Parameter estimation of load model using disturbance measurements provoked by taps of transformers

摘要

Load models represent the behaviour of a device or an aggregation of devices, and can be modelled as an induction motor (IM) and an impedance (Z), that could represent, respectively, dynamic and static loads. The article's goal is to correctly estimate the model's parameters, impacting on simulations of voltage and transitory stability. The problem in estimating load model's parameters is the need of measurements taken during disturbances (short circuit, faults, etc.), which are not frequent. For this paper, the measurements were obtained during a tap change, which is more frequent and makes the application easier. The model's outputs and real measurements are compared and the parameters are adjusted until the behaviours are close enough. The estimation process was applied to an initial set of parameters, converging in 5 iterations. The whole process took 2 seconds on a PC Core i7. The algorithm, created in Python, was efficient identifying parameters of a Z-IM load model, resulting in an output's behaviour close to the observed in real systems.