Hydrogen production is gaining increasing importance as a key component of the transition toward carbon-neutral energy systems. In this study, the prediction of hydrogen generation in membraneless alkaline water electrolyzers (MAWEs) is investigated using deep-learning-based time-series modeling. A single-input modeling framework is adopted, where only the system current is used as the input variable. Experimental current signals obtained from long-duration tests conducted at electrolyte concentrations between 5 and 35 g KOH (7200 s per experiment) are employed as the model inputs, while mass-based hydrogen production (in grams) is used as the output variable. Two recurrent neural network architectures, namely Long Short-Term Memory (LSTM) and Gated Recurrent Unit (GRU), are implemented, and their predictive performance is comparatively evaluated using RMSE, MAE, and R² metrics. In addition to deep learning models, classical approaches including Linear Regression, ARIMA, and Naïve Forecast are also considered for comparison. The results show that both models are capable of accurately reproducing the hydrogen-production dynamics across the entire concentration range. In particular, the prediction accuracy improves notably at medium and high electrolyte concentrations, where the coefficient of determination (R²) approaches 0.98. The residual distributions remain narrow and symmetric around zero, indicating the absence of systematic estimation bias. The results also show that classical models can achieve comparable performance under stable operating conditions, while deep learning models provide advantages in capturing nonlinear and dynamic behavior. While LSTM and GRU exhibit comparable accuracy, each architecture provides complementary advantages under different operating conditions. These findings indicate that deep-learning-based time-series modeling constitutes a lightweight and reliable framework for prediction and control applications in MAWE systems. Overall, this study demonstrates the applicability of data-driven models for the dynamic characterization of membraneless water electrolysis.

This study presents a phenomenological estimation framework for a membraneless alkaline water electrolyzer (MAWE), developed primarily from experimentally measured current signals and end-of-test mass-loss data. Thirteen KOH concentrations (5–35 g in 1 L deionized water) were investigated under a constant 12 V DC supply for 7200 s. The time-varying current was continuously recorded throughout each experiment, while the total gas production was determined from the net mass loss measured at the end of the electrolysis process. A time-resolved hydrogen-production representation was subsequently reconstructed from the measured current signal using Faraday’s law and constrained to be stoichiometrically consistent with the experimentally observed total mass loss. The term “charge-consistent” used throughout this study does not imply a new electrochemical principle, but rather refers to maintaining physical consistency between the experimentally measured current signal, Faraday-based charge transfer, and the experimentally observed end-of-test mass loss within the proposed phenomenological framework. Experimental results indicate that both the current response and the cumulative gas production exhibit a strong and distinctly nonlinear dependence on the KOH concentration. Two phenomenological modeling approaches were examined. The first is a static polynomial formulation describing the nonlinear relationship between the measured current signal and the reconstructed production rate. The second is a semi-empirical grey-box formulation in which the Faraday-based theoretical production term is corrected using an experimentally identified efficiency coefficient. Model performance was assessed using train/test data partitioning, residual analysis, autocorrelation functions, and Ljung–Box tests, demonstrating a high degree of internal charge consistency and macroscopic agreement with the reconstructed experimental representation. The proposed framework provides a reduced-order and experimentally accessible approach for representing reconstructed production behavior in MAWE systems without resorting to detailed multi-physics modeling or EIS-based characterization and offers a physically consistent baseline for comparison with more complex data-driven or control-oriented modeling strategies.

Growing concerns about the energy consumption of heating and cooling systems in buildings have led to an increased demand for construction materials with enhanced insulation performance. Accurate determination of the thermal properties of lightweight concrete (LCW) is vital for evaluating its energy performance; however, conventional high-precision tests are both time-intensive and expensive. In this study, the Group Method of Data Handling (GMDH) model is proposed as a tool for directly predicting the thermal properties of LCW from its mechanical properties. For the experimental study, LCWs were produced using waste rubber, pumice, and expanded perlite aggregates as substitutes. Three prediction scenarios were established to accurately estimate thermal conductivity (λ), specific heat (c _p ), and thermal diffusivity (α) from key mechanical properties, including compressive strength (σ _c ), tensile strength (σ _t ), porosity (ϕ), bulk density (ρ), and ultrasonic pulse velocity (upv). Using the GMDH model, mathematical relationships were established between mechanical inputs and thermal outputs, with the most influential parameters identified through layer-by-layer transformations. A topology consisting of two hidden layers, each with three neurons, was adopted, and the model was benchmarked against five machine learning algorithms. The results showed that GMDH provided high predictive accuracy, particularly for thermal conductivity, with an R ² of 0.9957 and low error metrics. These findings indicate that GMDH can serve as an effective and cost-efficient approach for predicting the thermal behavior of LCW, while supporting the sustainable use of alternative aggregates in energy-efficient construction.

Sadece metindeki bozuklukları (kelime birleşmeleri ve satır sonu kaymaları) düzelttim. Herhangi bir kelime eklemedim, silmedim veya stil değişikliği yapmadım: This paper presents a novel approach to designing a proportional-integral-derivative (PID) controller for DC-DC buck converters, which are widely employed for their voltage regulation capabilities despite inherent non-linearities. To address the need for robust and efficient control, the Reptile Search Algorithm, a recently introduced metaheuristic optimization technique inspired by crocodile hunting strategies, is utilized to fine-tune the proportional-integral-derivative controller parameters. Simulations conducted in the MATLAB/Simulink environment evaluated the objective function based on maximum overshoot, steady-state error, settling time, and rise time. The effectiveness of Reptile Search Algorithm was validated through benchmark functions, statistical boxplot analysis, Wilcoxon signed-rank test, transient response, frequency response, disturbance rejection, and robustness analyses. The Reptile Search Algorithm–optimized proportional-integral-derivative controller achieved a 30% reduction in settling time, a 50% decrease in overshoot, and a 15% improvement in phase margin compared to established optimization algorithms, such as Artificial Ecosystem Optimization, Particle Swarm Optimization, and Differential Evolution. Convergence speed analysis revealed that Reptile Search Algorithm converges 20% faster than Particle Swarm Optimization and 15% faster than Differential Evolution, with a 12% lower computational cost per iteration. Real-time testing on a laboratory prototype confirmed a 30% lower overshoot under a 20% disturbance signal. These findings highlight Reptile Search Algorithm’s potential to significantly enhance the dynamic performance and voltage regulation of buck converters in applications such as consumer electronics, motor control systems, and renewable energy setups, improving efficiency and stability. However, the study is limited to simulation and a laboratory prototype, with no extensive experimental validation across diverse operating conditions, suggesting a need for future real-world testing to confirm scalability and robustness. Cite this article as: D. Sevim and N. Bozkurt, "Optimization of proportional-integral-derivative controller for DC-DC buck converters using the reptile search algorithm," Electrica, 26, 0076, 2026. doi:10.5152/electrica.2026.25076.

Proton Değişim Membranlı (PEM) elektrolizörlerde, suyun hidrojen ve oksijene ayrışması, uygulanan Doğru Akım (DA) kadar sistemin nem yönetimine de bağlıdır. Nem, proton iletkenliğini artırarak elektrokimyasal süreci desteklerken, aşırı birikimi (flooding) gaz difüzyonunu engelleyerek sistem performansını ciddi şekilde düşürmektedir. Bu çalışmada, nemin PEM elektrolizörlerdeki etkileri detaylı şekilde analiz edilmiş, flooding mekanizmaları incelenmiş ve Gaz Difüzyon Katmanı (GDL) tasarımının bu sorunun hafifletilmesine yönelik katkıları değerlendirilmiştir. Optimum nem seviyelerinde maksimum elektrolitik verim elde edilirken, yetersiz nem koşulları proton iletkenliğini sınırlandırmakta, aşırı su birikimi ise elektrot yüzeylerini tıkayarak gaz çıkışını engellemektedir. Bu bağlamda, GDL'nin hidrofobik kaplamalar, gradyan gözenek yapıları, mikrokanallar ve hibrit malzeme kombinasyonları ile optimize edilmesi, etkin su yönetimi sağlayarak flooding riskini minimize etmektedir.