Artificial Intelligence for Heat Pumps

The aim of the "AI4HP" project was to develop a new generation of "smart heat pumps" that use artificial neural networks (ANNs) to adapt to changing operating conditions, thereby increasing energy efficiency whilst maintaining user comfort.

Initial Situation

Heat pumps represent an effective solution for reducing the energy consumption and environmental impact of buildings and for integrating renewable energy into the heating supply. To date, however, the control of heat pumps in residential buildings has been implemented using very simple heuristic methods that take into account neither actual user needs nor the prediction of external influences, including changes in user habits or the ageing or renovation of the building. The actual efficiency of heat pumps therefore does not always meet expectations in practice. 

Objective

The aim of this joint project between German and French experts in the fields of heat pumps and energy supply, as well as AI research, was to develop innovative artificial intelligence (AI) methods based on incremental learning using artificial neural networks (ANNs) for the adaptive control and monitoring of heat pumps. For the first time, these new intelligent, AI-supported heat pumps integrate new functionalities and interactions with a changing environment to offer users maximum energy efficiency and optimum comfort, simplify maintenance work and prevent performance losses through fault detection.

Approach

AI methods must become more robust and scalable to be implemented cost-effectively in a large number of heterogeneous buildings, where data often originates from sensors with low precision and reliability. Ultimately, only reliable and trustworthy methods that ensure safe operation will gain acceptance among heat pump manufacturers and their customers. The first part of the project therefore focused on the development of novel machine learning methods from the field of incremental and online learning for time-series data, which are necessary to ensure that KNNs learn autonomously and adaptively in a changing environment without having to store the complete set of past data or forget past knowledge. The adaptive AI pipeline has being developed for the three use cases "adaptive heating curve control", "adaptive control of domestic hot water heat pumps based on load forecasts" and "adaptive fault detection and diagnosis", integrated into heat pump control systems, and validated in laboratory tests and pilot buildings. By using advanced AI methods, the objective was to reach energy savings of up to 20% and a reduction in CO2 emissions without compromising on comfort.

Results

The project has resulted in an adaptive AI optimisation algorithm that adjusts the heating curve of heat pumps to actual demand, thereby increasing the energy efficiency of the heat pump. By integrating the intelligent controller, heat pump manufacturers can significantly enhance their products: practical tests show an increase in energy efficiency up to 10%  a 50% reduction in comfort deviations.