Optimization of the compensation of a meshed MV network by a modified Genetic Algorithm

This article discusses the utilization of a modified genetic algorithm (GA) for the optimization of the shunt compensation in meshed and radial MV distribution networks. The algorithm looks for minimum costs of the network power losses and minimum capital and operating costs of applied capacitors, all of this under limitations specified by a multicriteria penalization function. The parallel evolution branches in the GA are used for the purpose of the optimization acceleration. The application of this GA has been implemented in Matlab. The evaluation part of the GA implementation is based on the steady-state analysis using a linear one-line diagram model of a power network. The results of steady-state solution are compared with the results from the DIgSILENT Power Factory program. Its practical applicability is demonstrated on examples of 22 kV radial and meshed overhead distribution networks.