The Research Of A Multi-Mobile-Robots Control System Based On Pheromone Communication

Multi-Mobile-Robots mainly focus on how large number of relatively simple robots communication with themselves and the environment to generate group behavior. It is a new method for large-scale coordinative control of multi-robots system. Wireless Sensor Networks (WSN) is self-organization ad-hoc wireless Networks consisted of a large number of sensor nodes with the ability of apperceiving, computing and communicating. An important method to research Multi-Mobile-Robots collaboration is to mimic biological community collaboration mechanism. The main method for ants to communicate with each other is through pheromone.This paper mainly describes Multi-Mobile-Robots'control, combined with wireless sensor networks, and proposed a new Multi-Mobile-Robots Collaborative Control Algorithm by import the pheromone-based mechanism. The research results can provide a new approach for Swarm intelligence research, which has important theoretical significance.The general works are as follows:First of all, analyse domestic and international research achievements of wireless sensor networks, Multi-Mobile-Robots and pheromone technology. A special mobile robot was developed for the multi-robots system experiment. A IEEE 802.15.4 protocol stack was carried out based on GAINZ hardware platform, and then build a Multi-Mobile-Robots experiment system based on the robots and GAINZ nodes mentioned above.Then, the ant colony algorithm is analyzed. Based on this, a Pheromone-based Multi-Mobile-Robots Collaborative Control Algorithm (PMCA) was put forward. The algorithm defined 4 robot states,6 kinds of information and 4 kinds of robot's actions, given the basic method of the robot state transition, described the pheromone update and the path selection probability model in detail, and given the algorithm implementation process.Finally, demonstrate the PMCA algorithm in the experimental system. Design a Multi-Mobile-Robots robot path optimization experiment. The experiment is divided into three phases:the exploration phase, pheromone accumulation phase, and path optimization phases. Experiment results verified the effectiveness of the proposed algorithm.