Study Of Mobile Robot’s Autonomous Navigation And Obstacle Avoidance System

Autonomous navigation and obstacle avoidance is not only the precondition formobile robot to accomplishing all tasks, but also the hot research of robot technical field.With the development of sensor technology and intelligent control, mobile robot obstacleavoidance has tried many new devices and control algorithm, made a certainachievements, but it also has shortage. In hence, it has profound significance to make afurther research on robot obstacle avoidance.Mobile robots for environmental awareness, achieving environment information tobuilding the environment model by sensors which they carried, according to theenvironment model and obstacle avoidance algorithm to avoid obstacles, which it meetsin the process of moving. In this paper, TJ-1mobile robot experimental platform is usedfor autonomous mobile robot obstacle avoidance studies, the main contents are asfollows:Firstly, this paper analyzes the control system of the mobile robot platform what weused, establish the coordinate system model of the platform and the laser-range findingsensor coordinate system model, and also analyzes the error of the laser-range findingsensor what we used, in order to give the robot obstacle avoidance process provide atheoretical basis for data exchange, this paper also deduces the relationship among thethree coordination systems in mobile robot navigation. To meet the robot’s real-timerequirements of environmental awareness in the process of obstacle avoidance, this paperusing active window to complete perception of the environment.In order to decrease data storage and increase the system’s ability to process data, thispaper using store barrier points which will be influence the robot’s move, instead of storea large original obstetrical data that is in the form of trajectories’ vector direction otherthan in the form of grid, which is an improvement of grid method.Secondly, traditional artificial potential field algorithm is elaborated in the paper. An improved algorithm is proposed to overcome local minima problem of the traditionalartificial potential field algorithm, so that robot can successfully reach the target point,combine with walking along the wall behavioral for processing the typical concaveobstacle. The proposed algorithm is simulation by Matlab software, which varied its correctness.Finally, environmental information extraction methods and obstacle avoidancealgorithm are described in VC6.0software, and successfully transplanted to TJ-1platform; environmental modeling and obstacle avoidance are tested in the platform.Experiments show that the mobile robot in an unknown environment can successfullybypass the obstacles and reach the set target points，which further validated thealgorithm’s feasibility of the paper.