Resumo:
In the technical literature, different mathematical models are presented for modeling transmission lines (LT) that can be applied according to the nature of the LT and the electrical study carried out. In electrical studies, LT parameters are generally considered constant and possible significant variations over time may be neglected. Additionally, in the mathematical modeling of LTs, construction data is used that involves errors due to factors such as inaccuracies in soil resistance, considerations of the average value of the height of conductors in relation to the ground, among others that can lead to inaccuracies in the applied LT model.
Additionally, with the advent of new technologies, including PMU, it is possible to obtain time-synchronized phasors of voltage and current at points geographically distant from each other, enabling the evaluation of new methods, or improvements to methods, for determining TL parameters. In this sense, this work aims to propose a method capable of optimizing the estimated parameters of a TL over time. To carry out this activity, in addition to synchronized voltage and current data, mathematical models of TLs and optimization through genetic algorithms will be applied.
The proposed optimizer will aim to identify the concentrated parameters of a TL that best approximate its real model. The estimator's performance will be evaluated through computational simulations in ATP and MATLAB, as well as through real measurement data from a .
In order to seek validation of the proposed estimator, an application of the proposed solution will be presented in a fault location algorithm for real LT events, seeking to find better results for the calculated fault location values.
