This thesis deals with a Smart Grid implementation in a distribution network with substantial share of distributed generation in the form of photovoltaic power plants. The aim of the thesis is to determine optimal reactive power generation in individual photovoltaic power plants considering different operating conditions in the discussed network, bounds of individual power generation units and loads, bounds for the voltage profile in the network, and losses related with the electric power transmission. The objective function of nonlinear and bounded optimization problem will be power losses, which will be minimized in the optimization procedure. Stochastic search algorithm called Differential Evolution will be applied as the optimization tool. The result of optimization will be that distribution of reactive power generation among individual photovoltaic power plants, which will satisfy all limitations regarding voltage profiles, current bounds of the distribution lines, power generation and consumption, with the minimal power losses in the discussed network. The optimizations will be performed for an existing distribution network that belongs to the electricity distribution company Elektro Maribor, where smart gird implementation is in progress. A discussion of the obtained results will show the validity of the process, although other voltage controls are to completely satisfy the system control needs.

Outgoing