Holistic modelling and optimisation of thermal load forecasting, heat generation and plant dispatch for a district heating network

Finkenrath, Matthias; Faber, Till; Behrens, Fabian; Leiprecht, Stefan

doi:10.1016/j.energy.2022.123666

Cited by 15 publications

(3 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…As part of a plant dispatch optimisation process, the forecasting methods that were analysed in this study can help to operate district heating networks more efficiently and cost-effectively. [3] By developing a web interface, the best analysed forecasting models could be made available for use in live operation and optimisation by the district heating system operator.…”

Section: Discussionmentioning

confidence: 99%

“…The economic benefit of improved thermal load forecasting was shown in in a previous publication. [3] New methods in the field of artificial intelligence, especially in machine learning and "deep learning", offer considerable potential for improvement, in particular since machine learning techniques have become easy to deploy due to increasing amounts of training data coupled with cheaper and improved computing power. [4][5][6][7] This work is based on operating data from the district heating network for Ulm, a medium-sized city in southern Germany with about 125,000 inhabitants, as provided by the local utility company "Fernwärme Ulm GmbH" (FUG), which operates a district heating network about 150 km in length supplying some 600,000 MWh of thermal energy per year, equivalent to 45 percent of the city's heat requirements.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Benchmarking of State-Of-The-Art Machine Learning Methods for Highly Accurate Thermal Load Forecasting in District Heating Networks

Pressa,

Stefan,

Fabian

et al. 2023

36th International Conference on Efficiency, Cost, Optimization, Simulation and Environmental Impact of Energy Systems (ECOS 20

View full text Add to dashboard Cite

Decarbonisation of heat generation has become a priority for district heating network operators. In order to avoid the use of fossil-fired boilers, operators need to know peaks in heat demand in advance. Accurate thermal load forecasting is playing an increasingly important role in this respect. This paper presents the final results of the research project "deepDHC" (deep learning for district heating and cooling) funded by the German Federal Ministry for Economic Affairs and Climate Action (BMWK). The threeyear project focused on systematically benchmarking thermal load forecasts for district heating networks, based on state-of-the-art machine learning methods. The analysis covers a variety of machine learning techniques, such as neural networks -including latest deep learning methods -(e.g. LSTM, TFT, ESN, RC), decision trees (random forests, adaptive boosting, XGB) and statistical methods (SARIMAX). In addition, the impact of combining methods by so-called "stacking" was investigated. Training and validation of the machine learning algorithms was based on historical operating data from the district heating network for the city of Ulm in Germany, in combination with historical weather data, and weather forecasts. Thermal load forecaststypically for three days ahead -are presented and compared against one another. An automatic tuning routine was developed as part of the project, which enables regular re-training of the machine learning algorithms based on the latest operating data from the heating network. Furthermore, a web interface for real-time forecasting was developed and implemented at the power station.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Benchmarking of State-Of-The-Art Machine Learning Methods for Highly Accurate Thermal Load Forecasting in District Heating Networks

Pressa,

Stefan,

Fabian

et al. 2023

36th International Conference on Efficiency, Cost, Optimization, Simulation and Environmental Impact of Energy Systems (ECOS 20

View full text Add to dashboard Cite

show abstract

“…The accurate prediction of the short-term heat load trend of buildings helps to avoid energy waste, and provides a promising way to precisely regulate building energy based on energy demand [9][10][11][12][13]. With the expansion of heating scale and the growing complexity of heating data, there is a growing demand for processing the real-time data series of heat supply and predicting the short-term trend accurately for different heating districts and independent buildings [14][15][16][17][18][19][20]. The single prediction model with specific mathematical assumptions and applicable conditions can hardly meet the strict mathematical preconditions and hypotheses concerning the load trend prediction of actual heat supply projects in the background of big data [21][22][23][24].…”

Section: Introductionmentioning

confidence: 99%

Application of Optimal Combinatory Mathematics Model in Heat Load Trend Prediction

Liu¹,

Liu²,

Jiang³

et al. 2022

IJHT

View full text Add to dashboard Cite

The accurate prediction of the short-term heat load trend of buildings helps to prevent energy waste. Combinatory prediction methods need to be studied, as the processing of real-time data series of heat supply and the accurate prediction of short-term trend are more and more demanded by different heating districts and independent buildings. Therefore, this study explores the application of optimal combinatory mathematics model in heat load trend prediction. Firstly, feature extraction was performed on the historical weather data at the locality of the building and the historical data on heat load, creating a short-term trend prediction dataset with days as the unit, and an ultra-short-term trend prediction dataset with hours as the unit. On this basis, a combinatory mathematics model was created for heat load trend prediction. Furthermore, the authors detailed the principles of the two methods, namely, extreme gradient boosting tree (EGBT) and support vector regression (SVR), and explains the combination pattern of the single models in the combinatory model. Then, the weights were optimized by the simulated annealing (SA) algorithm, and the steps of the combinatory model were presented. The proposed model was proved effective through experiments.

show abstract

The predictive management in campus heating system based on deep reinforcement learning and probabilistic heat demands forecasting

Chen,

Xie,

Sun

et al. 2023

Applied Energy

View full text Add to dashboard Cite

Holistic modelling and optimisation of thermal load forecasting, heat generation and plant dispatch for a district heating network

Cited by 15 publications

References 20 publications

Benchmarking of State-Of-The-Art Machine Learning Methods for Highly Accurate Thermal Load Forecasting in District Heating Networks

Benchmarking of State-Of-The-Art Machine Learning Methods for Highly Accurate Thermal Load Forecasting in District Heating Networks

Application of Optimal Combinatory Mathematics Model in Heat Load Trend Prediction

The predictive management in campus heating system based on deep reinforcement learning and probabilistic heat demands forecasting

Contact Info

Product

Resources

About