This paper introduces a novel fault location model based on Adaptive Neuro-Fuzzy Inference System (ANFIS). For the purpose of performance improvement, a meta-heuristic algorithm known as Non-Dominated Sorting Genetic Algorithm type 2 (NSGA-II), which has the ability of fast searching for the optimal point and escaping the local optimality trap, has been used for ANFIS training. The fault current and voltage are usually the two parameters used as the inputs to the ANFIS, even though they cannot truly specify the fault characteristics on their own. Here, Discrete Wavelet transform (DWT) has been used to extract the effective and relevant features of the fault current. To demonstrate superiority of the proposed ANFIS-NSGA-II-DWT model, a comparative study will be performed against the traditional Least-Squares and Back-Propagation (LS+BP) and Chaotic Dynamic Weight Particle Swarm Optimization (CDW-PSO) methods. The simulation results show that the proposed estimation model has a superior performance compared to the previous models in terms of both convergence speed and mean square error (MSE).<
Alan : Eğitim Bilimleri; Fen Bilimleri ve Matematik; Sağlık Bilimleri; Sosyal, Beşeri ve İdari Bilimler
Dergi Türü : Uluslararası
Benzer Makaleler | Yazar | # |
---|
Makale | Yazar | # |
---|