Cross-household Transfer Learning Approach with LSTM-based Demand Forecasting

With the rapid increase in residential heat pump (HP) installations, optimizing hot water production in households is essential, yet it is faced with major technical and scalability challenges. Adapting production to actual household needs necessitates accurate forecasting of hot water demand to ensure comfort and, most importantly, to reduce energy waste. However, the conventional approach of training separate machine learning models for each household becomes computationally expensive at scale, particularly in cloud-connected HP deployments.This study introduces DELTAiF, a transfer learning (TL)-based framework that offers scalable and accurate prediction of households’ consumption of hot water. By predicting large usage of hot water, such as in showers, DELTAiF enables adaptive yet scalable hot water production at the household level. DELTAiF leverages learned knowledge from a representative household and fine-tunes it across others, eliminating the need to train separate machine learning models for each HP installation. This approach reduces overall training time by approximately 67% while maintaining high predictive accuracy values between 0.874–0.991 and mean absolute percentage error values of 0.001-0.017. The results show that TL is particularly effective when the source household exhibits regular consumption patterns, hence enabling hot water demand forecasting at scale.