314 | 0 | 229 |
下载次数 | 被引频次 | 阅读次数 |
群体智能是当前新一代信息技术发展理论、实践的聚焦点,对其本质及其研究路线进行哲学反思,对于深化认识和推动群体智能发展具有重要意义。群体智能研究依循仿生学路径,推动了仿生理念技术研究和人机协同系统研究两种逻辑演进,表征出基于仿生理念的两个特点,其研究历程呈现出基于约束条件的阶段性演进特征。当前群体智能研究本质上是机器隐喻下对具有非机器特质的智能群体的探究,在自然智能群体所特有的动态性、觉悟性和探索性方面呈现出一定限度。基于中国哲学智慧中的“体用不二”“知行合一”“理势相成”等观点反思群体智能问题求解的反馈机制、自组织机制和学习机制,能够认识其认知的实体与映射、自发与自觉、确定性与不确定性的辩证统一本性,为还原论范式下的复杂性研究提供拓展范式内在性、连续性、普遍性的启示。
Abstract:Swarm intelligence is the focus of the development theory and practice of new information technology generation. Philosophical reflection on its essence and the route of its research is of great significance for deepening the understanding and effectively promoting the development of swarm intelligence. Following the path of bionics, the research of swarm intelligence has promoted two kinds of logical evolution. Its research process are phased evolution characteristics. The current research on swarm intelligence is essentially an exploration of intelligent groups with non-machine characteristics under the machine metaphor, which presents a certain limit in the unique dynamic, enlightened and exploratory nature of natural intelligent groups. Based on the view of “the entity of substance and function”, “the harmony of knowledge and practice” and “the unity of principle and potential” in Chinese philosophical wisdom, the feedback mechanism, self-organization mechanism and learning mechanism of swarm intelligence problems-solving are reconsidered. It can help understand the dialectical unity nature of entity and mapping, spontaneity and consciousness, certainty and uncertainty, and provide enlightenment to expand the immanence, continuity and universality of the complexity research under the reductionist paradigm.
[1] 国务院关于印发新一代人工智能发展规划的通知[J].中华人民共和国国务院公报,2017(22):7-21.
[2] 肖人彬,冯振辉,王甲海.群体智能的概念辨析与研究进展及应用分析[J].南昌工程学院学报,2022,41(1):1-21.
[3] 肖人彬.群集智能特性分析及其对复杂系统研究的意义[J].复杂系统与复杂性科学,2006(3):10-19.
[4] 吴文峻,郑志明,王怀民,等.群体智能及产业集群发展战略研究[J].中国工程科学,2024,26(1):89-100.
[5] Dorigo M,Maniezzo V,Colorni A.Positive Feedback as a Search Strategy[R].Technical Report No.91-016.Milano:Politecnico di Milano,1991.
[6] Li W,Wu W J,Wang H M,et al.Crowd Intelligence in AI 2.0 Era[J].Frontiers of Information Technology & Electronic Engineering,2017,18(1):15-43.
[7] 李成凤,张阳伟,邵俊倩,等.多智能体群集系统分群行为研究进展[J].电光与控制,2022,29(6):62-67+92.
[8] 王玫,朱云龙,何小贤.群体智能研究综述[J].计算机工程,2005(22):204-206.
[9] 张燕,康琦,汪镭,等.群体智能[J].冶金自动化,2005(2):1-4.
[10] 熊十力.新唯识论[M].上海:中华书局,1985.
[11] 白方周,张雷.定性仿真导论[M].合肥:中国科学技术大学出版社,1998:1.
[12] 陈鹏.现代新儒学研究[M].福州:福建人民出版社,2006.
[13] Dorigo M,Maniezzo V,Colorni A.The Ant System:Optimization by a Colony of Cooperating Agents[J].IEEE Transactions on Systems,Man,and Cybernetics,Part B,1996,26(1):29-41.
[14] Ning J,Zhang Q,Zhang C,et al.A Best-Path-Updating Information Guided Ant Colony Optimization Algorithm[J].Information Sciences,2018,433/434:142-162.
[15] 吴光,钱明,董平,等.王阳明全集:传习录·答顾东桥书[M].上海:上海古籍出版社,2011.
[16] Sosa E.Judgement and Agency[M].Melbourne:Oxford University Press,2015.
[17] 吴光,钱明,董平,等.王阳明全集(卷6):文录·答友人问(丙戌)[M].上海:上海古籍出版社,2011:232.
[18] 王夫之.船山全书:尚书引义(卷四)[M].长沙:岳麓书社,2011:335.
[19] 肖人彬,陶振武.群体智能研究进展[J].管理科学学报,2007(3):80-96.
[20] 王夫之.船山全书:宋论[M].长沙:岳麓书社,2011:88.
[21] Sutton R S.Reinforcement Learning Architectures for Animates[C]// Proceedings of the First International Conference on the Simulation of Adaptive Behavior.Cambridge:The MIT Press,1991:105-124.
[22] 郑志明,吕金虎,王亮,等.社会大数据跨尺度系统学习理论与方法[J].中国科学:信息科学,2024,54(9):2083-2097.
[23] 冯埔,吴文峻,罗杰,等.基于群体熵的机器人群体智能汇聚度量[J].智能科学与技术学报,2022,4(1):65-74.
[24] 王夫之,舒士彦.读通鉴论(卷十二)[M].北京:中华书局,2013:1702.
[25] 郑志明,吕金虎,韦卫,等.精准智能理论:面向复杂动态对象的人工智能[J].中国科学:信息科学,2021,51(4):678-690.
[26] Wooldridge M.A Introduction to Multiagent Systems[M].Chichester:John Wiley & Sons,2002:678.
[27] 金岳霖.知识论[M].北京:商务印书馆,2019:174-176.
基本信息:
DOI:10.19484/j.cnki.1000-8934.2025.04.008
中图分类号:B2;TP18
引用信息:
[1]于金龙,贺彩.群体智能系统的研究路线及其反思——基于中国哲学智慧和复杂性思维[J].自然辩证法研究,2025,41(04):12-20.DOI:10.19484/j.cnki.1000-8934.2025.04.008.
基金信息:
国家社会科学基金重大项目“系统哲学思想史”(19ZDA037);国家社会科学基金后期资助项目“科学实践哲学中的规范性问题研究”(24FZXB072)