Electrical substation service-area estimation using Cellular Automata: an initial report

Abstract

The service areas for electric power substations can be estimated using a Cellular Automata (CA) model. The CA model is a discrete, iterative process whereby substation cells 'acquire' service area by claiming neighboring cells. The service area expands from a substation until a neighboring substation service area is met or the substation's total capacity or other constraints are reached. The CA-model output is dependent on the rule set that defines cell interactions. The rule set is based on a hierarchy of quantitative metrics that represent real-world factors such as land use and population density. Together, the metrics determine the rate of cell acquisition and the upper bound for service area size. These are emergent properties that result from the rule set. Assessing the CA-model accuracy requires comparisons to actual service areas. These actual service areas can be extracted from distribution maps. Quantitative assessment of the CA-model accuracy can be accomplished by a number of methods. Some are as simple as finding the percentage of cells predicted correctly, while others assess a penalty based on the distance from an incorrectly predicted cell to its correct service area. Estimation of the service areas of substations has growing importance from the deregulation of the electric-power industry. This is an initial report of a work in progress.