On the Conception of Intelligent Power Plants Based on Multiple Agent Systems

Lately, great efforts have been made to develop effective hybrid power systems, which consist of a mixture of renewable and conventional power plants, energy storage systems and power consumers. The very dissimilar characteristics of these elements, as well as the ever increasing performance requirements imposed to them, makes the design of control systems for power generation plants a remarkably challenging task. A promising approach to provide effective solutions to this problem is by applying the paradigms of Intelligent Agents and Multi-Agent Systems. In this paper, the definition of an Intelligent Multi-Agent System for Supervision and Control (iMASSC) is proposed to create intelligent power plants for either renewable or conventional power generation units. A Multi-Agent System with a generic structure is used instead of a single specific Intelligent Agent. This approach is more realistic in that it takes into account the complexity of current power plants. Later, the community of intelligent power plants, through autonomous and coherent collaboration, will achieve the objectives of the hybrid power system. Hence, the iMASSC model is expected to provide feasible solutions to the operation of modern intelligent hybrid power systems and smart grids.