Enhancing Electrical Load Prediction Using a Bidirectional LSTM Neural Network

Electronics, Journal Year: 2023, Volume and Issue: 12(22), P. 4652 - 4652

Published: Nov. 15, 2023

Precise anticipation of electrical demand holds crucial importance for the optimal operation power systems and effective management energy markets within domain planning. This study builds on previous research focused application artificial neural networks to achieve accurate load forecasting. In this paper, an improved methodology is introduced, centering around bidirectional Long Short-Term Memory (LSTM) (NN). The primary aim proposed LSTM network enhance predictive performance by capturing intricate temporal patterns interdependencies time series data. While conventional feed-forward are suitable standalone data points, consumption characterized sequential dependencies, necessitating incorporation memory-based concepts. model designed furnish prediction framework with capacity assimilate leverage information from both preceding forthcoming steps. augmentation significantly bolsters capabilities encapsulating contextual understanding Extensive testing performed using multiple datasets, results demonstrate significant improvements in accuracy compared simpleRNN-based framework. successfully captures underlying dependencies data, achieving superior as gauged metrics such root mean square error (RMSE) absolute (MAE). outperforms models, a remarkable RMSE, attesting its forecast impending precision. extended contributes field leveraging forecasting accuracy. Specifically, BiLSTM’s MAE 0.122 demonstrates accuracy, outperforming RNN (0.163), (0.228), GRU (0.165) approximately 25%, 46%, 26%, best variation all networks, at 24-h step, while RMSE 0.022 notably lower than that (0.033), (0.055), respectively. findings highlight significance incorporating memory advanced architectures precise prediction. has potential facilitate more efficient planning market management, supporting decision-making processes systems.

Language: Английский

---

Stochastic scheduling of energy sharing in reconfigurable multi-microgrid systems in the presence of vehicle-to-grid technology

Electric Power Systems Research, Journal Year: 2024, Volume and Issue: 231, P. 110285 - 110285

Published: March 11, 2024

On the one hand, inherent intermittency in demands and renewable energy sources (RES) frequently bring challenges such as overload or surplus generation within microgrids. other electric vehicle aggregations (EVAs) have garnered substantial attention a pivotal strategy to address climate change serve sustainable substitute for petroleum-based vehicles. However, uncoordinated deployment of EVAs microgrids, especially face intermittent nature RES, poses potential threat secure operation microgrid systems. To tackle mentioned issues, this research concentrates on interconnecting group scattered microgrids create multi-microgrid system. In more detail, by developing an management reconfigure interconnections among efficient exchange power these systems is facilitated, addressing variability load amidst stochastic patterns RESs. Besides, grid-to-vehicle (G2V) vehicle-to-grid (V2G) concepts are synchronized reconfigurable structure enhance flexibility model. evaluate model under realistic situations, scenario-based method also employed reflect effects uncertainties The proposed approach, characterized its mathematical convexity, allows employing solvers like CPLEX, ensuring attainment feasible global solution finite timeframe. effectiveness demonstrated through implementation modified 33-bus test system operated results show approach promising tool optimizing presence EVAs, leading operational cost reduction voltage profile enhancement.

Language: Английский

---

Optimal planning for integrated electricity and heat systems using CNN-BiLSTM-attention network forecasts

Energy, Journal Year: 2024, Volume and Issue: 309, P. 133042 - 133042

Published: Aug. 31, 2024

Language: Английский

---

Optimization of power system load forecasting and scheduling based on artificial neural networks

Energy Informatics, Journal Year: 2025, Volume and Issue: 8(1)

Published: Jan. 8, 2025

Abstract This study seeks to enhance the accuracy and economic efficiency of power system load forecasting (PSLF) by leveraging Artificial Neural Networks. A predictive model based on a Residual Connection Bidirectional Long Short Term Memory Attention mechanism (RBiLSTM-AM) is proposed. In this model, normalized time series data used as input, with network capturing bidirectional dependencies residual connections preventing gradient vanishing. Subsequently, an attention applied capture influence significant steps, thereby improving prediction accuracy. Based forecasting, Particle Swarm Optimization (PSO) algorithm employed quickly determine optimal scheduling strategy, ensuring safety system. Results show that proposed RBiLSTM-AM achieves 96.68%, precision 91.56%, recall 90.51%, F1-score 91.37%, significantly outperforming other models (e.g., Recurrent Network which has 69.94%). terms error metrics, reduces root mean square 123.70 kW, absolute 104.44 percentage (MAPE) 5.62%, all are lower than those models. Economic cost analysis further demonstrates PSO strategy costs at most points compared Genetic Algorithm (GA) Simulated Annealing (SA) strategies, being 689.17 USD in first hour 2214.03 fourth hour, both GA SA. Therefore, demonstrate benefits PSLF, providing effective technical support for optimizing scheduling.

Language: Английский

---

Bearing fault diagnosis method based on maximum noise ratio kurtosis product deconvolution with noise conditions

Measurement, Journal Year: 2023, Volume and Issue: 221, P. 113542 - 113542

Published: Sept. 6, 2023

Language: Английский

---

Development and Implementation of a Flexibility Platform for Active System Management at Both Transmission and Distribution Level in Greece

Applied Sciences, Journal Year: 2023, Volume and Issue: 13(20), P. 11248 - 11248

Published: Oct. 13, 2023

This work focused on prescribing, designing, implementing, and evaluating a pilot project conducted in the Greek power system that addressed balancing congestion management issues operators (SOs) face within clean energy era. The considered fully development of F-channel platform, including idea behind this application, steps were taken process, outcomes performed activities fitting into overall picture OneNet project. specified platform is web-based, client-server application uses artificial intelligence (AI) techniques cloud computation engines to improve active for TSO-DSO coordination. flexibility grid’s resources was identified, an integrated monitoring based precise forecasting variable generation demand implemented. focus areas management, frequency control, voltage control services, which corresponding network models created close cooperation with operators. obtained results are essential remaining demonstration because they offer incredibly accurate basis further research their use other weather-related enhanced transmission distribution planning operation practices.

Language: Английский

---

A Flexibility Platform for Managing Outages and Ensuring the Power System’s Resilience during Extreme Weather Conditions

Processes, Journal Year: 2023, Volume and Issue: 11(12), P. 3432 - 3432

Published: Dec. 14, 2023

It is challenging for the European power system to exactly predict RES output and match energy production with demand due changes in wind sun intensity unavoidable disruptions caused by severe weather conditions. Therefore, order address so-called “flexibility challenge” implement variable production, Union needs flexible solutions. In accommodate quicker reactions, compared those performed today, adaptive exploitation of flexibility, grid operators must adjust their operational business model, as electrical transitions from a fully centralized largely decentralized system. OneNet aspires complete this crucial step setting up new generation services that can utilize distributed generation, storage, responses while also guaranteeing fair, open, transparent conditions consumer. Using AI methods cloud-computing approach, current work anticipates active management TSO–DSO coordination will be improved web-based client-server application F-channel. work, user’s experience platform Business Use Case (BUC) under scenario presented. The aims increase reliability outage maintenance plans (SOs) granting them more accurate insight into which may forced operate upcoming period challenges it might face based on way, methodology applied case could, via AI-driven data exchange analyses, help SOs change so potential grave consequences avoided. have forecasts relevant parameters at disposal used achieve set targets. main results presented are has major contribution optimal allocation available resources, ensures voltage frequency stability system, provides an early warning hazardous regimes.

Language: Английский

---

Remote work might unlock solar PV's potential of cracking the ‘Duck Curve’

Applied Energy, Journal Year: 2024, Volume and Issue: 367, P. 123378 - 123378

Published: May 16, 2024

Integrating renewable energy technologies into a decentralised smart grid presents the 'Duck Curve' challenge — disparity between peak demand and solar photovoltaic (PV) yield. Smart operators still lack an effective solution to this problem, resulting in need maintain standby fossil fuel-fired plants. The COVID-19 pandemic-induced lockdowns necessitated shift remote work (work-from-home) home-based education. primary objective of study was explore mitigating strategies for duck curve by investigating notable behaviour examining effect education on PV electricity use 100 households with battery storage southwest UK. This examined 1-min granular consumption data April–August 2019 2020. findings revealed statistically significant disparities demand. Notably, there 1.4—10% decrease average from April August 2020 (during following lockdown) compared corresponding months 2019. Furthermore, household reduced 24—25%, while self-consumption systems increased 7—8% during lockdown May increase particularly prominent morning afternoon, possibly attributed growing prevalence work-from-home dynamic shifts patterns emphasised role meeting evolving needs unprecedented societal changes. Additionally, might unlock PV's potential resolving Curve', urging further investigation implications infrastructure policy development.

Language: Английский

---

Enhancing Electricity Load Forecasting with Machine Learning and Deep Learning

Technologies, Journal Year: 2025, Volume and Issue: 13(2), P. 59 - 59

Published: Feb. 1, 2025

The electricity load forecasting handles the process of determining how much will be available at a given time while maintaining balance and stability power grid. accuracy plays an important role in ensuring safe operation improving reliability systems is key component operational planning efficient market. For many years, conventional method has been used by using historical data as input parameters. With swift progress improvement technology, which shows more potential due to its accuracy, different methods can applied depending on identified model. To enhance forecast load, this paper introduces proposes framework developed graph database technology archive large amounts data, collects measured from electrical substations Pristina, Kosovo. includes weather parameters collected over four-year timeframe. proposed designed handle short-term forecasting. Machine learning Linear Regression deep Long Short-Term Memory algorithms are multiple datasets mean absolute error root square calculated. results show promising performance effectiveness model, with high

Language: Английский

---

Cloud-based estimation of lithium-ion battery life for electric vehicles using equivalent circuit model and recurrent neural network

Journal of Energy Storage, Journal Year: 2025, Volume and Issue: 114, P. 115718 - 115718

Published: Feb. 10, 2025

Language: Английский

---