Publications Search Results

Now showing 1 - 10 of 77
  • Publication
    Approximating multi-purpose AC optimal power flow with reinforcement trained Artificial Neural Network
    Solving AC-Optimal Power Flow (OPF) problems is an essential task for grid operators to keep the power system safe for the use cases such as minimization of total generation cost or minimization of infeed curtailment from renewable DERs (Distributed Energy Resource). Mathematical solvers are often able to solve the AC-OPF problem but need significant computation time. Artificial neural networks (ANN) have a good application in function approximation with outstanding computational performance. In this paper, we employ ANN to approximate the solution of AC-OPF for multiple purposes. The novelty of our work is a new training method based on the reinforcement learning concept. A high-performance batched power flow solver is used as the physical environment for training, which evaluates an augmented loss function and the numerical action gradient. The augmented loss function consists of the objective term for each use case and the penalty term for constraints violation. This training method enables training without a reference OPF and the integration of discrete decision variable such as discrete transformer tap changer position in the constrained optimization. To improve the optimality of the approximation, we further combine the reinforcement training approach with supervised training labeled by reference OPF. Various benchmark results show the high approximation quality of our proposed approach while achieving high computational efficiency on multiple use cases.
  • Publication
    Soziotechnisches Design eines lernenden Bildverarbeitungssystems für einen Arbeitsplatz im Wareneingang eines Logistikzentrums
    ( 2022) ;
    Braun, Martin
    ;
    Die künstliche Intelligenz erlaubt die Entwicklung lernfähiger, adaptiver Technologien. Ziel einer humanzentrierten Entwicklung ist die schrittweise Adaption von technischen Systemen an das menschliche Verhalten. Assistenzsysteme übernehmen hierbei die Rolle eines Interaktionspartners zwischen künstlicher Intelligenz und arbeitendem Menschen. In diesem Kontext bilden das System of Systems Engineering, das System Thinking und der soziotechnische Ansatz eine bewährte Plattform zum integrierten Design von künstlicher Intelligenz, Arbeits- und Assistenzsystemen. Das Vorgehen wird pilothaft am Beispiel einer lernfähigen Bildverarbeitung im Wareneingang vorgestellt und nach arbeitswissenschaftlichen Kriterien diskutiert.
  • Publication
    Impact of Dynamic Electricity Tariff and Home PV System Incentives on Electric Vehicle Charging Behavior: Study on Potential Grid Implications and Economic Effects for Households
    ( 2022) ; ;
    Al-Khzouz, Hadi
    ;
    ;
    Zhou, Xian
    The rapid increase of electric vehicles (EVs) would lead to a rise in load demand on power grids but create different potential benefits as well. Those benefits comprise EVs serving as a mobile energy storage system to participate in adjusting the load on the power grids and helping manage renewable energy resources. This paper evaluates the effect of dynamic electricity prices and home photovoltaic (PV) system incentives on users EVs charging behavior and potential impacts on grid load and household economy. This has been done by establishing and assessing three different optimized charging configurations and comparing them to an uncontrolled charging strategy. In this study, the charging incentives are applied to a representative sample of 100 households with EVs and PV systems in a metropolitan area. The results show that an optimized charging strategy based on the dynamic electricity tariff can reduce charging costs by 18.5%, while a PV-based optimized strategy can reduce the costs by 33.7%. Moreover, the PV-integrated optimization strategies significantly increase the utilization of PV energy by almost 46% on average, compared to uncontrolled charging. In addition, the simulations of this research have depicted the capability of using home PV systems incentives to smoothen the charging profiles and hence significantly reduce the maximum grid load. However, the electricity price optimization strategy increases the aggregated charging peaks, which can only be slightly reduced by peak shaving. Therefore, an identical price signal for all households might be critical. Further analyses have shown that direct charging occurs simultaneously with household electricity assigned to a specific low-voltage grid while PV and price incentive charging configurations shift the charging peaks away from household load peaks.
  • Publication
    Optimized characteristic-curve-based local reactive power control in distribution grids with distributed generators
    ( 2021)
    Liu, Zheng
    ;
    ;
    Bornhorst, Nils
    ;
    Kraiczy, Markus
    ;
    ;
    Kerber, Tobias
    ;
    Braun, Martin
    The local reactive power control in distribution networks with massive penetration of distributed energy resources (DERs) is essential in future power system operation. An appropriate control characteristic curve for DERs providing reactive power supports stable and efficient distribution network operation. In this paper, an innovative approach to obtain an optimized characteristic curve for DERs is proposed. With the power system analysis tool pandapower [1] and an artificial neural network-based optimization tool, a case study with a real German distribution grid is carried out with quasi-static time-series simulation with fast parallel power flow solver backend [2]. The performance of the new approach is compared to the three standard local reactive control concepts, fixed cos Ï control, Q(P)-control, and Q(V)-control.
  • Publication
    Considering Control Approaches for Electric Vehicle Charging in Grid Planning
    ( 2021)
    Schoen, Andrea
    ;
    ; ;
    Junge, Eric
    ;
    Bott, Christian
    ;
    ;
    In this publication, we introduce a methodology for power system planning that considers grid-friendly electric vehicle (EV) charging, which is developed within the research project ""Ladeinfrastruktur 2.0"" [1]. This publication shows how control strategies for EV charging can be integrated into probabilistic, time-series-based grid planning approaches to determine necessary grid reinforcement and extension measures. Since this method can be computationally expensive, the efficient integration of control strategies into this process is crucial. In order to compare practical and simulated findings, control strategies based on the field test project ""E-Mobility-Allee"" by the German distribution system operator (DSO) Netze BW [2] are applied in simulations. The methodology presented in this paper is developed for this purpose and applied in a case study with a real low voltage (LV) grid. Finally, conclusions regarding the field test and real-life applications are drawn based on the results of the case study, which indicate that the selected control strategies can lead to a reduction of necessary grid reinforcement and extension measures.
  • Publication
    Testing Automated Operation and Control Algorithms for Distribution Grids Using a Co-Simulation Environment
    ( 2021)
    Schoen, Andrea
    ;
    ; ; ;
    Wille-Haussmann, Bernhard
    ;
    Marchand, Sophie
    ;
    ; ;
    This publication presents a co-simulation framework that enables the coupling of various simulation components of energy systems that are not only modeled at three geographically distributed Fraunhofer Institutes, but are also vastly different in terms of their functionality, control algorithms, time resolution, speeds and used tools. This framework facilitates their joint usage despite these differences especially with regard to their time resolutions (real-time combined with non-real-time systems) and is applied in the DistributedGridLab of the Fraunhofer Cluster of Excellence Integrated Energy Systems (CINES) [1]. The DistributedGridLab thus allows users - e.g. grid operators, manufacturers and research institutes - to test their solutions before field deployment with special consideration of their interaction with other solutions at different remote testing facilities, and without the need to use the same software and hardware setups. A demonstrator of the DistributedGridLab is introduced where an electric energy system is modeled, which contains a medium voltage grid with three connected low voltage grids. The different control approaches interact to stabilize the entire system without a centrally controlled simulation of the systems.
  • Publication
    Human Factors and Ergonomics in Digital Manufacturing
    The digital transformation of "the working society introduces a further phase in the management of economic activities.""Digital Manufacturing" continues the concept of "Computer Integrated Manufacturing" announced in the 1970s on the basis of, available, Internet-based technologies. The digital transformation should strengthen the competitiveness of commercial enterprises, while enabling desirable ways of working at the same time. The digital transformation is an opportunity to shed some light on the role of the human individual in digital manufacturing and the business development potential of a human-centered work design. Measures taken to adapt work to people refer to technical-ergonomic conditions and/or to substantive-organizational conditions of the work activity. Immersion, that is, complete immersion in the virtual world by means of head-mounted displays or projections, addresses as many human senses as possible, in order to achieve a realistic simulation. The digital transformation is changing human work and workload patterns.
  • Publication
    Menschzentrierte KI-Anwendungen in der industriellen Produktion
    Den Systemen künstlicher Intelligenz (KI) werden erhebliche Nutzenpotenziale in der industriellen Produktion zugesprochen. Ihre Anwendungsfelder reichen von der Qualitätssicherung über die vorbeugende Instandhaltung bis hin zur intuitiven Mensch-Roboter-Kollaboration. Allerdings liegen nur wenige Erkenntnisse zur menschzentrierten Umsetzung von KI-Anwendungen in der industriellen Produktion vor. Um die Wissensbasis zu erweitern, wurde in elf Experteninterviews ein aktueller Umsetzungsstand von KI-Anwendungen vornehmlich in mittelständischen Produktionsbetrieben erhoben. Die Befragungskategorien beziehen sich auf die KIStrategie, die Implementierung, das Datenmanagement, die Qualifizierung, die Mensch-Technik-Interaktion sowie auf förderliche bzw. hemmende Faktoren. Der Beitrag dokumentiert konsolidierte Befragungsergebnisse. Zudem wird die Bedeutung einer menschzentrierten Gestaltung erörtert.
  • Publication
    Active Power Curtailment in Power System Planning
    In regions with high penetration of Distributed Energy Resources (DER), grid congestion is mostly caused by power feed-in of generators. In this context, conventional approaches for solving the grid congestion problems are Active Power Curtailment (APC) of DER and grid reinforcement. In the wake of increased DER generation, it is not only of interest how to obtain the grid reinforcement, but also how to consider APC during grid planning. The challenge of such grid planning is to determine the best combination of reinforcement and APC. We present a new grid planning method and use it to find the trade-off between APC and grid reinforcement in a case study of a High-Voltage (HV) grid with a future renewable energy scenario. We compare the proposed reinforcement method to two existing methods. With the combination of APC and grid reinforcement, the proposed method shows a decrease regarding the total expenditures.