Fuzzy EP algorithm and dynamic data structure for optimal capacitor allocation in radial distribution systems

Optimal reactive-power compensation in a radial distribution system requires the determination of the best set of locations for siting capacitors of minimum sizes. The total cost of compensation should be the least and must yield the maximum energy-loss reduction accounting for various load levels. Other controls such as transformer taps, reconfiguration options and existing reactive-power sources must be considered while searching for the optimal solution. Optimal selection of a few best sites from among a large set is a problem of high combinatorial order and difficult to solve using conventional optimisation techniques. Optimal sizing is a problem of a continuous nature. The paper proposes a single dynamic data structure for an evolutionary programming (EP) algorithm that handles the problems of siting and sizing of new shunt capacitors simultaneously while considering transformer taps, existing reactive-power sources and reconfiguration options, accounting for different load levels and time durations. A fuzzy model of the objective function is developed for optimisation in the EP framework. The proposed fuzzy EP method is tested on two cases of a 69-bus radial distribution system.