The potential of deep learning to reduce complexity in energy system modeling

Abstract

Summary In order to cope with increasing complexity in energy systems due to rapid changes and uncertain future developments, the evaluation of multiple scenarios is essential for sound scientific system analyses. Hence, efficient modeling approaches and complexity reductions are urgently required. However, there is a lack of scientific analyses going beyond the scope of traditional energy system modeling. For this reason, we investigate the potential of metamodels to reduce the complexity of energy system modeling. In our explorative study, we investigate their potential and limits for applications in the fields of electricity dispatch and design optimization for heating systems. We first select a suitable metamodeling approach by conducting pre‐tests on a small scale. Based on this, we selected artificial neural networks due to their good performance compared to other approaches and the multiple possibilities of network topologies and hyperparameter settings. As for the dispatch model, we show that a high accuracy of price replication can be achieved while substantially reducing the runtimes per investigated scenario (from 2 hours on average down to less than 30 seconds). With the design optimization model, we find double‐edged results: while we also achieve a substantial reduction of runtime in this case (from ~0.8 hours to less than 30 seconds), the simultaneous forecasting of several interdependent variables proved to be problematic and the accuracy of the metamodel shows to be insufficient in many cases. Overall, we demonstrate that metamodeling is a suitable approach to complemement traditional energy system modeling rather than to replace them: the loss of traceability in (black‐box) metamodels indicates the importance of hybrid solutions that combine fundamental models with metamodels.

Accelerating Energy-Economic Simulation Models via Machine Learning-Based Emulation and Time Series Aggregation

“…Also, on model with a multivariate output could have been trained instead. According to [2], this leads to the need for more training data. For this reason these models were disregarded.…”

“…In [2], Köhnen et al further define "hybrid meta models" as meta-models that are trained based on a simulation or optimization model, rather than on real data.…”

“…Meta-models ("a model of models") or surrogate models substitute the entire simulation model, e.g., by ML-based regression or classical polynomial functions [2]. The schematic difference of simulation, emulation, and surrogate/meta-models is shown in Figure 1.…”

Metaheuristics‐optimized deep learning to predict generation of sustainable energy from rooftop plant microbial fuel cells

“…Conventional methods of processing noise include dimensionality reduction and feature extraction, which can result in the loss of valuable information. The use of deep learning techniques enables the effective learning of data features, 30 from which internal patterns and important attributes are identified 31 . Such techniques are more accurate than conventional forecasting models 32 …”

Exploring Grassroots Renewable Energy Transitions: Developing a Community-Scale Energy Model

“…MLP models are very flexible in terms of architecture, size, and hyperparameter selection. As such, they are good candidates for energy system model applications [33]. However, particularly complex data relationships (typically image recognition) can require neural nets that are very deep for good accuracy [34], and these additional layers can make training difficult with traditional MLPs [32].…”