Power Quality Monitor Allocation Based on Singular Value Decomposition and Genetic Algorithm

Abstract

This paper proposed a novel method for allocating power quality meters with the main objective of ensuring a completely observable power distribution system. The method was designed to establish the quantity, location and type of measurement to be performed (voltage and/or current) for a given distribution system by employing genetic algorithms and the principles of state estimation based on the singular value decomposition. Moreover, a limitation in the number of current measuring channels was inserted in the mathematical formulation as an alternative for reducing costs. The method has been validated by running a three-phase harmonic state estimation using the IEEE 34 and 37 bus distribution test feeders. The results demonstrated the effectiveness of the method for designing power quality monitoring systems in distribution grids, ensuring full observability for state estimation purposes.