Almost half of the energy consumed globally is under the form of heat, produced mainly through fossil fuels. Switching to using renewable energy instead is a real challenge. Combining renewable thermal energy with thermal storage is a complex system to operate. To harness the full potential of thermal plants, advanced control strategies need to be implemented. Dynamic real-time optimization (DRTO) seems promising to fine tune controller setpoints of plants. The goal of our study is to ultimately enable DRTO by using Optimica because of its ease of use and Modelica’s modularity. This paper presents a Modelica library developed to first perform offline dynamic optimization with Optimica, and would ultimately be used in a DRTO strategy. The library enables to model a renewable thermal plant composed of solar thermals, heat pumps and thermal storages. The model of each subcomponent has been validated. Initial dynamic optimizations of plant operation give promising results.

In this work, we adress challenges associated with multi parameter calibration of complex system models of high computational expense.

We propose to replace the Modelica Model for screening of parameter space by a computational effective Machine-Learning Surrogate, followed by a polishing with a gradient-based optimizer coupled to the Modelica Model.

Our results show the superiority of this approach compared to common-used optimization strategies. We can resign on determining initial optimization values while using a small number of Modelica model calls, paving the path towards efficient global optimization. The Machine Learning Surrogate, namely a Physics Enhanced Latent Space Variational Autoencoder (PELS-VAE), is able to capture the impact of most influential parameters on small training sets and delivers sufficiently good starting values to the gradient-based optimizer.

In order to make this paper self-contained, we give a sound overview to the necessary theory, namely Global Sensitivity Analysis with Sobol Indices and Variational Autoencoders.

Strict operating boundaries on commercial lithium ion cells are defined to mitigate the effect of aging (loss of capacity and increase in internal resistance), as well as avoiding safety hazards, like the appearance of lithium plating during charge, which can lead to internal short circuit and subsequent thermal runnaway. Therefore, to develop fast charge algorithms that maximize charging speeds, electrochemical and thermal constraints must be considered. Most studies so far have focused on the single cell problem, whereas pack-level fast charge challenge has been tackled directly by the industry. The reason is that the temperature difference between cells within a battery pack is often considered small, and therefore that optimal charging profiles can be extrapolated from single cell investigations. In practice, temperature spread can reach up to 10 K from coldest to warmest points in the pack, and at least 5 K between same position of different cells. With this in mind, a Nonlinear Model Predictive Control (NLMPC) scheme is proposed that considers both electrochemical and thermal constraints at pack level, establishing, at least on a theoretical basis, the practical limits of fast charge. We demonstrate how active thermal management, i.e., controlling the fluid inlet temperature, is critical to reducing charging times below 40 min (from 0% to 80% state of charge), and discuss some challenges when using online optimization-based control techniques

Domain-overarching system models are crucial to investigate sector coupling concepts. Specifically, the coupling of building and electrical energy systems becomes crucial to integrate renewable energy sources such as photovoltaic power systems (PV). For such interdisciplinary simulation models, Modelica is a suitable language. However, most open-source Modelica libraries are either domain-specific or lack simple-to-parameterize PV models. We close this gap by developing a PV model for the IBPSA Modelica library. The model comprises two I-V-characteristic models and three mounting-dependent approaches to calculate the cell temperature. The I-V-characteristic models follow a single- and two-diodes approach. This study uses measurement data from a rooftop PV system in Berlin, Germany, for validation. The focus lies on comparing the implemented single- and two-diodes approach. Results prove that both models accurately calculate the modules’ DC power output and cell temperature. The two-diodes approach slightly outperforms the single-diode one at the expense of a higher parameterization effort.