A Mathematical Programming Approach for Optimization of Distribution Networks with Photovoltaic Sources

Resumo

This paper presents a mathematical formulation for minimizing the active losses of the electric power
distribution network with photovoltaic (PV) generation sources. The modeling considers the distribution network
reconfiguration (DNR) and the adjustment of the photovoltaic inverter power factor (fppv). Depending on the used
technology, the PV inverters can inject active and reactive power by adjusting its fppv, controlling the voltage
profile, and minimizing network losses. Another way to minimize losses and improve the voltage profile is by
performing the DNR. The DNR is performed by opening/closing switches to form a new network structure,
reducing power losses while satisfying operation constraints, and improves the voltage profile. This mathematical
formulation involves continuous and discrete variables, constraints, nonlinear, and non-convex objective
functions. To solve this problem, the use of a mathematical programming approach was proposed. The problem
aims to find the network configuration and the fppv of the PV inverters adjustment that minimize the losses of
active power and improve the voltage profile. The tests were performed in the 33-bus modified system using
AMPL software and Knitro solver. The results showed that the methodology is effective and the combination of
PV inverters control and DNR can reduce network losses by more than 85% in some situations of load and
generation.