Path Planning For Robot Based On ARM-LINUX Platform

Since the beginning of the21st century, with the rapid development of computer science,information technology, software engineering, and industrial control, the mobile robottechnology which covers these disciplines has been continuously developed. One importantsubject of the mobile robot is path planning which is the most basic and important technology ofthe autonomous mobile robots. Path planning technology includes the perception of the robot onthe environment, multi-sensor fusion, and coordination and cooperation mechanisms. Based onthe ARM-LINUX platform, we study the path planning and obstacle shape recognition for thestatic indoor environment. The obstacle shape recognition system and robot path planningsystem are designed and the efficiency is verified by simulation.The main contribution of the present work includes three parts.1). An embedded operationsystem for robot is developed based on the ARM-LINUX platform. The developed system canmeet the basic requirements for robot path planning and obstacle shape recognition.2). Using theprinciple of obstacle shape recognition, an ultrasonic detection apparatus for transmitting andreceiving ultrasonic echo is designed by a so-called “one-emit-three-receive” scheme and adigital oscilloscope is used to analyze waveform diagram of the echo. Combining the distanceinformation with the amplitude of the echo waveform, the simple shape information of theobstacle can be identified by the three echo-receivers. The efficiency of the designed apparatus isalso verified.3). The local minimum problem in the traditional artificial potential field isexplained by using the principle of artificial potential field path planning and solved byimproving the function of the repulsion potential field. Based on it, we develop the driving andapplication programs for the robot sensing system, the software design ideas of improvedartificial potential field, and the realization by C algorithm. The validity of the improvedartificial potential field method is verified by the simulation experiments on Matlab. We also useobstacle discretization method to perform pilot simulation of a mobile robot based on the map ofour laboratory. An improved scheme for how to recognize obstacles when moving along aplanned path is also proposed.