A new paradigm based on Wasserstein Generative Adversarial Network and time-series graph for integrated energy system forecasting

Energy Conversion and Management, Год журнала: 2025, Номер 326, С. 119484 - 119484

Опубликована: Янв. 13, 2025

Язык: Английский

---

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

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

Опубликована: Янв. 2, 2025

Язык: Английский

---

An innovative lost circulation forecasting framework utilizing multivariate feature trend analysis

Physics of Fluids, Год журнала: 2025, Номер 37(2)

Опубликована: Фев. 1, 2025

The prompt and precise prediction of lost circulation is essential for safeguarding the security drilling operations in field. This study introduces a model convolutional neural networks-long short-term memory-feature-time graph attention network-transformer (CL-FTGTR) that combines improved complete ensemble empirical mode decomposition with adaptive noise (ICEEMDAN) data trend reconstruction. A notable feature this utilization an innovative logging analysis technique processing fluid engineering parameters, synthesis two consecutive encoding modules: Feature-GAN-transformer (FGTR) time-GAN-transformer (TGTR). Experimental results confirm following: ① ICEEMDAN algorithm can effectively filter out extract components, minimizing impact on outcomes. ② Convolutional memory (CLSTM) position module, substituting traditional sin-cos encoding, significantly improves model's ability to encapsulate global information within input data. ③ FGTR TGTR modules are capable efficiently handling time dimension data, leading significant enhancement performance model. CL-FTGTR was experimentally tested across four wells same block, essentiality its confirmed by five metrics. attained peak precision, recall, F1PA%K, area under curve values 0.908, 0.948, 0.967, 0.927, respectively. findings demonstrate predicting boasts high precision dependability.

Язык: Английский

---

GenFormer: a deep-learning-based approach for generating multivariate stochastic processes

Data-Centric Engineering, Год журнала: 2025, Номер 6

Опубликована: Янв. 1, 2025

Abstract Stochastic generators are essential to produce synthetic realizations that preserve target statistical properties. We propose GenFormer, a stochastic generator for spatio-temporal multivariate processes. It is constructed using Transformer-based deep learning model learns mapping between Markov state sequence and time series values. The data generated by the GenFormer marginal distributions approximately capture other desired properties even in challenging applications involving large number of spatial locations long simulation horizon. applied simulate wind speed at various stations Florida calculate exceedance probabilities risk management.

Язык: Английский

---

A novel probabilistic carbon price prediction model: Integrating the transformer framework with mixed-frequency modeling at different quartiles

Applied Energy, Год журнала: 2025, Номер 391, С. 125951 - 125951

Опубликована: Апрель 21, 2025

Язык: Английский

---

Risk of groundwater depletion in Jaipur district, India: a prediction of groundwater for 2028 using artificial neural network

Rendiconti lincei. Scienze fisiche e naturali, Год журнала: 2024, Номер unknown

Опубликована: Окт. 18, 2024

Язык: Английский

---