A comparative study of different deep learning methods for time-series probabilistic residential load power forecasting DOI Creative Commons

Liangcai Zhou,

Yi Zhou, Linlin Liu

et al.

Frontiers in Energy Research, Journal Year: 2024, Volume and Issue: 12

Published: Oct. 18, 2024

The widespread adoption of nonlinear power electronic devices in residential settings has significantly increased the stochasticity and uncertainty systems. original load data, characterized by numerous irregular, random, probabilistic components, adversely impacts predictive performance deep learning techniques, particularly neural networks. To address this challenge, paper proposes a time-series prediction technique based on mature network point technique, i.e., decomposing data into deterministic stochastic components. component is predicted using technology, fitted with Gaussian mixture distribution model parameters are great expectation algorithm, after which obtained generation method. Using study evaluates six different methods to forecast power. By comparing errors these methods, optimal identified, leading substantial improvement accuracy.

Language: Английский

Xu Weight is All that Models Need! A Short-Term Power Load Forecasting Method Based on a Novel Adaptive Feature Selection Method and Xu Weight DOI

Jingqi Xu,

Xueman Wang,

Hui Hou

et al.

Published: Jan. 1, 2024

Language: Английский

Citations

0
A comparative study of different deep learning methods for time-series probabilistic residential load power forecasting DOI Creative Commons

Liangcai Zhou,

Yi Zhou, Linlin Liu

et al.

Frontiers in Energy Research, Journal Year: 2024, Volume and Issue: 12

Published: Oct. 18, 2024

The widespread adoption of nonlinear power electronic devices in residential settings has significantly increased the stochasticity and uncertainty systems. original load data, characterized by numerous irregular, random, probabilistic components, adversely impacts predictive performance deep learning techniques, particularly neural networks. To address this challenge, paper proposes a time-series prediction technique based on mature network point technique, i.e., decomposing data into deterministic stochastic components. component is predicted using technology, fitted with Gaussian mixture distribution model parameters are great expectation algorithm, after which obtained generation method. Using study evaluates six different methods to forecast power. By comparing errors these methods, optimal identified, leading substantial improvement accuracy.

Language: Английский

Citations

0