Aug 2026· Engineering, Technology & Applied Science Research· 0 citations· 19 references
TL;DR
A hybrid Sine Cosine Optimization–Bidirectional Long Short-Term Memory (BiLSTM)–Deep Reinforcement Learning (DRL) framework for smart grid energy management that improves energy utilization efficiency, reduces operating costs, and enhances electrical grid stability is proposed.
Abstract
Smart grid technologies have evolved rapidly to enable intelligent monitoring, forecasting, and control of modern energy systems. However, the increasing integration of renewable energy sources, variable demand, and dynamic grid conditions poses significant challenges for effective energy management. Traditional methods are often not well-suited to handling the nonlinear and time-varying nature of smart grid data, which can lead to reduced prediction accuracy and scheduling efficiency. To address these challenges, this paper proposes a hybrid Sine Cosine Optimization (SCO)–Bidirectional Long Short-Term Memory (BiLSTM)–Deep Reinforcement Learning (DRL) framework for smart grid energy management. The BiLSTM model is employed to accurately forecast energy demand, whereas the SCO algorithm optimizes model parameters to enhance prediction performance. Experimental results show that the proposed model achieves lower Mean Absolute Error (MAE) and Root Mean Square Error (RMSE) values and higher prediction accuracy than conventional models, including Support Vector Machine (SVM), Long Short-Term Memory (LSTM), and BiLSTM. The proposed framework improves energy utilization efficiency, reduces operating costs, and enhances electrical grid stability. Overall, the proposed approach provides a scalable and efficient solution for next-generation smart grid energy management.
For smart grid power system planning and operation, short-term load forecasting is crucial. Important decisions including determining system safety, scheduling fuel, economically dispatching electricity, and selling energy can be aided by accurate day-ahead estimates. However, due to its reliance on external variables like weather, the process is intricate and computationally intensive. In order to address this issue, the paper proposes an LSTGR-based architecture that systematically enhances STLF in Smart Grids. The input characteristics are first scaled correctly using data normalisation. A hybrid feature selection method combining XGB and RF is utilised to determine the most essential features. Afterwards, RFE is employed to eliminate superfluous attributes. To facilitate learning, a hybrid deep learning model is trained using the updated dataset. This model combines LSTM and GRU. Using assessment criteria such as MAPE, MAE, MSE, and RMSE, the results demonstrate that the LSTGR model outperforms other models. With an RMSE of only 1.8%, the model clearly excels at producing accurate predictions. All things considered, the model successfully improves the reliability of forecasts while being computationally efficient. Because of this, it is an excellent option for smart grid applications in the actual world.
L. Jayavani, Banoth Ashwini, Kolkur Swabhavika et al.· 2026 7th International Confe...· 0 citations
The increasing adoption of electric bicycles and motorcycles has intensified the demand for sustainable and reliable charging infrastructures, particularly in campus environments characterized by fluctuating mobility patterns and renewable energy variability. This study proposes an intelligent hybrid solar–wind renewable charging framework integrated with Long Short-Term Memory–Marine Predators Algorithm (LSTM-MPA) optimization to improve adaptive forecasting, charging stability, and renewable energy utilization. The methodology combined systematic meta-analysis and deep learning simulation approaches. A total of 30 empirical studies were analyzed using PRISMA-based selection procedures, risk-of-bias assessment, effect-size evaluation, and publication bias analysis. Experimental renewable energy datasets consisting of photovoltaic (PV), wind turbine, and charging parameters were modeled using the proposed hybrid LSTM-MPA architecture. The results demonstrated that Solar-Wind-Deep Learning systems achieved the highest average effect size of 20.653. The forecasting model achieved Root Mean Square Error (RMSE) values of 115.70–125.65 and Mean Absolute Error (MAE) values of 90.34–93.64, while training and validation losses decreased by more than 39%, indicating stable convergence. Operational analysis showed that hybrid renewable generation consistently exceeded campus charging demand, maintaining battery State of Charge (SoC) near 100% and resulting in a cumulative renewable energy generation of 43,235.30 kWh. Annual renewable energy exceeded 2,070 kWh with a maximum energy balance of 882.64 kWh. The system was economically feasible, achieving an IRR of 14.9% and positive cumulative cash flow after the seventh operational year, while reducing cumulative CO₂ emissions by more than 22,450 kg over 20 years.
M. S. Mauludin, Yuki Trisnoaji, Arif Rifan Rudiyanto et al.· Energy Storage and Conversio...· 0 citations
The integration of renewable energy sources (RES) and electric vehicle (EV) management in the smart grids has significantly introduced new challenges, because of the intermittent nature of the RES and stochastic charging behavior. Existing studies highlight that they focus on modeling either the renewable generation or EV charging demand but not focused on a combined approach effect on the smart grid stability. To address this research gap, this study has proposed a Bi-Directional Long Short-Term Memory (Bi-LSTM) for the grid stability classification under the integrated RES-EV dynamics. The research proposes a Hyperparameter-Optimized Bi-Directional Long Short-Term Memory (HO-Bi-LSTM) framework to improve the classification performance and learning efficiency under dynamic grid operating conditions. Furthermore, this research has used the dataset which is obtained from the Kaggle repository, and this dataset comprises nearly 60000 records with the synthetically generated data samples for the EV charging features to simulate the realistic grid conditions. Based on the experimental evaluation, the HO-Bi-LSTM demonstrated significant performance. Accuracy of 96.34%, Precision of 98.27%, Recall of 95.97%, F1-score of 97.11%. The proposed HO-Bi-LSTM framework can provide the better solutions for improved grid resilience and ensuring the efficient energy distribution and supporting the smart grid operations. The HO-Bi-LSTM framework effectively captures bidirectional temporal dependencies in integrated RES-EV dynamics. Ablation studies confirm the value of synthetic features and hyperparameter optimization.
Neeraj Shrivastava¹, Sanjiv Kumar· Journal of Intelligent Decis...· 0 citations
The high penetration of renewable energy and the large-scale integration of electric vehicles have significantly enhanced the nonlinearity and non-stationarity of the smart grid load sequence. Traditional time series models struggle to meet the prediction requirements in complex scenarios. This paper focuses on three kinds of deep learning load forecasting methods: recurrent neural network variants, CNN hybrid architectures, and Transformers, and systematically conducts a comparative analysis of them from three dimensions: prediction accuracy, long-term dependency modeling ability and computational efficiency. The results show that the three types of architectures each have distinct advantages, and there is no universal optimal solution: the recurrent variant balances accuracy and efficiency in stationary short-term prediction, the CNN hybrid architecture is more robust in strong coupling scenarios such as extreme weather, and the Transformers' trend modeling ability is outstanding but the computational cost for long sequences is large. The pros and cons of each method are highly dependent on the stationarity of the specific scenario, the forecasting horizon, and resource constraints. The model selection must be based on the task characteristics rather than the pursuit of general solutions.
Qiuhui Jiang· Applied and Computational En...· 0 citations
Emerging research directions, such as explainable AI, federated learning, digital twins, edge intelligence, and physics-informed machine learning, are identified as promising strategies for developing resilient, intelligent, and sustainable future power grids.
Olatunde Ibiyinka, Tolu Omotoso, N. Ekekwe· Global Journal of Engineerin...· 0 citations
This study proposes a novel AI-powered smart grid management framework that integrates predictive analytics, machine learning, and adaptive control techniques to optimize energy distribution, minimize transmission losses, and improve overall cost efficiency. The proposed system utilizes a comprehensive dataset comprising energy consumption, energy generation, voltage, current, temperature, wind speed, solar irradiance, battery storage, and dynamic electricity pricing to develop an intelligent decision-making architecture. Advanced machine learning algorithms are employed for energy demand forecasting, power flow optimization, and loss minimization, thereby enhancing grid efficiency, reliability, and renewable energy integration. The results demonstrate that AI-based optimization significantly improves grid resilience, load balancing, adaptive pricing strategies, and operational efficiency, contributing to the development of scalable, intelligent, and sustainable smart grid systems for future energy management.
Sujata Hanumant Kale· Dandao Xuebao/Journal of Bal...· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
