Robot Exploration Mission Planning Based on Heterogeneous Interactive Cultural Hybrid Algorithm

Abstract

Interactive bionics-swarm co-evolutionary hybrid algorithm system architecture is presented by using cultural double evolutionary structure in this paper. The architecture includes the upper ceiling knowledge space based on good-point-set genetic algorithm (GGA), the bottom ceiling population space based on discrete particle swarm optimization (DPSO), the top-down influence mechanism and the bottom-up acceptance mechanism, which realize heterogeneous population interaction. Additionally, customer estimation interface is reserved to realize human-computer interaction. In order to improve particle swarm optimization performance, the population space is initialized with good-point-set to distribute the initial particles uniformly in feasible solutions. A novel evolution model is proposed and the particle evolution ability index is defined, which increases the population's diversity and improves the algorithm's stability. A neighborhood local search strategy is introduced to enhance search capability. Finally, the heterogeneous interactive cultural hybrid algorithm (HICHA) is tested with TSPLIB standard data. Experimental results show that HICHA is better than the other algorithms in stability, convergence speed and solution quality. HICHA provides a new way for solving the robot exploration mission planning problem.