Novel Dynamic Shaping Artificial Intelligence- Based Fault Location Determination Using Unsynchronized Measurements of Power Transmission Lines

Résumé

Abstract Accurately locating faults that occur in transmission lines is a very urgent process for reliable guidance of the power grid system. This process acquires synchronized measurements of voltage and current at the instance of fault, and this is more complex in practical application. For this reason, this work produces a fault location (FL) estimation smart and high-quality approach for a 138-mile transmission line (TL) using machine learning techniques depending on unsynchronized data. The line-to-ground fault cases are considered an example of the proposed algorithm; however, it can be generalized for the applications of the other fault types. The work is divided into two main axes, the first axis is to utilize unsynchronized measurements for the candidate work by shifting voltage and current angles, while the other axis is to utilize a smart novel feed-forward artificial neural network (FFN) of changeable configuration adapted depending upon the nature of the studied problem. For a fair evaluation of the work, a comparison process with (FFN) based on the k-fold cross-validation technique and random forest (RF) methods is implemented. Whereas the proposed method has proven to be more highly efficient and accurate in predicting the location of the fault than others.

Citer ce document

Accès au document

Texte intégral en lecture en ligne, réservé aux abonnés SPHAERO et aux membres de l'institution. Se connecter

Auteur(s)

Statistiques

Consultations : 1

Téléchargements : 0