Immunity-Based Adaptive Genetic Algorithm for Multi-robot Cooperative Exploration

The key to multi-robot exploration is how to select appropriate targets for robots to avoid collision and overlap. However, the distribution of targets for multiple robots is an NP hard problem. This paper presents a multirobot cooperative exploration strategy based on the immune genetic algorithm. With its random global searching and parallel processing, genetic algorithm is applied for multi-robots multiple targets combinatorial distribution. With its antibody diversity maintaining mechanism, the immune algorithm is used to get over the premature convergence of genetic algorithm. The selection probability is computed based on the similarity vector distance to guarantee the antibody’s diversity. The crossover and mutation probability are adjusted based on the fitness of antibody to decrease the possibility of local optimal. The extensive simulations demonstrate that the immunity-based adaptive genetic algorithm can effectively distribute the targets to multiple robots in various environments. The multiple robots can explore the unknown environment quickly.