer to achieve efficient,stable and accurate real-time positioning in the indoor structured environment.Solve the problem of relocation of intelligent mobile robots when they are "kidnapped" by external forces or "power off",the main research contents of this article are as follows:(1)Extracting the scanning environment of the lidar to obtain point cloud data,correcting the distortion of the point cloud through incremental odometer information,reducing the error of straight line feature extraction,and the principle is demonstrated and verified by experiments.Complete the construction of the state model of the mobile robot.Consider that the speed of the mobile robot moving indoors is slow.Using the kinematics model can complete the construction of a more accurate state model.(2)Completed the construction and linear storage of the map in the structured environment,and completed the positioning algorithm of the mobile robot through map matching.In this paper,the rapidity of the observed and matched features is considered,and the map storage is required to facilitate the recovery of "kidnapping" relocation.The linear array is used to construct and store the features of the observed map.For the solution of the positioning process,the initial matching between the observed straight line features in the predicted pose and the straight line features in the map is used to obtain the corresponding relationship between the straight line features.An ingenious optimization equation is established to reduce the complexity of the solution,and the obtained pose angle is solved by first solving,then complete the process of solving the position,and give the experimental verification.(3)Use "sliding window" statistics and "sub-map" and "map" matching methods to solve the problem of resolving the attitude and position of the relocation of "kidnapping".In this paper,the special relationship between the observed straight line features and the matched map straight line features is used to construct a pose angle matrix.It is proposed to use the "sliding window" method to obtain the most probable attitude angle value,and use the attitude angle information to construct the relationship matrix between the observed features and the relationship matrix between the sub-map features obtained by segmentation.The matching matrix is used to obtain the matching relationship between the observation straight line and the map straight line,and then an optimization equation is constructed to solve the repositioning pose.(4)Build a simulation platform and an experimental platform for an intelligent mobile robot,and experimentally verify the positioning and "abduction" relocation algorithm of the mobile robot based on the straight line feature in this paper.In the experiment,the above-mentioned straight line extraction algorithm and mobile robot's positioning algorithm are loaded into the robot system,and the corresponding experimental analysis is completed to verify the feasibility of the positioning algorithm and demonstrate the effectiveness of the "kidnapping" relocation algorithm through offline experiments.To sum up,in the face of the problem of accurate positioning of mobile robots in a structured environment,this paper proposes a method of positioning and "abduction" repositioning based on straight line features.By improving the straight line extraction algorithm to obtain more accurate straight line information,complete the positioning of the mobile robot based on the straight line features,and realize the repositioning of the mobile robot under the phenomenon of "abduction" through the data association between the straight line features.This method can achieve stable and reliable positioning of the mobile robot in a structured environment.It will have broad applications in future production and life.