Power prediction for generator based on Transformer-LSTM and oil wear status
- 2 April 2026
- journal article
- Published by SAGE Publications in Proceedings of the Institution of Mechanical Engineers, Part J: Journal of Engineering Tribology
- Vol. 240 (7) , 1334-1343
Abstract
Driven by the escalating requirements for efficiency and operational safety in power-generation units, data-driven forecasting of generator output has emerged as a critical research frontier. Generator performance is governed by a complex interplay between mechanical operating conditions and stochastic human factors. In this study, we performed online monitoring of turbine lubricating oil to assemble a comprehensive time-series dataset integrating oil wear signatures and unit power metrics. We propose a synergistic deep-learning architecture, the Transformer–LSTM, which leverages an LSTM-based auxiliary extractor to capture fine-grained local temporal patterns. Concurrently, positional encodings, timestamp embeddings, and value embeddings are employed to characterize global contextual features. By fusing these local and global representations, the model achieves a joint perception of short-term dynamics and long-range dependencies. Empirical results demonstrate that the proposed hybrid model significantly outperforms standalone LSTM and Transformer baselines across diverse evaluation metrics in terms of both predictive accuracy and robustness. This framework provides robust technical decision support for condition-based maintenance and optimized power-generation scheduling.Keywords
Funding Information
- National Natural Science Foundation of China (52571368)
This publication has 32 references indexed in Scilit:
- Short-Term Electricity-Load Forecasting by deep learning: A comprehensive surveyEngineering Applications of Artificial Intelligence, 2025
- Deep learning for time series forecasting: a surveyInternational Journal of Machine Learning and Cybernetics, 2025
- Local and global trend Bayesian exponential smoothing modelsInternational Journal of Forecasting, 2024
- NOA-LSTM: An efficient LSTM cell architecture for time series forecastingExpert Systems with Applications, 2023
- Review of automated time series forecasting pipelinesWIREs Data Mining and Knowledge Discovery, 2022
- District heater load forecasting based on machine learning and parallel CNN-LSTM attentionEnergy, 2022
- Transductive LSTM for time-series prediction: An application to weather forecastingNeural Networks, 2020
- The M4 Competition: 100,000 time series and 61 forecasting methodsInternational Journal of Forecasting, 2020
- Predicting residential energy consumption using CNN-LSTM neural networksEnergy, 2019
- Short-term electricity demand forecasting with MARS, SVR and ARIMA models using aggregated demand data in Queensland, AustraliaAdvanced Engineering Informatics, 2018