Uncertainty quantification of granular computing-neural network model for prediction of pollutant longitudinal dispersion coefficient in aquatic streams

Ghiasi, Behzad; Noori, Roohollah; Sheikhian, Hossein; Zeynolabedin, Amin; Yuanbin, Sun; Jun, Changhyun; Hamouda, Mohamed A.; Bateni, Sayed M.; Abolfathi, Soroush

doi:10.1038/s41598-022-08417-4

Cited by 77 publications

(18 citation statements)

References 78 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Furthermore, as uncertainty quantification remains a challenging and ubiquitous task in real-world ML applications (e.g. in engineering domains such as transportation engineering [70] or water and environmental applications [71]), it could be interesting to use Bayesian machine learning models in SSL (e.g. Deep Gaussian Processes) to quantify uncertainty in downstream prognostics tasks.…”

Section: Discussionmentioning

confidence: 99%

Self-Supervised Learning for Data Scarcity in a Fatigue Damage Prognostic Problem

Akrim¹,

Gogu²,

Vingerhoeds³

et al. 2023

Preprint

View full text Add to dashboard Cite

With the increasing availability of data for Prognostics and Health Management (PHM), Deep Learning (DL) techniques are now the subject of considerable attention for this application, often achieving more accurate Remaining Useful Life (RUL) predictions. However, one of the major challenges for DL techniques resides in the difficulty of obtaining large amounts of labelled data on industrial systems. To overcome this lack of labelled data, an emerging learning technique is considered in our work: Self-Supervised Learning, a sub-category of unsupervised learning approaches. This paper aims to investigate whether pre-training DL models in a self-supervised way on unlabelled sensors data can be useful for RUL estimation with only Few-Shots Learning, i.e. with scarce labelled data. In this research, a fatigue damage prognostics problem is addressed, through the estimation of the RUL of aluminum alloy panels (typical of aerospace structures) subject to fatigue cracks from strain gauge data. Synthetic datasets composed of strain data are used allowing to extensively investigate the influence of the dataset size on the predictive performance. Results show that the self-supervised pre-trained models are able to significantly outperform the non-pre-trained models in downstream RUL prediction task, and with less computational expense, showing promising results in prognostic tasks when only limited labelled data is available.

show abstract

Section: Discussionmentioning

confidence: 99%

Self-Supervised Learning for Data Scarcity in a Fatigue Damage Prognostic Problem

Akrim¹,

Gogu²,

Vingerhoeds³

et al. 2023

Preprint

View full text Add to dashboard Cite

show abstract

“…Therefore, the performance of the BP model is highly affected by the data used for training. Any change in the selected training data will bring huge uncertainty to the model output [30,31]. Ensemble learning accomplishes the learning task by developing multiple learners, which is known to generally achieve a better learning result than a single learner.…”

Section: Ensemble Learning Modelmentioning

confidence: 99%

Analysis of Factors Influencing Wave Overtopping Discharge from Breakwater Based on an MIV-BP Estimation Model

Chen

Peng

et al. 2022

Water

View full text Add to dashboard Cite

Aiming at the problem of calculating the overtopping of single-slope breakwaters, a mean impact value-backpropagation (MIV-BP) estimation model for predicting overtopping was established. Experimental data from the Tianjin Research Institute of Water Transport Engineering (TIWTE) were utilized to further enrich the dataset of the CLASH project for single-slope wave overtopping discharge. This paper established a comprehensive prediction model based on an ensemble learning average method combination strategy. There are 10 input parameters in the model, including the offshore effective wave height, average wave period, offshore water depth, toe submergence, toe width, slope tangent, armor rock surface roughness factor, crest height with respect to the static water level, wall height with respect to the static water level, and crest width; the output parameter is the mean overtopping discharge. Subsequently, a comparative analysis was conducted between this estimation model, the Chinese standard formula calculation model, and the European Van der Meer formula calculation model. Compared with the two formulas mentioned above, this estimation model’s coefficient of correlation increased by 0.23 and 0.26, respectively. Finally, a weight evaluation analysis of the 10 main factors affecting overtopping was carried out based on a MIV-BP neural network model. In the analysis, a positive correlation was found for factors, such as the wave height, average wave period, and water depth at the structure toe; a negative correlation was found for factors, such as the slope, crest height with respect to the static water level, wall height with respect to the static water level, and crest width. Overall, the results provide a significant basis and reference for optimizing the design of the wave overtopping control.

show abstract

“…Furthermore, rainfall causes changes in the hydrodynamic conditions of the river, leading to sediment resuspension. The resuspension of sediment can re-suspend the bed sediments that are home to large nutrient loads in the river [2]. Excessive nutrient pollution causing eutrophication triggers the proliferation of HABs and poses a negative impact on the environment.…”

Section: Introductionmentioning

confidence: 99%

Research on the Development and Application of a Deep Learning Model for Effective Management and Response to Harmful Algal Blooms

Kim

et al. 2023

Water

View full text Add to dashboard Cite

Harmful algal blooms (HABs) caused by harmful cyanobacteria adversely impact the water quality in aquatic ecosystems and burden socioecological systems that are based on water utilization. Currently, Korea uses the Environmental Fluid Dynamics Code-National Institute of Environmental Research (EFDC-NIER) model to predict algae conditions and respond to algal blooms through the HAB alert system. This study aimed to establish an additional deep learning model to effectively respond to algal blooms. The prediction model is based on a deep neural network (DNN), which is a type of artificial neural network widely used for HAB prediction. By applying the synthetic minority over-sampling technique (SMOTE) to resolve the imbalance in the data, the DNN model showed improved performance during validation for predicting the number of cyanobacteria cells. The R-squared increased from 0.7 to 0.78, MAE decreased from 0.7 to 0.6, and RMSE decreased from 0.9 to 0.7, indicating an enhancement in the model’s performance. Furthermore, regarding the HAB alert levels, the R-squared increased from 0.18 to 0.79, MAE decreased from 0.2 to 0.1, and RMSE decreased from 0.3 to 0.2, indicating improved performance as well. According to the results, the constructed data-based model reasonably predicted algae conditions in the summer when algal bloom-induced damage occurs and accurately predicted the HAB alert levels for immediate decision-making. The main objective of this study was to develop a new technology for predicting and managing HABs in river environments, aiming for a sustainable future for the aquatic ecosystem.

show abstract

Uncertainty quantification of granular computing-neural network model for prediction of pollutant longitudinal dispersion coefficient in aquatic streams

Cited by 77 publications

References 78 publications

Self-Supervised Learning for Data Scarcity in a Fatigue Damage Prognostic Problem

Self-Supervised Learning for Data Scarcity in a Fatigue Damage Prognostic Problem

Analysis of Factors Influencing Wave Overtopping Discharge from Breakwater Based on an MIV-BP Estimation Model

Research on the Development and Application of a Deep Learning Model for Effective Management and Response to Harmful Algal Blooms

Contact Info

Product

Resources

About