Research On Multiple Intelligent Agents’ Swarm-intelligence Self-organizing And Regulation Methods Based On Stigmergy And Neuroendocrine Regulation Mechanism

The rapid development of Internet of Things(Io T),Ubiquitous Sensing Network,Artificial Intelligence and Learning Evolution,Internet and Digital Communications,Big Data and New Computing Technologies has brought about changes in social and economic aspects and driven the emergence and evolution of accurate and agile supply chain,smart logistics and product life cycle-oriented networking and intelligent manufacturing with huge momentum.Besides,integrated production,operation control and business management systems based on social division of labor under contracts also expand day by day,and a lot of dynamically changing,greatly uncertain,highly nonlinear,high-order,high-dimension,high-coupling,heterogeneous,heteromorphic and anisotropic problems also appear,and there’re increasingly complicated needs for division of labor,integrated operation and collaborative work.Nonintelligent traditional methods cannot be adopted to effectively overcome these complicated issues.Yet,intelligent individual and swarm self-organization mechanism and collaboration method based on multi-agent system(MAS)and social biotic communities and human physiological systems help a lot in solving these complicated issues.In this study,intelligent multi-agent system selforganization and modulation method based on stigmergy and neuroendocrine immunomodulation is adopted for experiments and discussions,thus providing a new idea and a new method for intelligent multi-agent system self-organization collaboration solution to multi-collaboration tasks in complicated environments.For this end,by analyzing distributed artificial intelligence,small-world network,Stigmergybased bionic swarm intelligence and neuroendocrine immunomodulation,the importance and functional mechanism of self-organization modulation that imitates human organism and nature to the construction of intelligent systems and the presentation and emergence of multi-agent system-based swarm intelligence is revealed for collaborative solutions of complicated issues.Multi-agent system-based self-organization modulation model and scheme is proposed for multi-tasking collaboration solution.Furthermore,the Stigmergy-driven& MAS-based Method of Modelling Intelligent System(SMMMIS)with human cognition and behavior control mechanism and structure is proposed and built,which is based on Stigmergy’s cooperative consensus initiative and pheromones or hormones generation and spreading mechanism to achieve external interactive cooperation,thus providing a system framework and application paradigm for the construction of intelligent systems at different levels.In order to empower multi-agent systems with different intelligences in an unordered or unbalanced state to autonomously form orderly and balanced dynamic cooperation alliance in response to the changes in different complicated environments,by means of the understanding of selforganization connotations,the discovery of small world network theory and the revelation of information dissemination mechanism based on SMMMIS-based distributed intelligent system,a friend relay algorithm-based multi-agent system swarm-intelligence self-organization mechanism is proposed.Moreover,by comparing distributed/dispersed intelligent system’s traditional self-organization modulation method,social organism system and intelligent system,through recognition of consensus initiative,learning from interactions,joint actions and selfregulation mechanism between human neuroendocrine systems for multi-agent system collaborative solutions to complicated issues,a multi-agent system swarm-intelligence selforganization mechanism and modulation method based on friend cooperation consensus initiative and neuroendocrine immunomodulation is proposed,including DNA double helixbased pheromones or hormones models independently formed by friend consensus initiativedriven intelligent systems and its dissemination and interaction mechanism and affinity calculation method,as well as autocrine and hormone-based multi-agent system consistency control algorithm,laying a solid foundation for the emergence and role of multi-agent systembased swarm intelligence.Based on the above contents,by designing an integrated operation control model of smart warehousing and cashier-free supermarket for retailing in the new era and constructing multi-agent AGV collaborative handing control model,the self-organization method that supports multi-agent collaborative work under smart warehousing/cashier-free supermarket environments is adopted for the design,construction and exploration of simulating experimental systems as well as their result analyses.The simulation results show that the intelligent system modeling method for multi-task collaboration solution under complex environments is verified,and the feasibility and usability of friend relay search algorithm and friend collaboration consensus initiative-based selforganization mechanism and neuroendocrine immunomodulation-based swarm intelligent selforganization modulation and multi-agent consistency modulation algorithm driven by such mechanism are also validated,providing a new model,a new method and a new application paradigm for effective realization of a new generation of road-vehicle collaboration-based intelligent transportation,a new generation of retailing e-business and agile supply chain,smart logistics and warehousing,delivery express,cashier-free supermarket,intelligent manufacturing and other multi-tasking collaboration work under complex environments,which is of huge theoretical guidance and practical reference and application value.