A multi-task learning model for building electrical load prediction

Liu, Chien‐Liang; Tseng, Chun-Jan; Huang, T. C.; Yang, Jie-Si; Huang, Kaibin

doi:10.1016/j.enbuild.2022.112601

Cited by 15 publications

(6 citation statements)

References 41 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The action set a is the set of weight change rates of RNN, LSTM, and GRU in the Q-VMD-RLG model, the formula is The reward set of VMD-RLG is the absolute value of the RMSE of the ensemble model after the RNN, LSTM, and GRU weights are updated, as shown in Equation (15).…”

Section: Markov Dynamic Decision Process Of Q-vmd-rlg Modelmentioning

confidence: 99%

A novel ensemble deep reinforcement learning model for short‐term load forecasting based on Q‐learning dynamic model selection

He,

Zhao,

Zhang

et al. 2024

The Journal of Engineering

View full text Add to dashboard Cite

Short‐term load forecasting is critical for power system planning and operations, and ensemble forecasting methods for electricity loads have been shown to be effective in obtaining accurate forecasts. However, the weights in ensemble prediction models are usually preset based on the overall performance after training, which prevents the model from adapting in the face of different scenarios, limiting the improvement of prediction performance. In order to improve the accurateness and validity of the ensemble prediction method further, this paper proposes an ensemble deep reinforcement learning approach using Q‐learning dynamic weight assignment to consider local behaviours caused by changes in the external environment. Firstly, the variational mode decomposition is used to reduce the non‐stationarity of the original data by decomposing the load sequence. Then, the recurrent neural network (RNN), long short‐term memory (LSTM), and gated recurrent unit (GRU) are selected as the basic power load predictors. Finally, the optimal weights are ensembled for the three sub‐predictors by the optimal weights generated using the Q‐learning algorithm, and the final results are obtained by combining their respective predictions. The results show that the forecasting capability of the proposed method outperforms all sub‐models and several baseline ensemble forecasting methods.

show abstract

Section: Markov Dynamic Decision Process Of Q-vmd-rlg Modelmentioning

confidence: 99%

A novel ensemble deep reinforcement learning model for short‐term load forecasting based on Q‐learning dynamic model selection

He,

Zhao,

Zhang

et al. 2024

The Journal of Engineering

View full text Add to dashboard Cite

show abstract

“…In another study, ref. [31] utilized MTL for electricity load forecasting, while the auxiliary task was to predict the outdoor temperature. The authors of [32] provide a comprehensive survey of MTL methodologies, underlining their effectiveness across different domains, including time series forecasting.…”

Section: Multi-task Learningmentioning

confidence: 99%

Multi-Task Learning and Temporal-Fusion-Transformer-Based Forecasting of Building Power Consumption

Ji,

Cao,

2023

Electronics

View full text Add to dashboard Cite

Improving the accuracy of the forecasting of building power consumption is helpful in reducing commercial expenses and carbon emissions. However, challenges such as the shortage of training data and the absence of efficient models are the main obstacles in this field. To address these issues, this work introduces a model named MTLTFT, combining multi-task learning (MTL) with the temporal fusion transformer (TFT). The MTL approach is utilized to maximize the effectiveness of the limited data by introducing multiple related forecasting tasks. This method enhances the learning process by enabling the model to learn shared representations across different tasks, although the physical number of data remains unchanged. The TFT component, which is optimized for feature learning, is integrated to further improve the model’s performance. Based on a dataset from a large exposition building in Hangzhou, we conducted several forecasting experiments. The results demonstrate that MTLTFT outperforms most baseline methods (such as LSTM, GRU, N-HiTS) in terms of Root Mean Squared Error (RMSE) and Mean Absolute Percentage Error (MAPE), suggesting that MTLTFT is a promising approach for the forecasting of building power consumption and other similar tasks.

show abstract

“…The paper of Chen-Liang Liu [26] mentions that when training the relationship between power load and weather data, the hyperparameter c of auxiliary weights is introduced to fit the auxiliary regression, which turns the single-task regression into a multi-task one. The regression effect of the model is optimized by 21% compared with the single regression.…”

Section: Introduction Of Machine Learningmentioning

confidence: 99%

Research on Optimized Design of Rural Housing in Cold Regions Based on Parametrization and Machine Learning

Sun,

Xue,

Wang

2024

Sustainability

View full text Add to dashboard Cite

With the rapid development of urbanization, the emergence of more self-built buildings in the countryside has brought about energy waste problems and decreased comfort. Achieving the low-carbon goal and improving the quality of the human living environment through architectural and planning means have become vital issues. In this study, from a parametric perspective, model building and performance simulation are carried out using Rhino and Grasshopper, and a multi-objective optimization method and a neural network model are used as the theoretical basis to train the prediction model after data collection and processing. The model validation of R2 = 0.988 and MSE = 0.0148 indicates that the model can accurately reflect the program’s performance. By establishing a rapid prediction model for the performance of rural residential buildings, decision-makers can perform performance predictions under various parameter combinations at the early design stage, facilitating the screening of building types with high energy consumption and costs. The method can improve the efficiency of decision-making at the early stage of design, help save decision-making costs by screening high-energy-consuming building types, improve the living conditions of residents, reduce carbon emissions, and contribute to the sustainable development of residential building renewal design in rural areas.

show abstract

A multi-task learning model for building electrical load prediction

Cited by 15 publications

References 41 publications

A novel ensemble deep reinforcement learning model for short‐term load forecasting based on Q‐learning dynamic model selection

A novel ensemble deep reinforcement learning model for short‐term load forecasting based on Q‐learning dynamic model selection

Multi-Task Learning and Temporal-Fusion-Transformer-Based Forecasting of Building Power Consumption

Research on Optimized Design of Rural Housing in Cold Regions Based on Parametrization and Machine Learning

Contact Info

Product

Resources

About