Artificial neural network-based optimal capacitor switching in a distribution system

One of the most important control decision functions in a modern distribution automation system is volt–var control. The objective of volt–var control is to supply controlled reactive power by switching optimally the switchable capacitors installed in the distribution system such that the voltage drop and real power loss is minimum. Traditionally, this problem of optimal capacitor switching has been solved through various optimization techniques. However, as the time taken by these traditional optimization methods are quite significant, these methods may not be much suitable for online application. To reduce the time required to solve the optimal capacitor switching problem, an artificial neural network (ANN)-based approach has been developed in this paper. It has been found that the ANN-based technique is at least a 100 times faster than the traditional optimization methods for a practical number of capacitors in the system. Moreover, as the number of capacitors in the system increases, the effectiveness of the ANN over the traditional approach (in terms of the solution time) increases.