Intelligent Multi-Agent Based Information System of Business Process Management

In this paper, a multi-agent based process management system is proposed to modularize, simulate and analyze the existing industrial business process environment dynamically. The different computational agents of the system are presented in a hierarchical architecture, which offers intelligent behavior at different system level. At the bottom level, agents modularize and simulate the typical existing business process environments for its basic data acquisition and pre-processing, which will generate concerned event based on rules embedded inside the system. At the middle layer, agents enable the assessments and reconciliation of above requests through different computation ways, finally return the analysis result to the database management system for top level decision. The top level relies on the statistical analysis reports from the database management system and the rule-based knowledge system to give the final decision and schedule through planning agents to the action subsystem for execution. The central control and coordination subsystem monitors and coordinates all agents inside the system at meta-management-level so as to make the whole system performance more robust and coherent.