Distribution network reconfiguration and reactive power compensation using a hybrid Simulated Annealing – Minimum spanning tree algorithm

Abstract

Change of the climate, environmental pollution and lack of classical energy sources forced many countries to address renewables. In modern strategies of smart distribution network reconfiguration should be applied in order to minimize power losses and economic costs. Also, capacitors switching (placement of shunt capacitors) and presence of distribution generation are always incorporated in modern networks. Further, On Load Tap Changer in the supplying substation has to be used. Simulated Annealing and Minimum spanning tree algorithms work separately in order to find the best appropriate solution. The main goals are minimization of power losses and costs. In addition, many constraints are attended to find the hourly reconfiguration and capacitor banks position. Distributed generators involve both wind generators and solar panels. Two methods described in literature are mixed and implemented for locating fixed nodes for distributed generators. Gauss and Weibull changing nodal loads and wind generators output respectively, as well as daily load curves for working and weekend day with insolation of solar panels are included in analysis. Per unit load diagrams are determined according to real one-year measurement results in one substation in Belgrade, Serbia. IEEE 69 bus network is analyzed. The main contribution of exposed research is near optimal solution for wide time range, realistic network operation/planning forecast (stochastic variation of consumption and production on hour basis) with satisfaction of all technical constraints that are minimum and maximum nodal voltages, limiting feeder’s capacities and not-overcompensation of the network, accompanied with elaborate economic background.