David Wölfle | Searcharxiv

arXiv2025

Open Energy Services -- Forecasting and Optimization as a Service for Energy Management Applications at Scale

This article aims at facilitating the widespread application of Energy Management Systems (EMSs), especially on buildings and cities, in order to support the realization of future carbon-neutral energy systems. We claim that economic viability is a severe issue for the utilization of EMSs at scale and that the provisioning of forecasting and optimization algorithms as a service can make a major contribution to achieve it. To this end, we present the \emph{Energy Service Generics} software framework that allows the derivation of fully functional services from existing forecasting or optimization code with ease. This work documents the strictly systematic development of the framework, beginning with a requirement analysis, from which a sophisticated design concept is derived, followed by a description of the implementation of the framework. Furthermore, we present the concept of the \emph{Open Energy Service} community, our effort to continuously maintain the service framework but also provide ready-to-use forecasting and optimization services. Finally, an evaluation of our framework and community concept, as well as a demarcation between our work and the current state of the art, is presented.

cs.SE↗

arXiv2022

Which Model to Trust: Assessing the Influence of Models on the Performance of Reinforcement Learning Algorithms for Continuous Control Tasks

The need for algorithms able to solve Reinforcement Learning (RL) problems with few trials has motivated the advent of model-based RL methods. The reported performance of model-based algorithms has dramatically increased within recent years. However, it is not clear how much of the recent progress is due to improved algorithms or due to improved models. While different modeling options are available to choose from when applying a model-based approach, the distinguishing traits and particular strengths of different models are not clear. The main contribution of this work lies precisely in assessing the model influence on the performance of RL algorithms. A set of commonly adopted models is established for the purpose of model comparison. These include Neural Networks (NNs), ensembles of NNs, two different approximations of Bayesian NNs (BNNs), that is, the Concrete Dropout NN and the Anchored Ensembling, and Gaussian Processes (GPs). The model comparison is evaluated on a suite of continuous control benchmarking tasks. Our results reveal that significant differences in model performance do exist. The Concrete Dropout NN reports persistently superior performance. We summarize these differences for the benefit of the modeler and suggest that the model choice is tailored to the standards required by each specific application.

cs.LG↗