Estimation of dielectric response equivalent circuit parameters of transformers using QPSO

Abstract

The initial slope of return voltage is not only an important characteristic quantity to assess insulation aging, but also of great significance to estimate the parameters of equivalent circuit based on linear dielectric response theory. This paper establishes a mathematical model with the data of initial slope for estimating circuit parameters, which is a group of nonlinear equations with exponential function. It is difficult to solve the problem using nonlinear least square optimization method. The most advantage of QPSO based on swarm intelligence is that it does not require any prior knowledge or space limitations and has very good performance on the complex problems. Therefore, this paper presents the QPSO to solve the parameters estimation problem. Based on estimated parameters, we can calculate the RVM curves from the equivalent circuit and compare the calculated data with the measured. The result shows that the established model is feasible and effective. It provides a new idea for assessment of insulation aging and research on new characteristic quantity.