Research And Implementation Of Immune-based Co-Evolution Multi-Agent Model

With the technological advancement and social development, E-government system has involved with increasing number of function departments, intricate operational flows, complex features and massive data storages. Currently, processing workflow in a predifined sequence, E-government system has delayed information communication in workflow knots, shortened speed of systematic respond and devoided ability of intelligent service matching. Known as a complex real-time and dispersed system, E-government system requires the coordination both from service consumer and from service provider to achieve a better performance. During its coordination process, a research concentration has been unprecedentedly focused on the service flow and intelligent service matching. Workflow engine is a traditional choice catering for the demand of E-government flow process. However, first introduced to deal with simple workflow tasks, workflow engine was designed to solve sequenced activities. Even if workflow engine was optimized for customized flow process system, it could only process predefined flows, let alone optimize operational work flows and process urgent situations. It lacks the capability to allow all service participants to coordinate with workflow process or to modify the process and is devoid of the ability of intelligent service matching.Researching related papers, comparing experimental data and integrating agent technology with work flow engine, this paper introduced a practical solution to work flow optimization and intelligent processing in E-government system. Cascading reliable work engine’s stability, combining the flexibility of agent and immune technology and expanding the application area of multi-agent system that uses artificial immune technology, this paper, basing on the above concentrations, proposed an Immune based Co-Evolution Multi-Agent Model. This model not only requires all service participants coordinate with workflow, but also requires E-government system providing coordination mechanism to sync work flow process and to update service records.This paper designed an algorithm to solve travelling person problem, verified the correctness of corresponding algorithm in a K-TSP issue, revealed the model’s flexibility comparing to traditional model and confirmed the model’s practicability in real-world applications. Constructing flat E-government organization work flow and providing practical possibilities, the Immune based Co-Evolution Multi-Agent Model can create direct communications between low and high positions, can erase redundant middle knots, can prompt information transformation and can lower corresponding costs.In conclusion, integrating the model to a real E-government system, this paper exemplified the Immune based Co-Evolution Multi-Agent Model in an e-government system and designed practical solution to a real-world e-government system’s flow processing by using Service Oriented Architecture, Aglet platform and co-evolution approach. According to experiment result and comparing to traditional E-government system’s service query consuming time, predominating conclusions such as service flow optimization and intelligent service matching were drawn from the model.