A hybrid time series forecasting approach integrating fuzzy clustering and machine learning for enhanced power consumption prediction

Scientific Reports, Год журнала: 2025, Номер 15(1)

Опубликована: Фев. 22, 2025

Power demand estimation in Tetouan, Morocco, uses fuzzy clustering with machine learning-based time series forecasting models as the main subject of research. This paper tackles an important requirement for methods that accurately predict electricity use areas changing to enhance energy management capabilities. An evaluation 52,417 records containing six characteristics derived from three power networks formed basis this analysis. A comparison Random Forest, Support Vector Machine, K-Nearest Neighbors, Extreme Gradient Boosting, and Multilayer Perceptron took place through Root Mean Square Error, Absolute R² metric evaluation. Model performance improved after integration, resulting multilayer perceptron achieving its best results RMSE at 355.42, MAE 246.43, 0.9889. The hybrid approach is original practical solution improves accuracy consumption.

Язык: Английский

Unlocking renewable energy potential: Overcoming knowledge sharing hurdles in rural EU regions on example of poland, sweden and france

PLoS ONE, Год журнала: 2025, Номер 20(4), С. e0320965 - e0320965

Опубликована: Апрель 10, 2025

The optimal technological choice for sustainable development lies in renewable energy sources (RES). However, the potential offered by RES utilization poses significant challenges mobile technologies and everyday living. Despite extensive research information highlighting benefits of energy, there remains considerable debate, limited awareness persists. advantages are not fully comprehended, raising concerns about its consistent application. Regrettably, lack knowledge a fundamental understanding hinders effective dissemination. To gauge attitudes residents regions where is employed, this study employed questionnaire authored researcher. was conducted between June 2022 January 2023, with total 12,428 participants completing survey. sampling method utilized an online form distributed via various social media channels among local contacts authors Poland, Sweden, France. Gender allocation: 58% male 42% female. Respondents shared their perspectives on ecology disclosed familiarity utilization. Results indicate public skepticism regarding adequacy security measures level use. Insufficient experts, advocacy, reliance contribute to low awareness. In several EU countries, absence widely accepted easily accessible (RES) sharing adoption. EU’s efforts promote through directives subsidies, rural communities these countries often adequate education technologies. This gap contributes unfavorable perceptions, some viewing renewables as unreliable or economically unfeasible options compared traditional like coal natural gas. Additionally, bureaucratic hurdles inconsistent government policies further complicate transition discouraging investment innovation sector. As result, while aims future, barriers impede widespread growth hinder progress towards climate targets. Poland found that 76% respondents expressed favorable perceptions RES, indicating general inclination adopting clean solutions. analysis uncovered high environmental participants, 85% expressing concern degradation. awareness, 62% reported reservations affordability derived from sources. 48% uncertainty ambivalence RES. France, revealed similar energy. 59% sources, 53% cited perceived costs barrier Furthermore, 41% identified underdeveloped infrastructure hindrance wider acceptance quantitative data highlights complex landscape While issues positive solutions, security, affordability, remain These findings underscore importance targeted interventions educational address practices across Europe. Renewable represent critical avenue development, offering pathway mitigate degradation reduce dependence fossil fuels. investigates attitudes, levels, adoption areas unique socio-economic cultural factors influencing regions. Conducted survey, gathering responses countries. evaluated statements responsibility, application, obstacles, using five-point Likert scale. Key reveal high, persist knowledge, infrastructure, associated view but cost security. Swedish demonstrated strong (85%), yet voiced reliability. French similarly highlighted costs, identifying systems primary hindrance. underscores campaigns policy bridge gaps foster greater Tailored strategies addressing barriers—such financial incentives, community-based investments—are essential overcoming challenges. By exploring diverse three valuable insights broader discourse transitions EU.

Язык: Английский

Predicting Residential Energy Consumption in South Africa Using Ensemble Models

Applied Computational Intelligence and Soft Computing, Год журнала: 2025, Номер 2025(1)

Опубликована: Янв. 1, 2025

This study presents ensemble machine learning (ML) models for predicting residential energy consumption in South Africa. By combining the best features of individual ML models, reduce drawbacks each model and improve prediction accuracy. We present four models: by averaging (EA), stacking estimator (ESE), boosting (EB), voting (EVE). These are built on top Random Forest (RF) Decision Tree (DT). base predictor leverage historical patterns to capture temporal intricacies, including seasonal variations rolling averages. In addition, we employed feature engineering methodologies further enhance their predictive abilities. The accuracy was evaluated assessing various performance indicators, mean squared error (MSE), absolute (MAE), percentage (MAPE), coefficient determination R 2 . Overall, findings illustrate efficiency providing accurate predictions consumption. provides valuable insights researchers practitioners buildings benefits using building research domains.

Язык: Английский