Artificial Neural Networks for Photovoltaic Power Forecasting: A Review of Five Promising Models

IEEE Access, Journal Year: 2024, Volume and Issue: 12, P. 90461 - 90485

Published: Jan. 1, 2024

Solar energy is largely dependent on weather conditions, resulting in unpredictable, fluctuating, and unstable photovoltaic (PV) power outputs. Thus, accurate PV forecasts are increasingly crucial for managing controlling integrated systems. Over the years, advanced artificial neural network (ANN) models have been proposed to increase accuracy of various geographical regions. Hence, this paper provides a state-of-the-art review five most popular ANN forecasting. These include multilayer perceptron (MLP), recurrent (RNN), long short-term memory (LSTM), gated unit (GRU), convolutional (CNN). First, internal structure operation these studied. It then followed by brief discussion main factors affecting their forecasting accuracy, including horizons, meteorological evaluation metrics. Next, an in-depth separate analysis standalone hybrid provided. has determined that bidirectional GRU LSTM offer greater whether used as model or configuration. Furthermore, upgraded metaheuristic algorithms demonstrated exceptional performance when applied models. Finally, study discusses limitations shortcomings may influence practical implementation

Language: Английский

---

A novel prediction model for wind power based on improved long short-term memory neural network

Energy, Journal Year: 2022, Volume and Issue: 265, P. 126283 - 126283

Published: Dec. 3, 2022

Language: Английский

---

Hybrid Inception-embedded deep neural network ResNet for short and medium-term PV-Wind forecasting

Energy Conversion and Management, Journal Year: 2023, Volume and Issue: 294, P. 117574 - 117574

Published: Aug. 25, 2023

Language: Английский

---

Evolving long short-term memory neural network for wind speed forecasting

Information Sciences, Journal Year: 2023, Volume and Issue: 632, P. 390 - 410

Published: March 9, 2023

Wind speed forecasting plays a crucial role in reducing the risk of wind power uncertainty, which is vital for system planning, scheduling, control, and operation. However, it challenging to obtain accurate results since series contain complex fluctuations. In this paper, novel model proposed by using genetic algorithm (GA) long short-term memory neural network (LSTM), where GA used evolving architectures hyper-parameters LSTM, called EvLSTM, because there no clear knowledge determine these parameters. EvLSTM model, flexible gene encoding strategy, crossover operation, mutation operation are describe different LSTM during evolutionary process. addition, overcome weakness single method forecasting, ensemble (EnEvLSTM) negative constraint theory learning whose weight coefficients determined differential evolution algorithm. The EnEvLSTM models evaluated on two real-world farms located Inner Mongolia, China Sotavento Galicia, Spain. Experimental horizons demonstrate superiority terms three performance indices statistical tests.

Language: Английский

---

A hybrid short-term wind power point-interval prediction model based on combination of improved preprocessing methods and entropy weighted GRU quantile regression network

Energy, Journal Year: 2023, Volume and Issue: 288, P. 129904 - 129904

Published: Dec. 6, 2023

Language: Английский

---

A survey of artificial intelligence methods for renewable energy forecasting: Methodologies and insights

Renewable energy focus, Journal Year: 2023, Volume and Issue: 48, P. 100529 - 100529

Published: Dec. 20, 2023

The efforts to revolutionize electric power generation and produce clean sustainable electricity have led the exploration of renewable energy systems (RES). This form is replenished cost-effective in terms production maintenance. However, RES, such as solar wind energies, intermittent; this one drawbacks its usage. In order overcome limitation, studies been undertaken forecast availability output. current trending method forecasting generated by RES artificial intelligence (AI) method. with all potential, traditional AI, Artificial Neural Network (ANN), Support Vector Machine (SVM) many more, does not it all. Because this, metaheuristic algorithms are being explored optimization techniques increase performance accuracy these AI methods some challenges models. study presents an insightful survey (traditional metaheuristic) systems. A existing surveyed literature was presented. taxonomy formulated, theoretical backgrounds were Also, various forms improved versions applied optimize classical systems' output surveyed. conceptual framework hybrid application formulated. Finally, discussion, insight, models future directions

Language: Английский

---

4PL routing problem using hybrid beetle swarm optimization

Soft Computing, Journal Year: 2023, Volume and Issue: 27(22), P. 17011 - 17024

Published: May 23, 2023

Language: Английский

---

Short-Term Wind Power Forecasting Based on VMD and a Hybrid SSA-TCN-BiGRU Network

Applied Sciences, Journal Year: 2023, Volume and Issue: 13(17), P. 9888 - 9888

Published: Aug. 31, 2023

Wind power generation is a renewable energy source, and its output influenced by multiple factors such as wind speed, direction, meteorological conditions, the characteristics of turbines. Therefore, accurately predicting crucial for grid operation maintenance management plants. This paper proposes hybrid model to improve accuracy prediction. Accurate forecasting critical safe systems. To prediction, this incorporating variational modal decomposition (VMD), Sparrow Search Algorithm (SSA), temporal-convolutional-network-based bi-directional gated recurrent unit (TCN-BiGRU). The first uses VMD break down raw data into several components, then it builds an SSA-TCN-BIGRU each component finally, accumulates all predicted components obtain prediction results. proposed short-term was validated using measured from farm in China. VMD-SSA-TCN-BiGRU framework compared with benchmark models verify practicability reliability. Compared TCN-BiGRU, symmetric mean absolute percentage error, root square error reduced 34.36%, 49.14%, 55.94%.

Language: Английский

---

Resource efficient PV power forecasting: Transductive transfer learning based hybrid deep learning model for smart grid in Industry 5.0

Energy Conversion and Management X, Journal Year: 2023, Volume and Issue: 20, P. 100486 - 100486

Published: Oct. 1, 2023

This paper presents an innovative approach for enhancing power output forecasting of Photovoltaic (PV) plants in dynamic environmental conditions using a Hybrid Deep Learning Model (DLM). The hybrid DLM employs synergy Convolutional Neural Network (CNN), Long Short-Term Memory (LSTM) network, and Bidirectional LSTM (Bi-LSTM), effectively capturing spatial temporal dependencies within weather data crucial accurate predictions. To optimize the DLM's performance efficiently, unique Kepler Optimization Algorithm (KOA) is introduced hyperparameter tuning, drawing inspiration from Kepler's laws planetary motion. By leveraging KOA, attains optimal configurations, elevating prediction precision. Additionally, this study integrates Transductive Transfer (TTL) with deep learning models to enhance resource efficiency. knowledge gained previously learned tasks, TTL enables improve its capabilities while minimizing utilization. Datasets encompassing parameters PV plant-generated across diverse sites are employed training testing. Three models, amalgamating CNN, LSTM, Bi-LSTM techniques, evaluated. Comparative assessment these distinct yields insightful observations. Performance evaluation, focused on short-term forecasting, underscores superiority over individual CNN models. achieves remarkable accuracy resilience predicting under varying conditions, showcasing potential efficient plant management.

Language: Английский

---

Novel wind power ensemble forecasting system based on mixed-frequency modeling and interpretable base model selection strategy

Energy, Journal Year: 2024, Volume and Issue: 297, P. 131142 - 131142

Published: April 3, 2024

Language: Английский

---