Scholarlink

<p>This paper investigates power consumption prediction using two machine learning models, namely Naive Bayes Regression (NBR) and Stochastic Gradient Boosting Regression (SGBR), and employs the Osprey Optimization Algorithm (OOA) for this purpose. Preprocessing data, choosing features, training models, and evaluating them are all steps in the machine learning process. For a number of reasons, power consumption prediction is essential. More importantly, appropriate forecasts allow energy providers to be in a better position to coordinate transmission, distribution, and production, which further improves resource allocation and reduces operation costs. This balance between supply and demand decreases the possibility of blackouts or shortages and maintains grid stability. This is also instrumental in supporting energy efficiency through encouraging sustainability and environmental protection, in that they empower consumers to change their patterns of consumption. A higher <InlineEquation ID="IEq1"> <InlineMediaObject> <ImageObject Color="BlackWhite" FileRef="12046_2025_2833_Article_IEq1.gif" Format="GIF" Height="17" Rendition="HTML" Resolution="72" Type="Linedraw" Width="19" /> </InlineMediaObject> <EquationSource Format="TEX">\({\text{R}}^{2}\)</EquationSource> <EquationSource Format="MATHML"><math> <msup> <mrow> <mtext>R</mtext> </mrow> <mn>2</mn> </msup> </math></EquationSource> </InlineEquation> value indicates that the hybrid models outperform the single models. In the test phase, the OOA_STOCR model outperforms the STOCR single model by 7.14%, while performance increase for the OOA_XGBR model from the XGBR single model is 6.84%. According to the U95 measure, a comparison shows the performance of the OOA_XGBR model at 1449.823 against 1931.838 for the OOA_STOCR model. These results, with higher predicted accuracy with reliability, further demonstrate the inclusion of machine learning techniques into the Osprey Optimization Algorithm in power consumption prediction. This work will enhance the energy forecasting approaches and their implications toward the development of sustainability and energy management systems.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Forecasting future power via machine learning models for predicting energy consumption

  • 期刊论文

  • 发表日期：

    2025年9月2日

查看全文
  • Qing Hu,
  • Chonglin Fan,
  • Dayong Guo

摘要

This paper investigates power consumption prediction using two machine learning models, namely Naive Bayes Regression (NBR) and Stochastic Gradient Boosting Regression (SGBR), and employs the Osprey Optimization Algorithm (OOA) for this purpose. Preprocessing data, choosing features, training models, and evaluating them are all steps in the machine learning process. For a number of reasons, power consumption prediction is essential. More importantly, appropriate forecasts allow energy providers to be in a better position to coordinate transmission, distribution, and production, which further improves resource allocation and reduces operation costs. This balance between supply and demand decreases the possibility of blackouts or shortages and maintains grid stability. This is also instrumental in supporting energy efficiency through encouraging sustainability and environmental protection, in that they empower consumers to change their patterns of consumption. A higher \({\text{R}}^{2}\) R 2 value indicates that the hybrid models outperform the single models. In the test phase, the OOA_STOCR model outperforms the STOCR single model by 7.14%, while performance increase for the OOA_XGBR model from the XGBR single model is 6.84%. According to the U95 measure, a comparison shows the performance of the OOA_XGBR model at 1449.823 against 1931.838 for the OOA_STOCR model. These results, with higher predicted accuracy with reliability, further demonstrate the inclusion of machine learning techniques into the Osprey Optimization Algorithm in power consumption prediction. This work will enhance the energy forecasting approaches and their implications toward the development of sustainability and energy management systems.