Research On DHMAS Model And Dynamic Adjustment Mechanism For Complex Systems

With the rapid development of information technology and human society,modern software systems are increasingly demanding in terms of complexity,heterogeneity,and openness.However,traditional software system modeling methods are often incapable of dealing with such complex system modeling problems.Multi-Agent System(MAS)can deal with the complexity problem caused by the change of external environment and internal state during the running of software system.With the characteristics of openness,heterogeneity and evolving,the multi-agent system(MAS)has become a research hotspot in recent years.In a variety of different multi-agent system organization styles,the Holonic Multi-Agent System(HMAS)is one of the most likely methods for modeling complex systems due to its advantages in flexibility and adaptability.The concept of Holonic multi-agent system has been proposed since 1999.Up to now,research on Holonic multi-agent system has made great achievements.Especially in intelligent manufacturing system,it has been widely used.However,its research in complex software system modeling is still not wide enough and in-depth,and there are insufficient considerations for the operating environment and internal structure changes that may occur during the operation of complex software systems.The current Holonic multi-agent system still can not fully meet the modeling requirements of complex software systems.The industry urgently needs a unified and effective dynamic adjustment mechanism to meet the stability and efficiency of the software system in the running process.Therefore,this paper studies the Holonic multi-agent system and proposes a dynamic Holonic multi-agent system(DHMAS).The main research contents and conclusions of this article are as follows:(1)This paper studies the characteristics of the Holonic multi-agent system organization structure model,designs the Holon model of the system modeling unit with self-similar characteristics,and designs the structure and related parameters of the Holon model.(2)This paper studies the task allocation mechanism based on DHMAS modeling complex software,and designs a task assignment method based on load state and a task network assignment method based on capability for the task assignment process of the system.(3)This paper studies the software adaptive mechanism based on DHMAS,designs and implements the system self-adaptive mechanism,and provides support for the state monitoring,capability adjustment and structure adjustment of the software system during the operation process.(4)Taking the intelligent meta search engine as the experimental object,this paper implements a DHMAS-based intelligent meta search engine,further tests the feasibility of the DHMAS modeling method from a functional perspective,and tests the stability and high efficiency of the DHMAS system from a non-functional perspective.The experimental results show that the dynamic ensemble multi-agent architecture model designed in this paper has good usability and can provide guidance for the modeling of complex systems from both the static organization structure and the dynamic operation mechanism.