Research On Indoor Robot Navigation Technology Based On Multi-sensor Fusio

Indoor mobile robot navigation technology has shown great application prospects,and the ability to achieve navigation obstacle avoidance in complex and dynamic environments is an important indicator for evaluating the performance of mobile robots,which is also a major challenge that limits the development of this research.This paper first summarizes the previous studies of indoor mobile robots navigation technology in recent decades,and finds that the types of sensors used by most mobile robots are still relatively single and limited.The mapping and obstacle avoidance algorithms are relatively simple,causing problems such as inaccurate mapping and slow obstacle avoidance in practical applications,which cannot meet the needs of practical application scenarios.In order to improve the obstacle avoidance ability of indoor robots,this study designs an indoor robot navigation system based on multi-sensor fusion,so as to carry out path planning and effectively execute obstacle avoidance navigation.A hardware platform for experiments was built and effectively verify the feasibility and superiority of multi-sensor fusion navigation system.The specific content of this study is as follows:(1)Different hardware modules were selected with the matching operating software system.Then the mobile robot hardware platform was constructed,and the overall design scheme of the indoor robot navigation system was designed.(2)Aiming at the shortcomings of a single sensor which is greatly affected by ambient light and easily drop environmental information,a multi-sensor fusion method based on Bayesian estimation method is proposed.By this method,the robot can construct a grid map of the environment,and improve the accuracy and stability of mapping.(3)Aiming at the problems of long search time and redundant inflection points in the traditional A* algorithm,an improved A* algorithm is proposed,which increases the weight in the direction of the obstacle and shortens the search time by more than half;An additional cost function has also been added,and inflection point optimization has been performed,improving navigation efficiency by about ten percent in smaller environments.(4)Using the built robot platform,the algorithm mapping experiment of multi-sensor fusion and the improved A* algorithm navigation test experiment of multi-sensor fusion are carried out in the simulation environment and indoor environment,which verifies the practicality and superiority of the indoor robot navigation technology proposed in this paper.