Load Forecasting in District Heating Systems Using Stacked Ensembles of Machine Learning Algorithms

Faber, Till; Finkenrath, Matthias

doi:10.2991/ahe.k.210202.001

Cited by 4 publications

(3 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Usually, an ANN is trained during the model development phase using experimental or previous operational data [14]. Among others, the focus for ML predictions in recent years has often been on energy predictions for buildings [3,6,20] or whole district heating systems, where the usual load forecasts are based on decision trees or simple neural networks [2,5].…”

Section: Related Workmentioning

confidence: 99%

Increasing Energy Efficiency and Flexibility by Forecasting Production Energy Demand Based on Machine Learning

Trenz

Hoffmann

Lange

et al. 2023

Lecture Notes in Mechanical Engineering

View full text Add to dashboard Cite

The ability of manufacturing companies to compete depends strongly on the efficient use of production resources and the flexibility to adapt to changing production conditions. Essential requirements for the energetic infrastructure (EGI) result from the production itself, e.g., security of supply, efficiency and peak shaving. Since production always takes priority and must not be disturbed, the flexibility potential in terms of energy efficiency lies primarily in the EGI. Based on this, strategies will be developed that support companies in increasing their efficiency and flexibility by optimizing the configuration and operation of the EGI, while production processes are reliably supplied and not adapted. This is reached with intelligent operation strategies for the heating and cooling network based on forecasts, the use of energy storage systems, and the coupling of energy sectors. This paper presents an approach for energy forecasts used for the optimization of operation strategies. Hence, an energy-forecast-tool was developed, which is used for the prediction of electrical and thermal loads depending on the expected production. Therefore, machine learning models are trained with past weather, energy, and production data. Using production planning data and weather forecasts, the model can predict energy demands as input for an EGI optimization.

show abstract

Section: Related Workmentioning

confidence: 99%

Increasing Energy Efficiency and Flexibility by Forecasting Production Energy Demand Based on Machine Learning

Trenz

Hoffmann

Lange

et al. 2023

Lecture Notes in Mechanical Engineering

View full text Add to dashboard Cite

show abstract

“…Carbonneau, Laframboise, and Vahidov (2008) found that RNN and SVM models had the best predictive accuracy when applied to forecasting distorted demand in manufacturing supply chain management. Faber and Finkenrath (2021) proposed a model for predicting heat demand for heating using existing decision tree-based regression algorithms such as AdaBoost and Random Forest (RF), as well as ANN, feed forward neural networks (FNN), and Long Short-Term Memory (LSTM). They also proposed an optimized power plant layout method based on this model.…”

Section: Introductionmentioning

confidence: 99%

Methodology for Improving the Performance of Demand Forecasting Through Machine Learning

Yoon

Park

Song

et al. 2023

Preprint

View full text Add to dashboard Cite

Accurate demand forecasting is crucial for industries to make strategic decisions and maintain their competitive edge. However, existing demand forecasting methods have prodigious problems, especially when it comes to handling the uncertainty, complexity, and nonlinearity of demand forecasting. In addition, the lack of historical data and data biases can create unreliable sources, which discourages the utilization of demand forecasting at a higher level of implementation in businesses. In addition, lack of historical data and data biases can create unreliable sources, which discourages utilization of demand forecasting at higher level of implementation in businesses. The proposed hybrid model aims to improve demand forecasting performance by combining the strengths of existing methods such as K-means clustering, LASSO regression, and LSTM deep learning. By leveraging these techniques, the model can overcome the limitations of each method and improve the accuracy of demand forecasting in various industries. K-means clustering helps to group similar data points, LASSO regression helps to select the most relevant features, and LSTM deep learning helps to capture the temporal dependencies in the data. The combination of these techniques can result in a more accurate and robust demand forecasting model. The model was tested on 2,548 retail products, and outperformed three benchmarking models using the mMAPE, RMSE, and MAE indicators. The proposed model can be used in the retail industry to improve management performance and decision-making, and its ability to optimize variables for each cluster can improve resource allocation.

show abstract

“…Demand forecasting is one of the most complex yet fundamental parts of hospital construction [22]. Although there are many studies on energy optimization in the literature [23][24][25][26][27], there is a lack of information about the maximum power estimation of hospitals.…”

Section: Introductionmentioning

confidence: 99%

Electricity Demand Forecasting of Hospital Buildings in Istanbul

Soyler¹,

Izgi

2022

Sustainability

View full text Add to dashboard Cite

Electricity demand forecasting is essential for utilities. For the consumer, predictability of demand is vital for efficient operation, installation, sizing and maintenance planning. Hospitals, which are among the institutions with high-energy consumption, provide uninterrupted service 24 h a day, 7 days a week. Every hospital building is unique, and many do not conform to a typical shape or floor plan. Depending on the services provided, each hospital can differ significantly in terms of energy demand. Therefore, demand forecasting is one of the most complex elements of hospital construction. Although there are many studies on energy optimization related to hospital buildings in the literature, there is a knowledge gap regarding the maximum power estimation of hospitals. In this study, the annual electrical energy use of 23 public hospitals with over 100 beds in Istanbul is measured, and after determining the monthly peak loads, two new forecasting models are generated using regression techniques for maximum demand forecasting. It is determined that the design criteria used in power calculations in hospitals was very high. A positive result was obtained from the linear regression technique, which is one of the basic regression techniques, and it was shown that the maximum power needs of the hospital can be estimated with great confidence by determining a new design factor in the light of the determined values. This study allows designers to set maximum demands and select transformer and generator sizes with a single formula.

show abstract

Load Forecasting in District Heating Systems Using Stacked Ensembles of Machine Learning Algorithms

Cited by 4 publications

References 7 publications

Increasing Energy Efficiency and Flexibility by Forecasting Production Energy Demand Based on Machine Learning

Increasing Energy Efficiency and Flexibility by Forecasting Production Energy Demand Based on Machine Learning

Methodology for Improving the Performance of Demand Forecasting Through Machine Learning

Electricity Demand Forecasting of Hospital Buildings in Istanbul

Contact Info

Product

Resources

About