Abstract:  Considering the problem of premature convergence in fruit fly optimization algorithm (FOA), this paper proposes a new FOA based on history cognition (FOABHC). The new algorithm FOABHC optimizes the evolutionary equation by the strategy of adding “history cognition”, and by doing so, the potential global optimum can avoid losing the probability of being the ultimate global optimum, which results from the potential global optimum not considering its own historical track, and simply aggregating to the current optimum to make its search trajectory with many twists and turns. The value of “history” to individual learning in the iterative optimization process is adjusted by the linear increment dynamic variation coefficient ω. The ability to break away from the local optimum and find the global optimum is greatly enhanced. The comparative experimental results show that the new algorithm has the advantages of better balance searching abilities between global and local, faster convergence speed and better convergence precision.

Key words: fruit fly optimization algorithm (FOA), swarm intelligence, history cognition, convergence precision, premature convergence

摘要： 针对果蝇优化算法的早熟收敛问题，提出了一种新的基于历史认知的果蝇优化算法。新算法通过增加个体“历史认知”的改进策略，优化进化方程，从而避免潜在全局最优解因为不考虑自己的历史轨迹，仅依靠单纯的聚集行为，而使自己的寻优轨迹迂回曲折，错过成为全局最优解的可能；并且通过线性递增的动态变化系数ω调整在迭代寻优过程中个体的“历史”对本次学习的价值，增强算法跳出局部最优，寻找全局最优的能力。对几种经典测试函数进行了仿真和实例计算，结果表明新算法更好地平衡了全局和局部搜索能力，在收敛速度、收敛可靠性及收敛精度上比其他经典智能优化算法有较大的提高。

关键词: 果蝇优化算法, 群体智能, 历史认知, 收敛精度, 早熟收敛