Development of New Electricity System Marginal Price Forecasting Models Using Statistical and Artificial Intelligence Methods DOI Creative Commons
Mehmet Kızıldağ, Fatih Abut, Mehmet Fatih Akay

и другие.

Applied Sciences, Год журнала: 2024, Номер 14(21), С. 10011 - 10011

Опубликована: Ноя. 2, 2024

The System Marginal Price (SMP) is the cost of last unit electricity supplied to grid, reflecting supply–demand equilibrium and serving as a key indicator market conditions. Accurate SMP forecasting essential for ensuring stability economic efficiency. This study addresses challenges prediction in Turkey by proposing comprehensive framework that integrates machine learning, deep statistical models. Advanced feature selection techniques, such Minimum Redundancy Maximum Relevance (mRMR) Likelihood Feature Selector (MLFS), are employed refine model inputs. incorporates time series methods like Multilayer Perceptron (MLP), Long Short-Term Memory (LSTM), Bidirectional LSTM (Bi-LSTM), Convolutional (ConvLSTM) capture complex temporal patterns, alongside models Support Vector Machine (SVM), Extreme Gradient Boosting (XGBoost), Learning (ELM) modeling non-linear relationships. Model performance was evaluated using Mean Absolute Percentage Error (MAPE) across regular weekdays, weekends, public holidays. XGBoost combined with MLFS consistently achieved lowest MAPE values, demonstrating exceptional accuracy robustness. Among all models, superior results highlight inadequacy traditional ARIMA SARIMA capturing highly volatile reinforcing necessity advanced techniques effective forecasting. Overall, this presents novel approach tailored markets, significantly enhancing predictive reliability incorporating indicators sophisticated methods.

Язык: Английский

Modeling the Efficiency of Resource Consumption Management in Construction Under Sustainability Policy: Enriching the DSEM-ARIMA Model DOI Open Access
Pruethsan Sutthichaimethee, Grzegorz Mentel, Volodymyr Voloshyn

и другие.

Sustainability, Год журнала: 2024, Номер 16(24), С. 10945 - 10945

Опубликована: Дек. 13, 2024

The aim of this research is to study the influence factors affecting efficiency resource consumption under sustainability policy based on using DSEM-ARIMA (Dyadic Structural Equation Modeling Autoregressive Integrated Moving Average) model. performed Thailand experience. findings indicate that continuous economic growth aligns with country’s objectives, directly contributing social growth. This efficient planning. It demonstrates management goal achieving 5.0. Furthermore, considering environmental aspect, it found and impacts ecological aspect due significant in construction. construction shows a rate increase 264.59% (2043/2024), reaching 401.05 ktoe (2043), which exceeds carrying capacity limit set at 250.25 ktoe, resulting long-term degradation. Additionally, political have greatest environment, exacerbating damage beyond current levels. Therefore, model establishes new scenario policy, indicating leads degradation reduced 215.45 does not exceed capacity. Thus, if utilized, can serve as vital tool formulating policies steer toward 5.0 effectively.

Язык: Английский

Процитировано

5
Towards Efficient Electricity Management in Benghazi DOI Creative Commons
Asma Agaal,

Hend M. Farkash,

Mansour Essgaer

и другие.

Solar Energy and Sustainable Development, Год журнала: 2025, Номер 14(FICTS-2024), С. 110 - 136

Опубликована: Янв. 29, 2025

In Libya, the general electricity company is tasked with managing peak demand, often resorting to load shedding. This practice, while necessary, results in power outages, particularly impacting areas like Benghazi Electrical Grid. study aims bring predictability these events by exploring time series forecasting models namely: Autoregressive Integrated Moving Average (ARIMA), Seasonal ARIMA (SARIMA), and Dynamic Regression (DRARIMA). The were trained using data from May 2020 2021, subsequently tested on 2022. Performance was evaluated metrics such as mean squared error, absolute percentage accuracy. model achieved highest accuracy at 78.88% a error of 0.9. SARIMA model, which considers seasonal patterns, an 73.86% 0.11, but its complexity may lead overfitting. DRARIMA, incorporates exogenous variables, demonstrated 65.36% 0.15. Future projections for 2024 2025 indicate potential improvements shedding management highlight importance selection accurate forecasting. By improving accuracy, this research enhance effectiveness management, thereby reducing outages their socio-economic impacts regions Benghazi. These findings are valuable energy planners managers similar contexts, providing practical insights data-driven strategies.

Язык: Английский

Процитировано

0
Sustainable Development Goals and Supply Chains for Driving Positive Impact and Resilience DOI
Yasemin Ülker, Banu Demirel

IGI Global eBooks, Год журнала: 2025, Номер unknown, С. 1 - 32

Опубликована: Фев. 5, 2025

This chapter strives to draw attention the role of supply chains in delivering sustainable development increase awareness strategic value chain strength achieving several SDGs attain environmental sustainability particular. Using a quantitative methodology, study examines relationship between and performance BRIC MIKTA countries, based on Global Competitive Index Environmental Performance (EPI). Results reveal strong, significant positive association, with explaining 44.7% variance strength.

Язык: Английский

Процитировано

0
Analysis of SARIMA Models for Forecasting Electricity Demand DOI
Ahmet Aksöz, Saadin Oyucu, Emre Biçer

и другие.

Опубликована: Май 27, 2024

Язык: Английский

Процитировано

1
The Characteristics of ARMA (ARIMA) Model and Some Key Points to Be Noted in Application: A Case Study of Changtan Reservoir, Zhejiang Province, China DOI Open Access
Zhuang Liu, Yibin Cui,

Chengcheng Ding

и другие.

Sustainability, Год журнала: 2024, Номер 16(18), С. 7955 - 7955

Опубликована: Сен. 12, 2024

Accurate water quality prediction is the basis for good environment management and sustainable use of resources. As an important time series forecasting model, Autoregressive Moving Average Model (ARMA) plays a crucial role in environmental sustainability research. This study addresses factors that affect ARMA model’s forecast accuracy goodness fit. The research results show sample size used model parameters estimation main influencing factor fit affecting error model. Constructing stable reliable requires certain number samples parameters. However, using excessive will not further improve but rather increase workload difficulty data collection. suitable long-term because models increases with time, when exceeds limit, fitted values almost no longer change which means has lost its significance prediction. For periodic components, introducing adjustment into can reduce error. These findings enable managers researchers to apply more rationally, hence developing precise pollution control development plans.

Язык: Английский

Процитировано

0
Development of New Electricity System Marginal Price Forecasting Models Using Statistical and Artificial Intelligence Methods DOI Creative Commons
Mehmet Kızıldağ, Fatih Abut, Mehmet Fatih Akay

и другие.

Applied Sciences, Год журнала: 2024, Номер 14(21), С. 10011 - 10011

Опубликована: Ноя. 2, 2024

The System Marginal Price (SMP) is the cost of last unit electricity supplied to grid, reflecting supply–demand equilibrium and serving as a key indicator market conditions. Accurate SMP forecasting essential for ensuring stability economic efficiency. This study addresses challenges prediction in Turkey by proposing comprehensive framework that integrates machine learning, deep statistical models. Advanced feature selection techniques, such Minimum Redundancy Maximum Relevance (mRMR) Likelihood Feature Selector (MLFS), are employed refine model inputs. incorporates time series methods like Multilayer Perceptron (MLP), Long Short-Term Memory (LSTM), Bidirectional LSTM (Bi-LSTM), Convolutional (ConvLSTM) capture complex temporal patterns, alongside models Support Vector Machine (SVM), Extreme Gradient Boosting (XGBoost), Learning (ELM) modeling non-linear relationships. Model performance was evaluated using Mean Absolute Percentage Error (MAPE) across regular weekdays, weekends, public holidays. XGBoost combined with MLFS consistently achieved lowest MAPE values, demonstrating exceptional accuracy robustness. Among all models, superior results highlight inadequacy traditional ARIMA SARIMA capturing highly volatile reinforcing necessity advanced techniques effective forecasting. Overall, this presents novel approach tailored markets, significantly enhancing predictive reliability incorporating indicators sophisticated methods.

Язык: Английский

Процитировано

0