Building a Health Care Multi-Agent Simulation Sysmte with Role-Based Modeling

Abstract

Multi-Agent System (MAS) is a suitable programming paradigm for simulating and modeling health care systems and applications, where resources, data, control and services are widely distributed. We have developed a multi-agent software prototype to simulate the activities and roles inside a health care system. The prototype is developed using a framework called Role-based Agent Development Environment (RADE). In this chapter, the authors present an integrated approach for modeling, designing and implementing a multi-agent health care simulation system using RADE. They describe the definition of role classes and agent classes, as well as the automatic agent generation process. The authors illustrate the coordination problem and present a rule-based coordination approach. In the end, they present a runtime scenario of this health care simulation system, which demonstrates that dynamic task allocation can be achieved through the creation of role instances and the mapping from role instances to agents. This scenario also explains how agents coordinate their activities given their local constraints and interdependence among distributed tasks.