Measurement-based Parameter Estimation for Dynamic Load Modeling

Abstract

It is important to accurately estimate the behavior of the loads for dynamic simulations of power systems. An effective way to represent load behavior is through measurement-based dynamic load modeling. In this study, a measurement-based efficient approach for modeling and identifying dynamic loads has been presented. In simulations of dynamic load modeling, Exponential Recovery Load Model (ERLM) has been used for the parameter estimation of a dynamic load model under various conditions by using voltage and power measurements. The parameter estimation problem of dynamic load modeling has been formulated as an optimization problem and it is solved with the Genetic Algorithm (GA). The parameters of the dynamic load model are estimated by solving the optimization problem whose objective is to minimize the error between the real data obtained from measurements and the estimated data obtained by the proposed models. Utilizing data obtained from the numerical simulations, the proposed model's applicability and accuracy are carefully assessed and tested on the IEEE 9-bus test system under various loading conditions and network topologies.