Novel AI approaches in power systems

The electric utility industry has been subjected to rapid deregulatory reforms which have lead to a significant increase in the complexity of the system from the viewpoint of management and control. In this context, traditional artificial intelligence (AI) approaches are no longer as accurate as we would like them to be. Hence, a new set of paradigms are required to take into account the increasing complexity. Firstly, more accurate load forecasting tools are required and we discuss recurrent neural networks (RNNs), hybrid neurofuzzy designs and modular neural networks as the emerging alternatives to the traditional feedforward networks. We then deal with the paradigm of neurofuzzy anticipatory systems, with application to power system control. We also discuss applications of "intelligent agents" to the areas of power system management and power trading in competitive markets.