Logo do repositório

Percorrer por autor "Gonzalez, Gabriel Villarrubia"

A mostrar 1 - 3 de 3
  • Differentiable neural search architecture with zero-cost metrics for insulator fault prediction
    Publication . Seman, Laio Oriel; Buratto, William Gouvêa; Gonzalez, Gabriel Villarrubia; Leithardt, Valderi Reis Quietinho; Nied, Ademir; Stefenon, Stefano Frizzo
    Reliable monitoring of high-voltage insulators is critical for maintaining the stability of electrical power systems, particularly under environmental contamination that can lead to flashover. Traditional inspection techniques struggle to anticipate degradation dynamics, while data-driven models often rely on fixed neural architectures that inadequately capture the complex temporal patterns in leakage current signals. This work proposes a Differentiable Neural Architecture Search (DARTS) framework, based on zero-cost metrics, tailored for time series forecasting in insulator monitoring. The method based on DARTS integrates a mixed encoder-decoder design with learnable selection over long short-term memory, gated recurrent units, and transformer components, coupled with a cross-attention bridge featuring temporal bias and gating mechanisms. To ensure efficient architecture exploration, the search leverages metrics such as SynFlow and Jacobian covariance for early candidate screening, followed by a bilevel optimization stage with entropy and diversity regularization. Experiments on real-world leakage current data demonstrate that the discovered architectures outperform manually designed baselines, offering improved forecasting performance.
    2026-03Artigo de investigação Acesso aberto Ver mais
  • Optimized Harmonic-Multiscale Kolmogorov–Arnold Network denoised by EWT for hydroelectric dam useful volume forecast
    Publication . Stefenon, Stefano Frizzo; Seman, Laio Oriel; Matos-Carvalho, João Pedro; Nied, Ademir; Gonzalez, Gabriel Villarrubia; Yow, Kin-Choong; Stefenon, Stefano Frizzo; Elsevier
    Accurate forecasting of the useful volume of a hydroelectric reservoir is essential for energy planning and water resource management. To address this challenge, this paper proposes a novel Harmonic-Multiscale Kolmogorov–Arnold Network forecasting model. The model integrates sinusoidal (Sin) basis functions to model global oscillatory patterns and wavelet (Wav) bases to capture localized transient features within the Kolmogorov Arnold Network (KAN) framework (SinWavKAN). This hybrid representation enhances feature expressiveness by jointly modeling smooth long-term dependencies and short-term irregular dynamics. The input signals are first denoised using the Empirical Wavelet Transform (EWT) to isolate informative modes from noise, and the SinWavKAN’s hyperparameters are optimized via an Adaptive Tree-structured Parzen Estimator (ATPE). An ablation study confirmed the synergistic superiority of the sinusoidal-wavelet combination over other basis functions. Compared to state-of-the-art KAN variants (e.g., fastKAN, PyKAN, WavKAN), the proposed architecture, named in short Opt-EWT-SinWavKAN, demonstrated a reduction in mean absolute error by approximately 85%–93% across forecasting horizons from 1 to 30 days. The results conclusively show that the Opt-EWT-SinWavKAN provides a more reliable approach for hydroelectric reservoir volume forecasting.
    2026Artigo de investigação Acesso aberto Ver mais
  • Optimized wavelet-denoised temporal convolutional network for multi-step ahead predictions in power systems
    Publication . Buratto, William Gouvea; Gonzalez, Gabriel Villarrubia; Yamaguchi, Cristina Keiko; Nied, Ademir; Yow, Kin-Choong; Stefenon, Stefano Frizzo
    Accurate forecasting of power generation is essential for ensuring the reliable and cost-efficient operation of modern electrical grids. This paper introduces an optimized wavelet temporal convolutional network (TCN) designed for multi-step-ahead prediction of thermal power generation in Brazil. The proposed model, called optimized-wavelet-TCN, integrates a discrete wavelet transform denoising stage to effectively filter high-frequency noise from operational data while preserving underlying signal trends, followed by a TCN optimized via a tree-structured Parzen estimator for hyperparameter tuning. Evaluated on real-world hourly data spanning multiple years, the optimized-wavelet-TCN achieves superior forecasting performance with a mean absolute percentage error of 10.2%, 7.24%, and 8.54% for 15, 30, and 60-step ahead predictions, respectively. The model significantly reduces computational time to only 1.54 s for a 30-step forecast, markedly outperforming state-of-the-art transformer-based and recurrent neural network baselines in both accuracy and efficiency.
    2026-08-17Artigo de investigação Acesso aberto Ver mais