Research On Obstacle Avoiding Method For Multi-robot Based On Wirless Sensor Network

With the fast development of the technology of modern sensor,communication and computer, mobile robot technology has a great increase. Forthe lots of advantages in aspects of information awareness, acquisition and taskallocation, multiple mobile robots technology becomes the hot point of research.Due to the growing complexity and unpredictability of the current node operationenvironment, the cooperation among mobile robots becomes one of the mostsignificant research fields. This paper is applied to the complex environmentwhere is danger or inconvenient for people to complete tasks, considering themobile robot cooperate with other to finish work in unknown area. Here we putforward a obstacle avoiding method based on wireless sensor network for mobilerobots.In this paper, a whole hardware system is complished. The hardware systemmainly includes mobile robots and external sensors. The mobile robots part isdivided into motor driver module, power supply module. The sensors systeminvolves ultrasonic sensors,angle sensors and their interface circuits design.This paper puts forwords a obstacle-avoiding method employed in Z-Stackprotocal stack, as well as the interface monitoring software of the hostcomputer.Software part includes ZigBee network building, Z-stack runningmechanism, external sensors information fusion algorithm, the fuzzy obstacleavoidance algorithm design and pulse width modulation. In addition, the fuzzyalgorithm of obstacle avoidance of robot has been tested, the human-computerinteraction interface upper used for real-time display and control of mobile robottraces has finished.To solve various problems in the process of robots navigation and obstacleavoidance, a multiple sensors fusion system including electronic compass andultrasonic sensor platform is done in this thesis. Based on fuzzy control method,we design a fuzzy obstacle-avoiding algorithm where information input, memberfunctions design, fuzzy rules and final result are all tested and verified well.Results show that the obstacle avoiding algorithm designeed for mobilerobots could meet the design requirements, and provide a reliable platform forfurther research.