A Fuzzy-Logic Based Chaos GA for Cooperative Foraging of Multi-Robots in Unknown Environments

This paper investigates the foraging of multiple robots in completely unknown environments. The onboard robot sensor information and expert knowledge of foraging are used to forage the targets. The foraging problem in this paper is defined as a searching task, where the robots cooperate to find and reach all the targets in an efficient way. A novel fuzzy-logic based chaos genetic algorithm (FCGA) is proposed for target foraging in unknown environments. The fuzzy logic is used to avoid the disorder of the robot movement and reduce the search time when there is no information about the targets or the information density around the robots is the same. The chaos genetic algorithm enables the robots find the targets efficiently. In the proposed approach, the robot motion can be dynamically adjusted to guarantee that all the targets can be found, even in some difficult situations such as targets are at some locations difficult to find or obstacles are linked together. The proposed approach is capable of dealing with uncertainties, e.g., some robots break down. In comparison to the pure chaos genetic algorithm (PCGA) and the random-search approach, experimental results show that the proposed approach is more efficient in foraging all the targets.