As digital grid development in Guangdong Province advances, the accumulation of multi-source heterogeneous data and the low efficiency of fault disposition in distribution networks have become increasingly prominent issues. To facilitate the effective use of multi-source heterogeneous data and efficient decision-making for distribution network fault disposition, this paper proposes a method for constructing a fault disposition knowledge graph based on multi-source heterogeneous data. First, the characteristics of multi-source heterogeneous data in distribution networks were analyzed, and a framework for constructing a fault disposition knowledge graph was designed accordingly. Then, named entities were extracted from unstructured fault texts using EBERT-BiLSTM-CRF. Based on the recognized entities, BiLSTM-GA was used to identify semantic relations and generate knowledge triples. Finally, fault data from a distribution network in a specific region of Guangdong Province were used in a case study, and the resulting knowledge graph was visualized and managed using the GridGraph graph database. Within the proposed knowledge graph construction framework, the complete BiLSTM-CRF-GA extraction configuration achieved a mean precision of 91.16%, a mean recall of 92.94%, and a mean F1-score of 92.06% across five matched data splits, supporting its effectiveness in extracting fault-disposition knowledge from multi-source data.
Citation: Xingyu Wang, Hongpeng Liu. Construction and application of distribution network fault disposition knowledge graph based on multi-source data[J]. AIMS Energy, 2026, 14(5): 1164-1185. doi: 10.3934/energy.2026046
As digital grid development in Guangdong Province advances, the accumulation of multi-source heterogeneous data and the low efficiency of fault disposition in distribution networks have become increasingly prominent issues. To facilitate the effective use of multi-source heterogeneous data and efficient decision-making for distribution network fault disposition, this paper proposes a method for constructing a fault disposition knowledge graph based on multi-source heterogeneous data. First, the characteristics of multi-source heterogeneous data in distribution networks were analyzed, and a framework for constructing a fault disposition knowledge graph was designed accordingly. Then, named entities were extracted from unstructured fault texts using EBERT-BiLSTM-CRF. Based on the recognized entities, BiLSTM-GA was used to identify semantic relations and generate knowledge triples. Finally, fault data from a distribution network in a specific region of Guangdong Province were used in a case study, and the resulting knowledge graph was visualized and managed using the GridGraph graph database. Within the proposed knowledge graph construction framework, the complete BiLSTM-CRF-GA extraction configuration achieved a mean precision of 91.16%, a mean recall of 92.94%, and a mean F1-score of 92.06% across five matched data splits, supporting its effectiveness in extracting fault-disposition knowledge from multi-source data.
| [1] |
Ercan SU (2024) Power line Communication: Revolutionizing data transfer over electrical distribution networks. Eng Sci Technol, Int J 52: 101680. https://doi.org/10.1016/j.jestch.2024.101680 doi: 10.1016/j.jestch.2024.101680
|
| [2] |
Siddique MNI, Shafiullah M, Mekhilef S, et al. (2024) Fault classification and location of a PMU-equipped active distribution network using deep convolution neural network (CNN). Electr Power Syst Res 229: 110178. https://doi.org/10.1016/j.epsr.2024.110178 doi: 10.1016/j.epsr.2024.110178
|
| [3] |
Cui J, Ran Z, Shen W, et al. (2024) Study on multi-type flexible load control method of active distribution network based on dynamic time-sharing electricity price. Appl Energy 357: 122479. https://doi.org/10.1016/j.apenergy.2023.122479 doi: 10.1016/j.apenergy.2023.122479
|
| [4] |
Zhou Y, Lin Z, Lin J, et al. (2024) Knowledge graph learning algorithm based on deep convolutional networks. Intell Syst Appl 22: 200386. https://doi.org/10.1016/j.iswa.2024.200386 doi: 10.1016/j.iswa.2024.200386
|
| [5] |
Nawaz R, Qamar A, Wadood A, et al. (2026) Consensus-based feature selection and Bayesian-optimized stacking ensembles for comprehensive fault diagnosis in series-compensated ultra-high-voltage power systems. Sci Rep 16: 18678. https://doi.org/10.1038/s41598-026-49195-7. doi: 10.1038/s41598-026-49195-7
|
| [6] |
Albalawi H, Wadood A, Bukhari SBA, et al. (2025) Design of a novel adaptive weighted multiview deep neural network for enhanced islanding detection in smart distribution systems. IEEE Access 13: 90333–90344. https://doi.org/10.1109/ACCESS.2025.3571907. doi: 10.1109/ACCESS.2025.3571907
|
| [7] |
Chen C, Wang T, Zheng Y, et al. (2023) Reinforcement learning-based distant supervision relation extraction for fault diagnosis knowledge graph construction under industry 4.0. Adv Eng Inform 55: 101900. https://doi.org/10.1016/j.aei.2023.101900 doi: 10.1016/j.aei.2023.101900
|
| [8] |
Fu Y, Zhao H, Zhang H, et al. (2026) Fault detection and diagnosis of the AHU via combining a new hybrid monitoring strategy and an improved graph convolutional network. Energy Build 358: 117177. https://doi.org/ 10.1016/j.enbuild.2026.117177. doi: 10.1016/j.enbuild.2026.117177
|
| [9] |
Zhang H, Zhao H, Fu Y, et al. (2026) The class labels and spatial information based fault diagnosis of air handling unit via combining kernel Fisher discriminant analysis with an improved graph convolutional neural network. Measurement 257: 118622. https://doi.org/10.1016/j.measurement.2025.118622. doi: 10.1016/j.measurement.2025.118622
|
| [10] |
Li Q, Wang Y, Dong J, et al. (2024) Multi-node knowledge graph assisted distributed fault detection for large-scale industrial processes based on graph attention network and bidirectional LSTMs. Neural Netw 173: 106210. https://doi.org/10.1016/j.neunet.2024.106210 doi: 10.1016/j.neunet.2024.106210
|
| [11] |
Mao Z, Wang H, Jiang B, et al. (2024) Graph convolutional neural network for intelligent fault diagnosis of machines via knowledge graph. IEEE Trans Ind Inform 20: 7862–7870. https://doi.org/10.1109/TII.2024.3367010 doi: 10.1109/TII.2024.3367010
|
| [12] |
Xiao X, Li C, Huang J, et al. (2022) Fault diagnosis of rolling bearing based on knowledge graph with data accumulation strategy. IEEE Sens J 22: 18831–18840. https://doi.org/10.1109/JSEN.2022.3201839 doi: 10.1109/JSEN.2022.3201839
|
| [13] |
Gao F, Li G, Gu J, et al. (2023) GridOnto: knowledge representation and extraction for fault events in power grid. IEEE Access 11: 58863–58878. https://doi.org/10.1109/ACCESS.2023.3284839 doi: 10.1109/ACCESS.2023.3284839
|
| [14] |
Ji Z, Wang X, Zhang J, et al. (2023) Construction and application of knowledge graph for grid dispatch fault handling based on pre-trained model. Global Energy Interconnection 6: 493–504. https://doi.org/10.1016/j.gloei.2023.08.009 doi: 10.1016/j.gloei.2023.08.009
|
| [15] |
Li Z, Huang Z, Guo L, et al. (2023) Cognitive knowledge graph generation for grid fault handling based on attention mechanism combined with multi-modal factor fusion. Comput Electr Eng 111: 108855. https://doi.org/10.1016/j. compeleceng.2023. 108855 doi: 10.1016/j.compeleceng.2023.108855
|
| [16] |
Xiao N, Peng B, Li X, et al. (2023) Research on the construction and implementation of power grid fault handling knowledge graphs. Energy Rep 9: 182–189. https://doi.org/10.1016/j.egyr.2023.02.073 doi: 10.1016/j.egyr.2023.02.073
|
| [17] |
Zhao Y, Zhang B, Gao D (2022) Construction of petrochemical knowledge graph based on deep learning. J Loss Prev Process Ind 76: 104736. https://doi.org/10.1016/j.jlp.2022.104736 doi: 10.1016/j.jlp.2022.104736
|
| [18] |
Wang Y, Cheng Y, Qi Q, et al. (2024) IDS-KG: An industrial dataspace-based knowledge graph construction approach for smart maintenance. J Ind Inf Integr 38: 100566. https://doi.org/10.1016/j.jii.2024.100566 doi: 10.1016/j.jii.2024.100566
|
| [19] |
Guo L, Li X, Yan F, et al. (2024) A method for constructing a machining knowledge graph using an improved transformer. Expert Syst Appl 237: 121448. https://doi.org/10.1016/j.eswa.2023.121448 doi: 10.1016/j.eswa.2023.121448
|
| [20] |
Fei Y, Yang H, Zhao G (2022) Research on intelligent construction technology of information-driven power grid security situation knowledge graph. IFAC-PapersOnLine 55: 102–107. https://doi.org/10.1016/j.ifacol.2022.05.018 doi: 10.1016/j.ifacol.2022.05.018
|
| [21] |
Tang X, Feng Z, Xiao Y, et al. (2023) Construction and application of an ontology-based domain-specific knowledge graph for petroleum exploration and development. Geosci Front 14: 101426. https://doi.org/10.1016/j.gsf.2022.101426 doi: 10.1016/j.gsf.2022.101426
|
| [22] |
Bano S, Khalid S, Tairan NM, et al. (2023) Summarization of scholarly articles using BERT and BiGRU: Deep learning-based extractive approach. J King Saud University-Comput Inf Sci 35: 101739. https://doi.org/10.1016/j.jksuci.2023.101739 doi: 10.1016/j.jksuci.2023.101739
|
| [23] |
Zhu Y (2024) A knowledge graph and BiLSTM-CRF-enabled intelligent adaptive learning model and its potential application. Alex Eng J 91: 305–320. https://doi.org/10.1016/j.aej.2024.02.011 doi: 10.1016/j.aej.2024.02.011
|
Figures(11) / Tables(7)
Xingyu Wang, Hongpeng Liu. Construction and application of distribution network fault disposition knowledge graph based on multi-source data[J]. AIMS Energy, 2026, 14(5): 1164-1185. doi: 10.3934/energy.2026046
DownLoad: