Particle Swarm Optimization-based approach for parameterization of power capacitor models fed by harmonic voltages

Abstract

The modeling of power capacitors has great importance on the allocation of such devices in power distribution systems. Many models based on electrical circuits are found in the literature, but their proper parameterization is difficult when these capacitors are fed by harmonic voltages. In this context, the objective of this paper is to propose a methodology to determine, by means of a Particle Swarm Optimization (PSO) algorithm, the parameters of three different capacitor models supplied by voltages with harmonic distortions. In such a methodology, the PSO algorithm determines the parameters of each model by comparing a reference current obtained in laboratory tests with the calculated current flowing through its terminals, considering four different voltage waveforms and two types of metallized polypropylene power capacitors (all-film and impregnated). This approach is intended to be applicable as a prior tool to support capacitor allocation studies in power distribution systems. The results show that the PSO is able to efficient and effectively estimate the unknown parameters of such models. The replicability of results provided by the PSO and the robustness of the proposed methodology are demonstrated by means of statistical analyses. The Mean Squared Error (MSE) between the reference and the calculated currents for ten consecutive executions of the proposed methodology considering each waveform, model and capacitor varied between 3.049 and 0.032A2, with variation coefficients smaller than 1%; to validate the choice of the PSO algorithm, a comparison of MSE values provided by the PSO and four nonlinear programming methods is also shown.