Data-Driven and Knowledge-Guided Heterogeneous Graphs and Temporal Convolution Networks for Flood Forecasting
Abstract
:1. Introduction
- (1)
- This article proposes a data-driven and knowledge-based dual-drive flood prediction model, which optimizes the initial conditions of the model through heterogeneous graph modules and introduces prior knowledge of the watershed. TCN with integrated attention is used to fit the model, significantly reducing data noise and enhancing the robustness of the model.
- (2)
- The proposed model effectively integrates the guiding framework technology of flood physics theory into data-driven models, and introduces physical constraints that comply with the physical laws of flood flow prediction principles, solving the problem of errors in common data-driven model predictions that do not conform to physical knowledge.
- (3)
- This article first discusses the combination of tensor low-rank approximation data preprocessing methods and flood prediction mechanisms to effectively compensate for tensor sparsity.
- (4)
- It is recommended to use interval prediction output architecture to quantify the output error of the model, in order to improve the accuracy of flood time-series data prediction. The experimental results show that compared to the baseline, the proposed model improves the index of PI coverage probability (PICP) by 11.4%.
2. Materials and Methods
2.1. Preliminaries
2.1.1. Tensor Decomposition
2.1.2. Graph Convolution
2.1.3. Temporal Convolution
2.2. Methods
2.2.1. Knowledge-Guided Framework Based on Physical Constraints
2.2.2. Pretreatment of Data Tensioning
2.2.3. Heterography Module
- Node attention module
- 2.
- Relational attention module
- 3.
- Among these, represents the weight of the r-th relationship to node j.
2.2.4. Temporal Feature Extraction
2.2.5. Model Train
3. Results
3.1. Datasets
3.2. Baseline and Parameter Settings
3.3. Evaluation Metrics
3.4. Experimental Results
3.4.1. Comparison
3.4.2. Robustness Analysis
3.4.3. Attention Module Analysis of DK-HTAN
4. Discussion
5. Conclusions
Author Contributions
Funding
Institutional Review Board Statement
Informed Consent Statement
Data Availability Statement
Conflicts of Interest
References
- Shi, Y.; Yao, Q.; Wen, J.; Xi, J.; Li, H.; Wang, Q. A spatial accessibility assessment of urban tourist attractions emergency response in Shanghai. Int. J. Disaster Risk Reduct. 2022, 74, 102919. [Google Scholar] [CrossRef]
- Yang, H.; Li, W. Data decomposition, seasonal adjustment method and machine learning combined for runoff prediction: A case study. Water Resour. Manag. 2023, 37, 557–581. [Google Scholar] [CrossRef]
- Li, K.; Huang, G.; Wang, S.; Razavi, S. Development of a physics-informed data-driven model for gaining insights into hydrological processes in irrigated watersheds. J. Hydrol. 2022, 613, 128323. [Google Scholar] [CrossRef]
- Wu, C.; Zhang, X.; Wang, W.; Lu, C.; Zhang, Y.; Qin, W.; Shu, L.; Tick, G.R.; Liu, B. Groundwater level modeling framework by combining the wavelet transform with a long short-term memory data-driven model. Sci. Total Environ. 2021, 783, 146948. [Google Scholar] [CrossRef] [PubMed]
- He, X.; Liu, R.; Anumba, C.J. Data-driven insights on the knowledge gaps of conceptual cost estimation modeling. J. Constr. Eng. Manag. 2021, 147, 04020165. [Google Scholar] [CrossRef]
- Xia, M.; Shao, H.; Ma, X.; de Silva, C.W. A stacked GRU-RNN-based approach for predicting renewable energy and electricity load for smart grid operation. IEEE Trans. Ind. Inform. 2021, 17, 7050–7059. [Google Scholar] [CrossRef]
- Wang, H.; Lu, B.; Li, J.; Liu, T.; Xing, Y.; Lv, C.; Hashemi, E.; Cao, D.; Li, J.; Zhang, J. Risk Assessment and Mitigation in Local Path Planning for Autonomous Vehicles With LSTM Based Predictive Model. IEEE Trans. Autom. Sci. Eng. 2021, 19, 2738–2749. [Google Scholar] [CrossRef]
- Cao, X.; Ren, N.; Tian, G.; Fan, Y.; Duan, Q. A three-dimensional prediction method of dissolved oxygen in pond culture based on Attention-GRU-GBRT. Comput. Electron. Agric. 2021, 181, 105955. [Google Scholar] [CrossRef]
- Wang, Q.; Huang, K.; Chandak, P.; Zitnik, M.; Gehlenborg, N. Extending the nested model for user-centric XAI: A design study on GNN-based drug repurposing. IEEE Trans. Vis. Comput. Graph. 2022, 29, 1266–1276. [Google Scholar] [CrossRef]
- Pradhyumna, P.; Shreya, G.P. Graph neural network (GNN) in image and video understanding using deep learning for computer vision applications. In Proceedings of the 2021 Second International Conference on Electronics and Sustainable Communication Systems (ICESC), Coimbatore, India, 4–6 August 2021; pp. 1183–1189. [Google Scholar]
- Yang, Z.; Dong, S. HAGERec: Hierarchical attention graph convolutional network incorporating knowledge graph for explainable recommendation. Knowl.-Based Syst. 2020, 204, 106194. [Google Scholar] [CrossRef]
- Wang, T.; Jin, D.; Wang, R.; He, D.; Huang, Y. Powerful graph convolutional networks with adaptive propagation mechanism for homophily and heterophily. In Proceedings of the 36th AAAI Conference on Artificial Intelligence, Virtual, 22 February–1 March 2022; Volume 36, pp. 4210–4218. [Google Scholar]
- Gupta, K.; Ajanthan, T. Improved gradient-based adversarial attacks for quantized networks. In Proceedings of the 36th AAAI Conference on Artificial Intelligence, Virtual, 22 February–1 March 2022; Volume 36, pp. 6810–6818. [Google Scholar]
- Lindemann, B.; Müller, T.; Vietz, H.; Jazdi, N.; Weyrich, M. A survey on long short-term memory networks for time series prediction. Procedia CIRP 2021, 99, 650–655. [Google Scholar] [CrossRef]
- Fan, J.; Zhang, K.; Huang, Y.; Zhu, Y.; Chen, B. Parallel spatio-temporal attention-based TCN for multivariate time series prediction. Neural Comput. Appl. 2023, 35, 13109–13118. [Google Scholar] [CrossRef]
- Zhou, M.; Wang, B.; Guo, S.; Watada, J. Multi-objective prediction intervals for wind power forecast based on deep neural networks. Inf. Sci. 2021, 550, 207–220. [Google Scholar] [CrossRef]
- Peng, X.; Wang, H.; Lang, J.; Li, W.; Xu, Q.; Zhang, Z.; Li, C.; Cai, T.; Duan, S.; Liu, F. EALSTM-QR: Interval wind-power prediction model based on numerical weather prediction and deep learning. Energy 2021, 220, 119692. [Google Scholar] [CrossRef]
- Piadeh, F.; Behzadian, K.; Alani, A.M. A critical review of real-time modelling of flood forecasting in urban drainage systems. J. Hydrol. 2022, 607, 127476. [Google Scholar] [CrossRef]
- Deutz, S.; Bardow, A. Life-cycle assessment of an industrial direct air capture process based on temperature–vacuum swing adsorption. Nat. Energy 2021, 6, 203–213. [Google Scholar] [CrossRef]
- Wei, Q.; Li, X.; Song, M. De-aliased seismic data interpolation using conditional Wasserstein generative adversarial networks. Comput. Geosci. 2021, 154, 104801. [Google Scholar] [CrossRef]
- Chen, X.; Wang, J.; Zhang, G.; Peng, Q. Tensor-Based Parametric Spectrum Cartography From Irregular Off-Grid Samplings. IEEE Signal Process. Lett. 2023, 30, 513–517. [Google Scholar] [CrossRef]
- Chen, X.; He, Z.; Wang, J. Spatial-temporal traffic speed patterns discovery and incomplete data recovery via SVD-combined tensor decomposition. Transp. Res. Part C Emerg. Technol. 2018, 86, 59–77. [Google Scholar] [CrossRef]
- Zhou, S.; Erfani, S.; Bailey, J. Online CP decomposition for sparse tensors. In Proceedings of the 2018 IEEE International Conference on Data Mining (ICDM), Singapore, 17–20 November 2018; pp. 1458–1463. [Google Scholar]
- Che, M.; Wei, Y. Randomized algorithms for the approximations of Tucker and the tensor train decompositions. Adv. Comput. Math. 2019, 45, 395–428. [Google Scholar] [CrossRef]
- Yuan, L.; Li, C.; Mandic, D.; Cao, J.; Zhao, Q. Tensor ring decomposition with rank minimization on latent space: An efficient approach for tensor completion. In Proceedings of the AAAI Conference on Artificial Intelligence, Honolulu, HI, USA, 27 January–1 February 2019; Volume 33, pp. 9151–9158. [Google Scholar]
- Wang, J.; Xu, G.; Li, C.; Wang, Z.; Yan, F. Surface defects detection using non-convex total variation regularized RPCA with kernelization. IEEE Trans. Instrum. Meas. 2021, 70, 1–13. [Google Scholar] [CrossRef]
- He, X.; Deng, K.; Wang, X.; Li, Y.; Zhang, Y.; Wang, M. Lightgcn: Simplifying and powering graph convolution network for recommendation. In Proceedings of the 43rd International ACM SIGIR Conference on Research and Development in Information Retrieval, Virtual, 25–30 July 2020; pp. 639–648. [Google Scholar]
- Hewage, P.; Behera, A.; Trovati, M.; Pereira, E.; Ghahremani, M.; Palmieri, F.; Liu, Y. Temporal convolutional neural (TCN) network for an effective weather forecasting using time-series data from the local weather station. Soft Comput. 2020, 24, 16453–16482. [Google Scholar] [CrossRef] [Green Version]
- Yang, J.H.; Zhao, X.L.; Ji, T.Y.; Ma, T.H.; Huang, T.Z. Low-rank tensor train for tensor robust principal component analysis. Appl. Math. Comput. 2020, 367, 124783. [Google Scholar] [CrossRef]
- Kong, X.; Yang, C.; Cao, S.; Li, C.; Peng, Z. Infrared small target detection via nonconvex tensor fibered rank approximation. IEEE Trans. Geosci. Remote Sens. 2021, 60, 5000321. [Google Scholar] [CrossRef]
- Chen, Z.; Liang, J.; Wang, T.; Tang, B.; So, H.C. Generalized MBI algorithm for designing sequence set and mismatched filter bank with ambiguity function constraints. IEEE Trans. Signal Process. 2022, 70, 2918–2933. [Google Scholar] [CrossRef]
- Dodangeh, E.; Panahi, M.; Rezaie, F.; Lee, S.; Bui, D.T.; Lee, C.W.; Pradhan, B. Novel hybrid intelligence models for flood-susceptibility prediction: Meta optimization of the GMDH and SVR models with the genetic algorithm and harmony search. J. Hydrol. 2020, 590, 125423. [Google Scholar] [CrossRef]
- Ding, Y.; Zhu, Y.; Wu, Y.; Jun, F.; Cheng, Z. Spatio-temporal attention LSTM model for flood forecasting. In Proceedings of the 2019 International Conference on Internet of Things (IThings) and IEEE Green Computing and Communications (GreenCom) and IEEE Cyber, Physical and Social Computing (CPSCom) and IEEE Smart Data (SmartData), Atlanta, GA, USA, 14–17 July 2019; pp. 458–465. [Google Scholar]
- Yan, L.; Feng, J.; Hang, T.; Zhu, Y. Flow interval prediction based on deep residual network and lower and upper boundary estimation method. Appl. Soft Comput. 2021, 104, 107228. [Google Scholar] [CrossRef]
Region | Model | PICP, Flood Season | PICP, Non-Flood Season | PINRW, Flood Season | PINRW, Non-Flood Season |
---|---|---|---|---|---|
Qijiang | SVR | 0.689 (0.361) | 0.651 (0.233) | 0.237 (0.119) | 0.241 (0.012) |
RNN | 0.724 (0.412) | 0.706 (0.246) | 0.205 (0.145) | 0.198 (0.023) | |
LSTM | 0.799 (0.698) | 0.758 (0.259) | 0.189 (0.189) | 0.194 (0.012) | |
STALSTM | 0.812 (0.634) | 0.791 (0.312) | 0.177 (0.259) | 0.179 (0.031) | |
AGCLSTM | 0.841 (0.287) | 0.820 (0.189) | 0.164 (0.012) | 0.172 (0.022) | |
DKHTAN | 0.904 (0.264) | 0.896 (0.174) | 0.157 (0.014) | 0.163 (0.027) |
Region | Model | MAPE, Flood Season | MAPE, Non-Flood Season | RMSE, Flood Season | RMSE, Non-Flood Season |
---|---|---|---|---|---|
Qijiang | SVR | 0.158 (0.361) | 0.163 (0.233) | 0.337 (0.119) | 0.341 (0.012) |
RNN | 0.149 (0.412) | 0.154 (0.246) | 0.335 (0.145) | 0.338 (0.023) | |
LSTM | 0.135 (0.698) | 0.141 (0.259) | 0.306 (0.189) | 0.304 (0.012) | |
STALSTM | 0.129 (0.634) | 0.133 (0.312) | 0.283 (0.259) | 0.297 (0.031) | |
AGCLSTM | 0.119 (0.287) | 0.120 (0.189) | 0.269 (0.012) | 0.282 (0.022) | |
DK-HTAN | 0.112 (0.264) | 0.116 (0.174) | 0.221 (0.014) | 0.203 (0.027) |
Region | Model | 5% | 10% | 15% | 20% |
---|---|---|---|---|---|
Qijiang | SVR | 0.611 (0.361) | 0.599 (0.233) | 0.437 (0.119) | 0.441 (0.012) |
RNN | 0.709 (0.412) | 0.656 (0.246) | 0.525 (0.145) | 0.538 (0.023) | |
LSTM | 0.712 (0.698) | 0.684 (0.259) | 0.570 (0.189) | 0.564 (0.012) | |
STALSTM | 0.807 (0.634) | 0.791 (0.312) | 0.634 (0.259) | 0.608 (0.031) | |
AGCLSTM | 0.813 (0.287) | 0.792 (0.189) | 0.673 (0.012) | 0.632 (0.022) | |
DK-HTAN | 0.824 (0.264) | 0.735 (0.174) | 0.711 (0.014) | 0.683 (0.027) |
Region | Model | 5% | 10% | 15% | 20% |
---|---|---|---|---|---|
Qijiang | SVR | 0.601 (0.361) | 0.589 (0.233) | 0.530 (0.119) | 0.511 (0.012) |
RNN | 0.719 (0.412) | 0.646 (0.246) | 0.623 (0.145) | 0.607 (0.023) | |
LSTM | 0.732 (0.698) | 0.674 (0.259) | 0.660 (0.189) | 0.644 (0.012) | |
STALSTM | 0.789 (0.634) | 0.721 (0.312) | 0.704 (0.259) | 0.698 (0.031) | |
AGCLSTM | 0.816 (0.287) | 0.842 (0.189) | 0.760 (0.012) | 0.733 (0.022) | |
DK-HTAN | 0.833 (0.264) | 0.815 (0.174) | 0.791 (0.014) | 0.763 (0.027) |
Model | Node Attention | Relational Attention |
---|---|---|
DK-HTAN | weighted | weighted |
DK-HTAN-a | weighted | mean |
DK-HTAN-b | mean | weighted |
DK-HTAN-c | mean | mean |
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content. |
© 2023 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).
Share and Cite
Shao, P.; Feng, J.; Wu, Y.; Wang, W.; Lu, J. Data-Driven and Knowledge-Guided Heterogeneous Graphs and Temporal Convolution Networks for Flood Forecasting. Appl. Sci. 2023, 13, 7191. https://doi.org/10.3390/app13127191
Shao P, Feng J, Wu Y, Wang W, Lu J. Data-Driven and Knowledge-Guided Heterogeneous Graphs and Temporal Convolution Networks for Flood Forecasting. Applied Sciences. 2023; 13(12):7191. https://doi.org/10.3390/app13127191
Chicago/Turabian StyleShao, Pingping, Jun Feng, Yirui Wu, Wenpeng Wang, and Jiamin Lu. 2023. "Data-Driven and Knowledge-Guided Heterogeneous Graphs and Temporal Convolution Networks for Flood Forecasting" Applied Sciences 13, no. 12: 7191. https://doi.org/10.3390/app13127191
APA StyleShao, P., Feng, J., Wu, Y., Wang, W., & Lu, J. (2023). Data-Driven and Knowledge-Guided Heterogeneous Graphs and Temporal Convolution Networks for Flood Forecasting. Applied Sciences, 13(12), 7191. https://doi.org/10.3390/app13127191