Temporal reasoning methodologies used in AI applications for power system control centers

The results already obtained with intelligent real time applications in power systems show that artificial intelligence techniques are adequate and useful for solving some problems for which traditional programming techniques are not able to provide good solutions. On the other hand, the duration of this kind of project is still rather long and the total number of deployed systems all over the world is still very low. This is due to the complex issues that real time applications for power systems must deal with, among which temporal reasoning plays a key role. This paper presents the most important temporal reasoning requirements for real-time applications in power systems. It also presents the methods used in some systems that have been developed to deal with temporal reasoning, with particular emphasis on SPARSE, an expert system for fault analysis and service restoration, developed for Portuguese Control Centers