关键词: 多智能体系统, 启发式搜索, 混合交叉策略, 凸变异

Abstract: By integrated with the local apperceive ability of multi-agent system and the strong search ability of genetic algorithm，a Hybrid Multi-Agent Genetic Algorithm（HMAGA） is proposed.It constructs heuristic search and a hybrid crossover strategy to complete the competition and cooperation of agents，a convex mutation operator and some local search to achieve the self-learning characteristic.Some different type of testing functions proved the effectiveness of this algorithm.The experimental results show that HMAGA has a good performance on numerical optimization，especially for the composition functions.

Key words: multi-agent system, heuristic search, hybrid crossover strategy, convex mutation