Multi agent systems : an example of power system dynamic reconfiguration

The energy management in complex power electronic systems especially in case of limited power availability is challenging. In these systems generators, loads, energy storage devices and electrical power routers must be dynamically controlled and reconfigured based on need and priority in order to accommodate the power demand. Furthermore, to achieve optimal use of the available energy, soft continuous reconfiguration is more effective than hard switching reconfiguration. The implementation of this capability in large systems calls for distribution on one hand and integration on the other hand, of control functionalities with monitoring, decision making and other ancillary functions. A multi-agent framework has been proposed in the literature to control these systems. Agents must perform different tasks such as measurement, control, computation, decision making, and simulation, each task being best performed in a specific hardware or software environment. The focus of this work is a heterogeneous agent-based monitoring and control system, while the aim is to allow the use of interactive and proactive capabilities proper of agents. For this purpose dedicated interfaces between the agents and the measurement, control and simulation environments have been implemented. The presented application is the load management in a limited-power, power electronic system