Research On The Uncertainty Problem Of Mobile Robot Environment Perception

The research of mobile robots' perception of the environment is an important part of mobile robot technology,especially its application in path planning.Many algorithms have been developed in the research of path planning technology,such as the classic A*,APF and RRT algorithms.When studying the path planning of mobile robots,most algorithms do not consider the uncertain factors of the mobile robot in perceiving the external environment information,especially the uncertain factors in the environment of narrow passages.The sources of uncertainty are mainly: the uncertainty of movement,the uncertainty of the sensor and the uncertainty of the map.The uncertainty in the path planning process largely affects the results of path planning.In order to solve the uncertainty of perception and movement in the narrow passage environment,the paper studies the following:(1)The paper studies the classic path planning algorithms,and systematically analyzes the advantages and disadvantages and simulation performance of the A*,APF and RRT algorithms.According to the characteristics of the algorithm and the environment of the narrow passage,the scheme of using the RRT algorithm to plan the path of the mobile robot is determined.(2)It is proposed to use Partially Observable Markov Decision Process(POMDP)to solve the uncertainty problem in the narrow channel environment.The method in this paper follows the general POMDP solution framework,in which the belief dynamics is represented by an Extended Kalman Filter(EKF),the value function is represented by an effective quadratic function near the trajectory,and the linear quadratic Gaussian method is used in In the belief space,the value is solved iteratively,so as to obtain the local optimal linear control strategy around the trajectory in the belief space.Execute the control strategy to generate a new trajectory,and continuously iterate this process until the best trajectory route is iterated.(3)A partition sampling strategy based on POMDP is proposed.The proposed strategy can effectively solve the sampling problem of mobile robots in a narrow passage environment.First,the environment space is divided into free space and narrow space by partition sampling,and then adaptive sampling is performed in different spaces.Then the local optimal solution of the initial trajectory is calculated through the solution of POMDP,so as to iterate the best route trajectory of the mobile robot movement.In this paper,an experimental simulation is carried out through a narrow passage environment,and the experimental results show the effectiveness of the research content in this paper.It is verified that the method in this paper can improve the path planning efficiency of mobile robots to a certain extent,and solve the problem of uncertainty in narrow passages.