The Research Of Cooperate Search Method Of Swarm Robots Based On The Foraging Behavior

With the development of robotics, swarm robots get more and more attentionbecause of their small size, quantity, simple structure, flexible and robust, easier tocontrol than a single robot, and they have strong fault tolerance, even if a robot isbroken, it will not affect the overall search and other features. Simultaneously withthe deepening of biological research, the researchers found that some animals canquickly find food through simple cooperation. Inspired by these creatures foragingbehavior, this paper proposes to simulate social foraging processes by swarm robots,namely in an unknown environment, swarm robots using the features of socialforaging behavior to search targets.This paper on the basis of the key technology in collaborative search by swarmrobot use grid method for modeling the environment. In order to improve searchefficiency, we proposed zoning policies, the entire search area is divided into severalsub-regional to search. We using the local communication technology to communicatedue to it. In order to reduce the burden of communication and storage, we using twomemory storages, one is permanent memory, one is erasable memory.In order to find the target as soon as possible, to reduce the residence time inthe search area, on the basis of the Hornets threshold model, we propose a regionalutility function. Through theoretical analysis and experimental simulation can be seen,robots use the utility regional function to select each sub-region for searching canmake the search more flexible, reducing the target trapped in time, while in thesimulation process, there are some drawbacks in using the regional utility functions,so we set some rules,proposed an improved regional utility function.Inspired by biological foraging behavior, we studied the hybrid searchalgorithm based on the foraging behavior of organisms. This hybrid algorithm iscomposed by a random search algorithm and particle swarm dynamic In case thetarget is not found, the search algorithm using the random search, and search efficiency was found in the target, using a dynamic search PSO accelerate time todetermine the target, while avoiding the phenomenon of missing the target, thesimulation results of our You can see that the proposed hybrid algorithm forsingle-target and multi-target is suitable, but also to ensure that the proposed hybridalgorithm in the search process is not lost goals. While in this paper we also studiedthe behavior transformation strategy that based on the hybrid algorithm, robots haverandom search behavior, organizational behavior, collaborative behavior, they use thehybrid method for conversion behavior in the search process.Finally, by comparing the experimental validation of the proposed group basedon the foraging behavior of the robot search methods in determining the targettime,finding the target number,the total time of completing searching and otheraspects superior to other methods, can achieve good search results.