Stephan Balduin's research while affiliated with Hanse Institut Oldenburg and other places

... The setup of the simulation environment follows the basic architecture of a SIL framework, as explained in [8]. ...

... Communication protocols are generally simulated with event-based simulators, while models of the physics of energy systems are usually based on differential and algebraic equations solved with continuous simulation [34]. Models for markets and market aggregators also require different paradigms for complex calculations and optimization, including machine learning [35]. Combining all these so-called domainspecific models with their different paradigms to simulate the entire cellular energy system necessitates co-simulation. ...

... While the former only considers the final, i.e., steady, state after the system's parameters change, transient state simulation also models the process, i.e., the transient behavior, between these steady states. Transient state simulation thus offers more insights at the cost of higher complexity for model parameters and calculations [6]. In Table 2, we classify different co-simulation approaches based on these paradigms and summarize to which extent they fulfill the requirements derived in §2.3. ...

... For example, the computational cost of training an support-vector machine (SVM) scales quadratically with the number of samples, rendering it infeasible for large data sets [16]. Likewise, to optimize models, many models are not trained only once, but often dozens of models are trained and compared to find the optimal hyperparameters [17,18]. ...

... Distributed control and optimization systems often involve multiple energy units that decide locally and communicate with each other to solve global problems. For instance, software agents can represent flexible energy loads that cooperatively aggregate flexibility to provide load dispatch options for balancing markets or congestion management [5] [6] [7]. ...

... To address this difficulty, most RL frameworks have considered the use of power system simulation models (e.g., [123]) that aim to simulate the real-world power system response, in the form shown in Figure 4c. Furthermore, the concepts called surrogate models [124] and response surface approximation have recently been intensively used in the context of operational sophistication of power systems of various scales [125][126][127] to simulate plausible system responses without access to detailed information on physical characteristics and still allow for fast and numerous trial evaluations. Hence, the physical model-free RL scheme [110] introducing a surrogate model of the power system, as shown in Figure 4d, is expected to be further sophisticated as an important approach to consider dynamic and computationally cost-effective optimization under various conditions, taking into account the real-world power flows that dynamically change according to the DER penetration phase. ...