一种面向空间分配问题的群智能劳动分工新方法

A new approach to labor division in swarm intelligence for space allocation problem

以装填问题为例,将群智能劳动分工的应用范围由任务分配推广到空间分配,提出一种群智能劳动分工新方法.首先分析了装填问题的空间分配特性,探讨了装填问题的空间分配与群智能劳动分工的任务分配之间的相似性.接着在此基础上引入群智能劳动分工响应阈值模型,根据装填问题的空间分配特点重新设计环境刺激、响应阈值、更新规则等模块,提出了基于响应阈值模型的劳动分工算法(response threshold model based labor division algorithm,RTMLDA).最后,选取装填问题的典型代表--等圆装填问题和不等圆装填问题进行仿真实验,对3组共54个代表性算例的计算结果及与其他算法的比较表明RTMLDA有效可行.

In order to extend the application of swarm intelligence labor division from task allocation to space allocation,this paper takes the packing problem as an example and proposes a new swarm intelligence labor division algorithm.First of all,the space allocation characteristic of the packing problem is analyzed,and the similarity between space allocation in the packing problem and the task allocation in swarm intelligence labor division is discussed.On this basis,the response threshold model of swarm intelligence labor division is introduced.According to the space allocation characteristic of the packing problem,the modules of environmental stimuli,response thresholds and updating rule are redesigned,and the response threshold model-based labor division algorithm(RTMLDA) is presented.Finally,two typical representatives of the packing problem-the equal circle packing problem and the unequal circle packing problem are selected for the experiment.Experimental results on three sets of benchmarks consisting of fifty-four representative instances indicate that RTMLDA is very efficient.