Research On SLAM Method Of Indoor Autonomous Mobile Robot

Autonomous mobile robots are currently a hot research direction.This robot is used in a wide range of fields and can replace humans for some dangerous or single boring tasks.Although there are many advantages,there are also disadvantages such as low positioning accuracy and poor path planning.Combining with the needs of society and the working mode of the robot,an autonomous mobile robot used indoors was designed in this thesis.Main control chip was STM32F103ZET6,the sensor module was odometer and radar,and the driving part was a motor.In this thesis,the model of the robot motion was deduced,the basic theory about SLAM positioning was discussed,the basic model of the system was established,Bayesian filtering was used as the basis for handling the positioning problem,and several filtering algorithms were discussed,including Kalman Filtering,the filtering algorithm and particle swarm filtering algorithm developed on the basis of Kalman Filtering,and corresponding simulation experiments for various aspects of first-order nonlinear systems were given,and then the real environment was simulated,and the experimental simulation of robot positioning and map construction were implemented by using RPBF algorithms.The simulation trajectory graph under the simulation situation was obtained.Environmental information was described by means of raster maps,and common path representations were introduced.Based on the study of the traditional A*algorithm for global path planning,an improved A*algorithm after head-to-tail bidirectional search was proposed and MATLAB simulation verified that the improved algorithm was more effective than the unimproved one.When dealing with the local path planning problem,the dynamic window method was used,while the DWA algorithm was easy to fall into the local dilemma.When dealing with this problem,the walking trajectory was optimized by improving the function,and the improved algorithm was verified by MATLAB simulation effectiveness.The robot was used as the experimental object,and the SLAM experiment was carried out under the static closed condition indoors.The results of simuliation and experiment show that robot positioning and map construction and improved algorithms are helpful.