从群智能优化到群智能进化  

From swarm intelligence optimization to swarm intelligence evolution

作　　者：肖人彬[1] 陈峙臻 XIAO Renbin;CHEN Zhizhen(School of Artificial Intelligence and Automation,Huazhong University of Science and Technology,Wuhan 430074,China;Business School,University of Greenwich,London,SE109LS,UK)

机构地区：[1]华中科技大学人工智能与自动化学院,湖北武汉430074 [2]格林威治大学商学院,英国伦敦SE109LS

出　　处：《南昌工程学院学报》2023年第1期1-10,共10页Journal of Nanchang Institute of Technology

基　　金：科技创新2030-“新一代人工智能”重大项目(2018AAA0101200)。

摘　　要：从认识世界的宏观视野出发,针对问题复杂性与方法有限性存在的鸿沟,按照问题导向的思路,将广泛意义下的解决问题归结为资源有限情况下的资源分配问题(简称分配问题),并将其分为优化型和协调型分配问题两大类,给出了多种优化问题和分配问题之间的关联关系。群智能优化主要是通过对群居生物觅食行为仿生来处理优化型分配问题,群智能进化则是以群居生物劳动分工主导的合作行为仿生来解决协调型分配问题,通过对两者的全方位考察,论述了群智能优化与群智能进化的异同并加以对比说明。聚焦韧性概念时代的到来是从群智能优化到群智能进化的现实推动力量。群智能进化采用协调处理思路,根据问题情景建立定制化的群智能模型,从而有效解决问题,适宜处理一般性分配问题和异常性分配问题,表现出显著的韧性。群智能优化与群智能进化都属于群智能范畴,其中蕴含着若干相互作用的对偶要素,由对偶要素形成的两股力量之间的互动结果决定了群智能类别的归属。按照系统学二象论原理,通过分析竞争与合作、开发与探索、吸引与排斥三对对偶要素,揭示了其与群智能优化和群智能进化类别归属的内在关系。研究表明群智能进化作为群智能研究的新领域,代表了其今后发展的主流方向。Based on the macro perspective of recognizing the world and the problem-oriented thinking,this paper summarizes problem solving in a broad sense as resource allocation problems in the case of limited resources(allocation problem for short),in view of the gap between the complexity of problems and the finiteness of methods.In this paper,the allocation problems are divided into two main categories,namely optimization-based and coordination-based,and the correlation between a variety of optimization and allocation problems is given.Swarm intelligence optimization deals with optimization-based allocation problems by modeling the foraging behavior of gregarious organisms,while swarm intelligence evolution solves coordination-based allocation problems through cooperative behavior dominated by labor division.By comparing the two types of allocation problem in all aspects,the similarities and differences between swarm intelligence optimization and swarm intelligence evolution are discussed and illustrated.The advent of the post-epidemic era is a realistic driving force for the progression from swarm intelligence optimization to swarm intelligence evolution,with a greater emphasis on the concept of resilience.Swarm intelligence evolution adopts the coordinated processing idea and establishes a customized model according to the problem scenario,so as to solve the problem effectively.It is suitable for dealing with general allocation problems and abnormal allocation problems,showing significant resilience.Swarm intelligence optimization and swarm intelligence evolution belong to the category of swarm intelligence,which contains a number of interactive dual elements.The interaction between the two forces formed by dual elements determine the attribution of the swarm intelligence category.According to the principle of systematics interduality theory,the internal relationship between swarm intelligence optimization and swarm intelligence evolution is revealed by analyzing three pairs of dual elements:competition and cooperatio

关 键 词：群智能优化 群智能进化 协调型分配问题 韧性 系统学二象论 问题导向 高质量发展 群智涌现计算 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]