Intelligent Planning and Execution of Tasks Using Hybrid Agents

Multi agent systems are being applied to a wide range of applications from territorial explorations to unmanned combat missions. Agents operate as distributed problem solvers and work together to accomplish the tasks in the system. Various control frameworks have been developed for execution of tasks by agents in the system. Control frameworks consider that the agents are aware of their tasks and provide for the controls to complete the tasks in the dynamic, changing and uncertain environment of the system. But tasks in multi agent systems are not always predetermined and may evolve over time requiring dynamic planning for not only the controls but also the determination of tasks for the agents. This paper presents an integrated task planning and control framework for multi agent systems. A dynamic planning framework is developed which in turn is integrated with a selected hybrid control framework for multi agent systems. Simulation result are presented for the effectiveness of the proposed work towards a comprehensive solution to the multi agent systems based applications.