A machine learning approach to fault detection in district heating substations. (September 2018)

• Record Type: Journal Article
• Authors: Månsson, Sara; Kallioniemi, Per-Olof Johansson; Sernhed, Kerstin; Thern, Marcus
• Is Part Of: Energy procedia. Volume 149(2018)
• Page Start: 226
• Page End: 235
• Publication Date: 2018-09
• Subjects: District heating substations -- fault detection -- machine learning; Power resources -- Congresses; Power resources -- Periodicals; Power resources; Conference proceedings; Periodicals; 333.7905
• DOI: 10.1016/j.egypro.2018.08.187
• Languages: English
• ISSNs: 1876-6102

Abstract

Abstract: The aim of this study is to develop a model capable of predicting the behavior of a district heating substation, including being able to distinguish datasets from well performing substations from datasets containing faults. The model developed in the study is based on machine learning algorithms and the model is trained on data from a Swedish district heating substation. A number of different models and input/output parameters are tested in the study. The results show that the model is capable of modelling the substation behavior, and that the fault detection capability of the model is high.