Model Predictive Control for room temperature control in buildings is an effective approach to energy management in buildings. However, the development and maintenance of physical models may be a bottleneck for widespread real life application. Data Predictive Control is an attempt to address this problem by learning the behaviour of the building from historical data and thus reducing the modelling effort. Here, we present an application of a Data Predictive Control approach, based on Random Forests with affine functions and convex optimization, to control the room temperature in a real life apartment. When compared to a conventional hysteresis controller, the applied approach saves 24.9 % of cooling energy while reducing the integral of comfort constraint violations by 72.0 % in a six-day experiment.

Bidirectional low temperature networks are a novel concept that promises more efficient heating and cooling of buildings. Early research shows theoretical benefits in terms of exergy efficiency over other technologies. Pilot projects indicate that the concept delivers good performance if heating and cooling demands are diverse. However, the operation of these networks is not yet optimized and there is no quantification of the benefits over other technologies in various scenarios. Moreover, there is a lack of understanding of how to integrate and control multiple distributed heat and cold sources in such networks. Therefore, this paper develops a control concept based on a temperature set point optimization and agent-based control which allows the modular integration of an arbitrary number of sources and consumers. Afterwards, the concept is applied to two scenarios representing neighborhoods in San Francisco and Cologne with different heating and cooling demands and boundary conditions. The performance of the system is then compared to other state-of-theart heating and cooling solutions using dynamic simulations with Modelica. The results show that bidirectional low temperature networks without optimization produce 26% less emissions in the San Francisco scenario and 63% in the Cologne scenario in comparison to the other heating and cooling solutions. Savings of energy costs are 46% and 27%, and reductions of primary energy consumption 52% and 72%, respectively. The presented operation optimization leads to electricity use reductions of 13% and 41% when compared to networks with free-floating temperature control and the results indicate further potential for improvement. The study demonstrates the advantage of low temperature networks in different situations and introduces a control concept that is extendable for real implementation.

To reduce the heating and cooling energy demand of buildings and districts novel control strategies are constantly being developed that require information on the future demand of the controlled entity. Demand forecasting is commonly done with deterministic white box models or fitted grey-box models, however, recently more and more data and machine learning based approaches are being developed. All approaches have weaknesses: white-box models require major modelling effort, grey-box approaches are limited by their model or parameter complexity and machine learning is dependent on hyperparameters, some of which are randomly chosen, and therefore considered unreliable. Here we develop a forecasting approach based on Artificial Neural Networks (ANN) and introduce error correction methods based on online learning and the learned autocorrelation of the forecasting error. We compare the approach to other regression based and grey-box methods in a real case study of a small-scale district energy system with mixed use and unknown lower-level control. We show that the proposed method outperforms the other forecasting methods in terms of average error and coefficient of determination. We further demonstrate that in our case study the error correction methods significantly reduce variance in ANN performance created by randomly initialized parameters in the networks.