Research on Ultra-short-term combination forecasting algorithm of power load based on machine learning DOI Open Access

Jinggeng Gao,

Kun Wang, Xiaohua Kang

и другие.

Journal of Physics Conference Series, Год журнала: 2024, Номер 2846(1), С. 012046 - 012046

Опубликована: Сен. 1, 2024

Abstract Power load forecasting is of great significance to the power grid marketing department. To obtain accurate results, a minute-by-minute method for electricity based on multi-stage proposed (TPE-WXL) by combining non-linear and time-series attributes. Firstly, historical series specific areas in city are pre-processed. Then, order accurately predicted XGBoost LightGBM applied extract attributes from build hybrid model. Moreover, TPE introduced enhance hyperparameters model series. Finally, dataset region used as an example conduct experimental analysis. Experimental results revealed that can forecast trend load, is, R 2 =0.981, RMSE =2.643.

Язык: Английский

Study on deterministic and interval forecasting of electricity load based on multi-objective whale optimization algorithm and transformer model DOI
Pei Du,

Yuxin Ye,

Han Wu

и другие.

Expert Systems with Applications, Год журнала: 2025, Номер 268, С. 126361 - 126361

Опубликована: Янв. 2, 2025

Язык: Английский

Процитировано

1
Short-term electric load forecasting based on series decomposition and Meta-Informer algorithm DOI
Lianbing Li,

Xingchen Guo,

Ruixiong Jing

и другие.

Electric Power Systems Research, Год журнала: 2025, Номер 243, С. 111478 - 111478

Опубликована: Фев. 8, 2025

Язык: Английский

Процитировано

1
EGNL-FAT: An Edge-Guided Non-Local network with Frequency-Aware transformer for smoke segmentation DOI
Yitong Fu, Haiyan Li, Yifan Wang

и другие.

Expert Systems with Applications, Год журнала: 2025, Номер unknown, С. 127621 - 127621

Опубликована: Апрель 1, 2025

Язык: Английский

Процитировано

0
Integrated multi-energy load prediction system with multi-scale temporal channel features fusion DOI
Dezhi Liu, Jiaming Zhu,

Mengyang Wen

и другие.

Measurement, Год журнала: 2025, Номер unknown, С. 117559 - 117559

Опубликована: Апрель 1, 2025

Язык: Английский

Процитировано

0
A hybrid load forecasting system based on data augmentation and ensemble learning under limited feature availability DOI

Qing Yang,

Zhirui Tian

Expert Systems with Applications, Год журнала: 2024, Номер unknown, С. 125567 - 125567

Опубликована: Окт. 1, 2024

Язык: Английский

Процитировано

3
The degree of population aging and carbon emissions: Analysis of mediation effect and multi-scenario simulation DOI Creative Commons
Shuyu Li, Shun Jia, Yang Liu

и другие.

Journal of Environmental Management, Год журнала: 2024, Номер 367, С. 121982 - 121982

Опубликована: Июль 30, 2024

The continuous deepening of aging has posed new challenges for global sustainable development. Measuring the impact population on carbon emissions is crucial next stage climate governance. However, structural changes in social production and consumption make it difficult to evaluate effects. Therefore, this study constructed a bidirectional fixed Space Durbin Model explore mediating pathway aging's emissions. Furthermore, we have established high-precision prediction models simulate evolution trajectory under multi-factor driving scenarios. main findings are as follows: (1) process emission reduction due significant energy hindrance effect industrial structure effect, while growth constrained by enhancement technology progress labor participation effect. (2) moderating effects technological innovation 10.74% 10.24%, respectively, force relatively weak. (3) goodness fit MNGM-ARIMA MNGM-BPNN over 97%. Carbon high regions show decreasing trend all scenarios except consumption-driven scenario, medium low decrease slowly only R&D- supply-driven This advocates developing heterogeneous measures based degree aging, accelerating supply side upgrading, increasing proportion green consumption.

Язык: Английский

Процитировано

2
Electricity Demand, Forecasting the Peaks: Development and Implementation of C-EVA Method DOI Creative Commons
Petros Theodorou, Demetris Christopoulos

Green and Low-Carbon Economy, Год журнала: 2024, Номер 2(4), С. 310 - 324

Опубликована: Май 29, 2024

Price spikes in electricity markets are very frequent, posing tremendous burden on household income and manufacturing cost. Electricity demand (load) can be divided two parts, energy (MWh) peak (MW) most of time is responsible for the price spikes. Literature review while devoting discussion to energy, lags investigation peak. In this research, a model analysis forecasting developed. The based portfolio cluster extreme value (C-EVA) methods using unit invariant knee, extremum distance estimator, weighted scale load innovations optimal determination clusters daily peaks divulgence. C-EVA method consists Clustering part number classification day month peak, Extreme Value Analysis computation statistical confidence interval maxima. after all currently available maxima, estimates statistically expected worst-case scenario loads. Load will determined by EVA an estimated bimodal distribution signaling prompt probability extremes. added proposed that does not reject values as methodologies do. maxima minima provide estimators highest lowest hourly load, giving return level optimization selection rolling window, period. It was found distributed generation renewables create camel effect which increases sharpness. methodology solved issue opening ground future research role storage, batteries well virtual power plants integrated generation. Received: 18 December 2023 | Revised: February 2024 Accepted: 19 May Conflicts Interest authors declare they have no conflicts interest work. Data Availability Statement database supports findings study made upon request only specific Excel format. Author Contribution Petros Theodorou Demetris Theodoros Christopoulos: Conceptualization, Methodology, Software, Validation, Formal analysis, Investigation, Resources, curation, Writing - original draft, & editing, Visualization, Supervision, Project administration.

Язык: Английский

Процитировано

1
Crude oil price forecasting with multivariate selection, machine learning, and a nonlinear combination strategy DOI
Yan Xu,

Tianli Liu,

Fang Qi

и другие.

Engineering Applications of Artificial Intelligence, Год журнала: 2024, Номер 139, С. 109510 - 109510

Опубликована: Окт. 31, 2024

Язык: Английский

Процитировано

1
The Degree of Population Aging and Carbon Emissions: Analysis of Mediation Effect and Multi-Scenario Simulation DOI
Shuyu Li, Shun Jia, Yang Liu

и другие.

Опубликована: Янв. 1, 2024

The continuous deepening of aging has posed new challenges for global sustainable development. Measuring the impact population on carbon emissions is crucial next stage climate governance. However, structural changes in social production and consumption make it difficult to evaluate effects. Therefore, this study constructed a bidirectional fixed Space Durbin Model explore mediating pathway aging's emissions. Furthermore, we have established high-precision prediction models simulate evolution trajectory under multi-factor driving scenarios. main findings are as follows: (1) process emission reduction due significant energy hindrance effect industrial structure effect, while growth constrained by enhancement technology progress labor participation effect. (2) moderating effects technological innovation 10.74% 10.24%, respectively, force relatively weak. (3) goodness fit MNGM-ARIMA MNGM-BPNN over 97%. Carbon high regions show decreasing trend all scenarios except consumption-driven scenario, medium low decrease slowly only R&D- supply-driven This advocates developing heterogeneous measures based degree aging, accelerating supply side upgrading, increasing proportion green consumption.

Язык: Английский

Процитировано

0
Research on Ultra-short-term combination forecasting algorithm of power load based on machine learning DOI Open Access

Jinggeng Gao,

Kun Wang, Xiaohua Kang

и другие.

Journal of Physics Conference Series, Год журнала: 2024, Номер 2846(1), С. 012046 - 012046

Опубликована: Сен. 1, 2024

Abstract Power load forecasting is of great significance to the power grid marketing department. To obtain accurate results, a minute-by-minute method for electricity based on multi-stage proposed (TPE-WXL) by combining non-linear and time-series attributes. Firstly, historical series specific areas in city are pre-processed. Then, order accurately predicted XGBoost LightGBM applied extract attributes from build hybrid model. Moreover, TPE introduced enhance hyperparameters model series. Finally, dataset region used as an example conduct experimental analysis. Experimental results revealed that can forecast trend load, is, R 2 =0.981, RMSE =2.643.

Язык: Английский

Процитировано

0