Refine
Year of publication
- 2021 (4) (remove)
Document Type
- conference proceeding (article) (2)
- Article (1)
- Handout (1)
Publication reviewed
- begutachtet (3)
- nicht begutachtet (1)
Keywords
- Maschinelles Lernen (4) (remove)
Institute
KI-Anwendung in der Energietechnik: Einsatz maschineller Lernverfahren für die Wärmelastprognose
(2021)
Maschinelles Lernen gilt als eines der vielversprechendsten Teilgebiete der künstlichen Intelligenz (KI). Sein Einsatz hat in den vergangenen Jahren zu enormen Fortschritten sowohl in der Bild- und Texterkennung als auch bei Zeitreihenprognosen geführt.
Der Vortrag demonstriert dies am Beispiel von Wärmelastprognosen für die Fernwärmebranche. Dabei werden maschinelle Lernverfahren genutzt, um den Wärmebedarf in Fernwärme-netzen über mehrere Tage im Voraus hochgenau vorherzusagen. Auf diese Weise können Energieversorger den Einsatz ihrer Wärmeerzeugungsanlagen optimal planen. Beispielsweise können bei vorhersehbaren Lastspitzen Wärmespeicher frühzeitig mit erneuerbar erzeugter Wärme gefüllt und so der Betrieb fossiler Spitzenlastkraftwerke vermieden werden. Als Folge kann der Betreiber sowohl CO2-Emissionen als auch Emissions-, Brennstoff- sowie An- und Abfahrkosten einsparen.
Der Vortrag basiert auf Ergebnissen aus mehreren Forschungsprojekten, die an der Hochschule Kempten seit 2016 gemeinsam mit Fernwärmeversorgern durchgeführt wurden. Dabei wurden maschinelle Lernverfahren sehr unterschiedlicher Komplexität systematisch untersucht und bewertet – von „einfachen“ bis hin zu anspruchsvollen Verfahren aus dem Bereich des sogenannten „Deep Learning“. Die Wärmelastprognosen werden unter Verwendung historischer Betriebs- und Wetterdaten sowie von Wetterprognosen vollautomatisiert erstellt und dem Betreiber über eine Web-Schnittstelle zur Verfügung gestellt, die auch im Vortrag gezeigt wird.
Die vorgestellte Methode bietet erhebliche Einsparpotenziale für den Anlagenbetreiber. Sie ist zudem auch auf andere Branchen mit ähnlichen Zielgrößen bzw. Fragestellungen übertragbar.
Precise forecasting of thermal loads is a critical factor for economic and efficient operation of district heating and cooling networks. If thermal loads are known with high accuracy in advance, use of renewable energies can be maximized, and – in combination with thermal storage units – fossil generation, in particular in peaking units, can be avoided. Machine learning has proven to be a powerful tool for time series forecasting, and has demonstrated significant advancements in recent years. This paper presents the scientific methodology and first results of the publicly funded research project “deepDHC”, which aims at a broad benchmarking of traditional and advanced machine learning methods for thermal load forecasting in district heating and cooling applications. The analysis covers autoregressive forecasting approaches, decision trees such as “adaptive boosting”, but also latest “deep learning” techniques such as the “long short-term memory” (LSTM) neural network. This work is based on data from the district heating network of the city of Ulm in Germany. First, different performance metrics for evaluating forecasting qualities are introduced. Second, approaches for data screening and results of a linear and non-linear correlation analysis are presented. Third, the machine learning tuning process is described. For thermal load forecasting, weather data are key input parameters. This work uses hourly weather forecasts from weather models provided by the German meteorological service. These weather data are updated automatically, and have been statistically corrected in order to represent very accurate point forecasts for up to ten days ahead. In addition, a user-friendly web interface has been developed for use by the district heating network operator. The performance of different machine-learning algorithms is compared based on 72 h heating load forecasts.
Decarbonisation of district heating networks requires a heat generation and storage portfolio that allows for maximised integration of renewable energies. However, it is also essential to have a precise forecast of the thermal loads to be expected in the network and – based on this forecast – an optimised dispatch strategy in order to best match the available generation and storage portfolio with the actual heat demands in the grid. This paper presents a holistic approach that combines modelling and optimisation activities related to these three aspects: first, detailed process modelling and optimisation of power plants and thermal storages; second, a numerical model for dispatch optimisation; and third, machine-learning-based load forecasting. This work, which was performed as part of the publicly funded research projects “KWKflex” and “deepDHC”, was based on operating data from the district heating network of the city of Ulm in Germany. The paper presents the modelling, validation and simulation results of a stationary and instationary process simulation for a 58 MW thermal biomass-fired combined heat and power plant. The analysis identifies a potential to integrate additional renewable power of up to 17 MW thermal power by “power-to-heat” technologies into different parts of the process. The economic benefit is quantified with a mixed-integer linear programming dispatch optimisation model of the district heating network. In order to allow for real-time optimisation of the power plant and thermal energy storage dispatch, a machine-learning-based thermal load forecasting method was developed. The performance of different machine learning algorithms, including decision trees and deep learning techniques, is compared based on a 72-hour forecast horizon. In addition, the economic impact of uncertainties in thermal load prediction is analysed with the numerical dispatch optimisation tool.
Load Forecasting in District Heating Systems Using Stacked Ensembles of Machine Learning Algorithms
(2021)
For district heating, heat demand forecasting is playing a key role for an optimised power plant dispatch. Machine Learning can help to significantly improve forecasts of thermal loads. The prediction quality of neural networks is higher than that of decision trees in most cases. However, compared to decision trees neural networks have weaknesses when extrapolating outside known ranges. This work presents a novel method called “Deep DHC” (Deep Learning for District Heating and Cooling), which combines these two approaches in order to benefit from strengths of both methods. On the one hand, the novel approach uses conventional decision tree based regression algorithms such as the AdaBoost and Random Forest, as well as artificial neural networks. In addition to common feed forward neural networks (FNN), a deep learning network structure, which consists of long short-term memory (LSTM) cells, is used for the first time. The LSTM method has already proven to be very powerful in modern speech recognition. In order to achieve best possible heat demand forecasts, the aforementioned methods for load forecasting are combined and weighted by an additional machine learning method. Results show that it is possible to achieve a further improvement in forecasting quality for district heating loads by purposefully combining individual forecasting methods. Hence, mean and absolute deviations are significantly reduced in comparison to the individual methods.