论文部分内容阅读
为提高人工蜂群算法的全局搜索能力,提出一种多策略蜂群算法(multi-strategy based artificial bee colony algorithm:m ABC)。该算法设计一种基于最优蜜源的位置计算策略,然后与经典蜜源位置计算策略合作,在引领蜂阶段,以随机进化模式对蜜源进行更新,在跟随蜂阶段,以组合进化模式对蜜源进行更新。同时,设计一种基于差分进化算子的侦察蜂进化模式。对10个经典测试函数和2个数字系统建模问题进行实验仿真,实验结果表明,相比标准蜂群算法,m ABC算法模式有能效地平衡算法的探索和开发能力,提高收敛速度和最优解的精度,具有良好全局搜索效率,是一种有效的求解全局优化问题的方法。
To improve the global searching capability of artificial bee colony algorithm, a multi-strategy based artificial bee colony algorithm (m ABC) is proposed. The algorithm designs a position calculation strategy based on the optimal nectar source, and then cooperates with the classical nectar source position calculation strategy to update the nectar source in a stochastic evolutionary mode while leading the bee staging phase. After following the bee staging phase, the nectar source is updated in a combined evolutionary mode . At the same time, a evolutionary model of reconnaissance bee based on differential evolution operator is designed. Experimental simulation of 10 classic test functions and 2 numerical system modeling problems shows that compared with the standard bee colony algorithm, the m ABC algorithm mode can effectively balance the exploration and development capabilities of the algorithm and improve the convergence speed and the optimal The accuracy of the solution and the good global search efficiency are an effective method to solve the global optimization problem.