Deep learning models for solar irradiance forecasting: A comprehensive review

Journal of Cleaner Production, Journal Year: 2021, Volume and Issue: 318, P. 128566 - 128566

Published: Aug. 11, 2021

Language: Английский

---

Physical energy and data-driven models in building energy prediction: A review

Energy Reports, Journal Year: 2022, Volume and Issue: 8, P. 2656 - 2671

Published: Feb. 10, 2022

The difficulty in balancing energy supply and demand is increasing due to the growth of diversified flexible building resources, particularly rapid development intermittent renewable being added into power grid. accuracy consumption prediction top priority for electricity market management ensure grid safety reduce financial risks. speed load are fundamental prerequisites different objectives such as long-term planning short-term optimization systems buildings past few decades have seen impressive time series forecasting models focusing on domains objectives. This paper presents an in-depth review discussion models. Three widely used approaches, namely, physical (i.e., white box), data-driven black hybrid grey were classified introduced. principles, advantages, limitations, practical applications each model investigated. Based this review, research priorities future directions domain highlighted. conclusions drawn could guide prediction, therefore facilitate efficiency buildings.

Language: Английский

---

Machine Learning, Deep Learning and Statistical Analysis for forecasting building energy consumption — A systematic review

Engineering Applications of Artificial Intelligence, Journal Year: 2022, Volume and Issue: 115, P. 105287 - 105287

Published: Aug. 12, 2022

Language: Английский

---

Building energy prediction using artificial neural networks: A literature survey

Energy and Buildings, Journal Year: 2021, Volume and Issue: 262, P. 111718 - 111718

Published: Nov. 26, 2021

Language: Английский

---

Photovoltaic power forecast based on satellite images considering effects of solar position

Applied Energy, Journal Year: 2021, Volume and Issue: 302, P. 117514 - 117514

Published: Aug. 13, 2021

Language: Английский

---

District heater load forecasting based on machine learning and parallel CNN-LSTM attention

Energy, Journal Year: 2022, Volume and Issue: 246, P. 123350 - 123350

Published: Feb. 1, 2022

Accurate heat load forecast is important to operate combined and power (CHP) efficiently. This paper proposes a parallel convolutional neural network (CNN) - long short-term memory (LSTM) attention (PCLA) model that extracts spatiotemporal characteristics then intensively learns importance. PCLA by derived spatial temporal features parallelly from CNNs LSTMs. The novelty of this lies in the following three aspects: 1) for forecasting proposed; 2) it demonstrated performance superior 12 models including serial coupled model; 3) using LSTMs better than one principal component analysis. dataset includes district heater related variables, load-derived weather forecasts time factors affect loads. accuracy reflected lowest values mean absolute squared errors 0.571 0.662, respectively, highest R-squared value 0.942. therefore previously proposed demand expected be useful CHP plant management.

Language: Английский

---

Predicting hourly heating load in a district heating system based on a hybrid CNN-LSTM model

Energy and Buildings, Journal Year: 2021, Volume and Issue: 243, P. 110998 - 110998

Published: April 21, 2021

Language: Английский

---

Two-stage Optimal Dispatching of AC/DC Hybrid Active Distribution Systems Considering Network Flexibility

Journal of Modern Power Systems and Clean Energy, Journal Year: 2023, Volume and Issue: 11(1), P. 52 - 65

Published: Jan. 1, 2023

The increasing flexibility of active distribution systems (ADSs) coupled with the high penetration renewable distributed generators (RDGs) leads to increase complexity. It is practical significance achieve largest amount RDG in ADSs and maintain optimal operation. This study establishes an alternating current (AC)/direct (DC) hybrid ADS model that considers dynamic thermal rating, soft open point, network reconfiguration (DNR). Moreover, it transforms dispatching into a second-order cone programming problem. Considering different control time scales dispatchable resources, following two-stage framework proposed. ① day-ahead dispatch uses hourly input data goal minimizing grid loss dropout. obtains 24-hour schedule determine plans for DNR energy storage system. ② intraday 15 min 1-hour rolling-plan but only executes first dispatching. To eliminate error between actual operation plan, divided three 5-min step-by-step executions. each step trace tie-line power greatest extent at minimum cost. measured are used as feedback after executed. A case shows comprehensive cooperative can release line capacity, reduce losses, improve rate RDGs. Further, frame-work handle source-load fluctuations enhance system stability.

Language: Английский

---

Comparison of reinforcement learning and model predictive control for building energy system optimization

Applied Thermal Engineering, Journal Year: 2023, Volume and Issue: 228, P. 120430 - 120430

Published: March 27, 2023

Language: Английский

---

Short-Term Load Forecasting Models: A Review of Challenges, Progress, and the Road Ahead

Energies, Journal Year: 2023, Volume and Issue: 16(10), P. 4060 - 4060

Published: May 12, 2023

Short-term load forecasting (STLF) is critical for the energy industry. Accurate predictions of future electricity demand are necessary to ensure power systems’ reliable and efficient operation. Various STLF models have been proposed in recent years, each with strengths weaknesses. This paper comprehensively reviews some models, including time series, artificial neural networks (ANNs), regression-based, hybrid models. It first introduces fundamental concepts challenges STLF, then discusses model class’s main features assumptions. The compares terms their accuracy, robustness, computational efficiency, scalability, adaptability identifies approach’s advantages limitations. Although this study suggests that ANNs may be most promising ways achieve accurate additional research required handle multiple input features, manage massive data sets, adjust shifting conditions.

Language: Английский

---