Affiliations: Department of Electronic and Electrical Engineering, University of Strathclyde, Glasgow, UK. E-mail: keith.bell@eee.strath.ac.uk | Department of Computer and Information Sciences, University of Strathclyde, Glasgow, UK. E-mails: {firstname.lastname}@cis.strath.ac.uk
Note: [] Corresponding author: Andrew Coles, Department of Computer and Information Sciences, University of Strathclyde, Glasgow, G1 1XH, UK. E-mail: andrew.coles@cis.strath.ac.uk.
Abstract: The management of power substations is a challenging task, with two opposing criteria: to reduce wear-and-tear on equipment and to ensure the voltage remains within a specified range. At present, a two-stage process is used. Voltage targets are determined for each time of day by electrical engineers using a time-consuming and costly manual process. Then, a reactive control system at the substation is used to satisfy these targets. In this article, we present a novel application of AI planning as part of an intelligent automated system for devising voltage targets. Within the system, both the cost and fault-tolerance implications of voltage target decisions are considered, and hence an efficient and effective set of voltage targets is produced. Using AI planning affords a great deal of flexibility and we show how the system can handle known exogenous events for a given day to reduce forecasted operational costs.
Keywords: Applications of planning, decision support, autonomous control
