Martin Oberascher, Claudia Maussner, Andrea Cominola, R. Sitzenfrei
Model-based leakage localisation in water distribution networks requires accurate estimates of nodal demands to correctly simulate hydraulic conditions. While digital water meters installed at household premises can be used to provide high-resolution information on water demands, questions arise regarding the necessary temporal resolution of water demand data for effective leak localisation. In addition, how do temporal and spatial data gaps affect leak localisation performance? To address these research gaps, a real-world water distribution network is first extended with the stochastic water end-use model PySIMDEUM. Then, more than 700 scenarios for leak localisation assessment characterised by different water demand sampling resolutions, data gap rates, leak size, time of day for analysis, and data imputation methods are investigated. Numerical results indicate that during periods with high/peak demand, a fine temporal resolution (e.g., 15 min or lower) is required for the successful localisation of leakages. However, regardless of the sampling frequency, leak localisation with a sensitivity analysis achieves a good performance during periods with low water demand (localisation success is on average 95%). Moreover, improvements in leakage localisation might occur depending on the data imputation method selected for data gap management, as they can mitigate random/sudden temporal and spatial fluctuations of water demands.
{"title":"Sensitivity of model-based leakage localisation in water distribution networks to water demand sampling rates and spatio-temporal data gaps","authors":"Martin Oberascher, Claudia Maussner, Andrea Cominola, R. Sitzenfrei","doi":"10.2166/hydro.2024.245","DOIUrl":"https://doi.org/10.2166/hydro.2024.245","url":null,"abstract":"\u0000 Model-based leakage localisation in water distribution networks requires accurate estimates of nodal demands to correctly simulate hydraulic conditions. While digital water meters installed at household premises can be used to provide high-resolution information on water demands, questions arise regarding the necessary temporal resolution of water demand data for effective leak localisation. In addition, how do temporal and spatial data gaps affect leak localisation performance? To address these research gaps, a real-world water distribution network is first extended with the stochastic water end-use model PySIMDEUM. Then, more than 700 scenarios for leak localisation assessment characterised by different water demand sampling resolutions, data gap rates, leak size, time of day for analysis, and data imputation methods are investigated. Numerical results indicate that during periods with high/peak demand, a fine temporal resolution (e.g., 15 min or lower) is required for the successful localisation of leakages. However, regardless of the sampling frequency, leak localisation with a sensitivity analysis achieves a good performance during periods with low water demand (localisation success is on average 95%). Moreover, improvements in leakage localisation might occur depending on the data imputation method selected for data gap management, as they can mitigate random/sudden temporal and spatial fluctuations of water demands.","PeriodicalId":54801,"journal":{"name":"Journal of Hydroinformatics","volume":null,"pages":null},"PeriodicalIF":2.2,"publicationDate":"2024-07-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141807618","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Nhu Y. Nguyen, Tran Ngoc Anh, Huu Duy Nguyen, Kha Dinh Dang
Accurate precipitation is crucial for hydrological modelling, especially in sparse gauge regions like the Lam River Basin (LRB) in Vietnam. Gridded precipitation data sets derived from satellite and numerical models offer significant advantages in such areas. However, satellite precipitation estimates (SPEs) are subject to uncertainties, especially in high variable of topography and precipitation. This study focuses on enhancing the accuracy of Integrated Multi-satellitE Retrievals for Global Precipitation Measurement (IMERG), Climate Prediction Center morphing technique (CMORPH) using the Quantile Mapping (QM) technique, aligning the cumulative distribution functions of the observed precipitation data with those of the SPEs, and assessing the impact on hydrological predictions. The study highlights that the post-correction IMERG precipitation using QM performs better than other data sets, enhancing the hydrological model's performance for the LRB at different temporal scales. Nash–Sutcliffe efficiency values increased from 0.60 to 0.77, surpassing the original IMERG's 0.52 to 0.74, and correlation coefficients improved from 0.79 to 0.89 (compared with the previous 0.75–0.86) for hydrological modelling. Additionally,Per cent Bias (PBIAS) decreased from approximately −1.66 to −2.21% (contrasting with the initial −20.22 and 4.6%) with corrected SPEs. These findings have implications for water resource management and disaster risk reduction initiatives in Vietnam and other countries.
{"title":"Quantile mapping technique for enhancing satellite-derived precipitation data in hydrological modelling: a case study of the Lam River Basin, Vietnam","authors":"Nhu Y. Nguyen, Tran Ngoc Anh, Huu Duy Nguyen, Kha Dinh Dang","doi":"10.2166/hydro.2024.225","DOIUrl":"https://doi.org/10.2166/hydro.2024.225","url":null,"abstract":"\u0000 \u0000 Accurate precipitation is crucial for hydrological modelling, especially in sparse gauge regions like the Lam River Basin (LRB) in Vietnam. Gridded precipitation data sets derived from satellite and numerical models offer significant advantages in such areas. However, satellite precipitation estimates (SPEs) are subject to uncertainties, especially in high variable of topography and precipitation. This study focuses on enhancing the accuracy of Integrated Multi-satellitE Retrievals for Global Precipitation Measurement (IMERG), Climate Prediction Center morphing technique (CMORPH) using the Quantile Mapping (QM) technique, aligning the cumulative distribution functions of the observed precipitation data with those of the SPEs, and assessing the impact on hydrological predictions. The study highlights that the post-correction IMERG precipitation using QM performs better than other data sets, enhancing the hydrological model's performance for the LRB at different temporal scales. Nash–Sutcliffe efficiency values increased from 0.60 to 0.77, surpassing the original IMERG's 0.52 to 0.74, and correlation coefficients improved from 0.79 to 0.89 (compared with the previous 0.75–0.86) for hydrological modelling. Additionally,Per cent Bias (PBIAS) decreased from approximately −1.66 to −2.21% (contrasting with the initial −20.22 and 4.6%) with corrected SPEs. These findings have implications for water resource management and disaster risk reduction initiatives in Vietnam and other countries.","PeriodicalId":54801,"journal":{"name":"Journal of Hydroinformatics","volume":null,"pages":null},"PeriodicalIF":2.2,"publicationDate":"2024-07-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141809182","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Sanjit Kumar, B. Kirar, Mayank Agarwal, Vishal Deshpande, Upaka S. Rathnayake
Sewer systems are usually built with a self-cleaning system that keeps the bottom of the channel free of sediment to lessen the effects of the constant buildup of sediment particles. Because of this, it is important to accurately predict the particle Froude number (Fr) when making sewer systems. For the prediction of Fr, five different sets of input variables were looked at. For the training and testing of the machine learning (ML) model, we used 10-fold cross-validation methodologies to prevent overfitting. M5Prime (M5P) model as a standalone and Bagging-M5P as a hybrid model were utilized, and the results were compared with the empirical equations proposed in the literature. Models perform best when all input variables are used for training and testing of models. The hybrid BA-M5P model performed better than the M5P model and empirical equations. We performed sensitivity analysis and compared the result based on MAE and MSE value, and we found sediment concentration (Svc) is the most important variable to predict the particle Froude number under non-deposition with deposited bed by best performing model BA-M5P. Hence, for the self-cleaning system, we prefer the BA-M5P ML model 26 with Svc the most required variable.
{"title":"Efficient functioning of a sewer system: application of novel hybrid machine learning methods for the prediction of particle Froude number","authors":"Sanjit Kumar, B. Kirar, Mayank Agarwal, Vishal Deshpande, Upaka S. Rathnayake","doi":"10.2166/hydro.2024.065","DOIUrl":"https://doi.org/10.2166/hydro.2024.065","url":null,"abstract":"\u0000 Sewer systems are usually built with a self-cleaning system that keeps the bottom of the channel free of sediment to lessen the effects of the constant buildup of sediment particles. Because of this, it is important to accurately predict the particle Froude number (Fr) when making sewer systems. For the prediction of Fr, five different sets of input variables were looked at. For the training and testing of the machine learning (ML) model, we used 10-fold cross-validation methodologies to prevent overfitting. M5Prime (M5P) model as a standalone and Bagging-M5P as a hybrid model were utilized, and the results were compared with the empirical equations proposed in the literature. Models perform best when all input variables are used for training and testing of models. The hybrid BA-M5P model performed better than the M5P model and empirical equations. We performed sensitivity analysis and compared the result based on MAE and MSE value, and we found sediment concentration (Svc) is the most important variable to predict the particle Froude number under non-deposition with deposited bed by best performing model BA-M5P. Hence, for the self-cleaning system, we prefer the BA-M5P ML model 26 with Svc the most required variable.","PeriodicalId":54801,"journal":{"name":"Journal of Hydroinformatics","volume":null,"pages":null},"PeriodicalIF":2.2,"publicationDate":"2024-07-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141808859","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Seith N. Mugume, James Murungi, Philip M. Nyenje, J. Sempewo, John Okedi, Johanna Sörensen
The need to develop new and computationally efficient artificial intelligence models that accurately simulate river flows in data-scarce regions, considering not only current but also projected future climate change conditions is vital. In this study, a hybrid artificial neural network (ANN) model that combines HEC-HMS and the feed-forward neural network (FFNN) was developed in the Python programming language and applied to simulate future stream flows in the River Mayanja catchment in Central Uganda. The study results suggest that the performance of the validated hybrid HEC-HMS-ANN model during calibration and validation (NSE and R2 > 0.99) was more superior to the corresponding performance obtained using individual HEC-HMS (NSE and R2 > 0.50), MIKE HYDRO (NSE and R2 > 0.42), and ANN models (NSE and R2 > 0.56). Using the developed hybrid ANN model, future average daily stream flows are projected to increase by up to 17.3% [2.2–39.5%] and 18.5% [0.8–42.7%] considering the SSP2-4.5 and SSP5-8.5 future climate change scenarios. The study demonstrates that well-trained hybrid ANN models could provide more computationally efficient models for the simulation of future stream flow and for undertaking water resource assessments in catchments with limited in situ observed data.
{"title":"Development and application of a hybrid artificial neural network model for simulating future stream flows in catchments with limited in situ observed data","authors":"Seith N. Mugume, James Murungi, Philip M. Nyenje, J. Sempewo, John Okedi, Johanna Sörensen","doi":"10.2166/hydro.2024.066","DOIUrl":"https://doi.org/10.2166/hydro.2024.066","url":null,"abstract":"\u0000 \u0000 The need to develop new and computationally efficient artificial intelligence models that accurately simulate river flows in data-scarce regions, considering not only current but also projected future climate change conditions is vital. In this study, a hybrid artificial neural network (ANN) model that combines HEC-HMS and the feed-forward neural network (FFNN) was developed in the Python programming language and applied to simulate future stream flows in the River Mayanja catchment in Central Uganda. The study results suggest that the performance of the validated hybrid HEC-HMS-ANN model during calibration and validation (NSE and R2 > 0.99) was more superior to the corresponding performance obtained using individual HEC-HMS (NSE and R2 > 0.50), MIKE HYDRO (NSE and R2 > 0.42), and ANN models (NSE and R2 > 0.56). Using the developed hybrid ANN model, future average daily stream flows are projected to increase by up to 17.3% [2.2–39.5%] and 18.5% [0.8–42.7%] considering the SSP2-4.5 and SSP5-8.5 future climate change scenarios. The study demonstrates that well-trained hybrid ANN models could provide more computationally efficient models for the simulation of future stream flow and for undertaking water resource assessments in catchments with limited in situ observed data.","PeriodicalId":54801,"journal":{"name":"Journal of Hydroinformatics","volume":null,"pages":null},"PeriodicalIF":2.2,"publicationDate":"2024-07-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141824278","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Yunshuo Cheng, Zhiwei Li, Guo‐An Yu, Weiwei Yao, Bang Chen
A recent discovery of two unique meandering streams near the Yarlung Tsangpo Grand Canyon facilitates the present study. Given the contrasting channel patterns compared with the surrounding bedrock and braided reaches, as well as their recent formation due to dam-induced topographic changes within the valley, this study offers critical insights into the formation and evolution processes of meandering channels. It is found that, first, the prolonged sedimentation process due to the backwater of the mainstream of the floodplain proves a material base for the formation of the meandering river. Proper bank strength provided by the floodplain (stratified layer of root-soil composite and silty clay) contrasts the stream from a braided pattern into a single-threaded pattern, then the alternate bar in the upstream preludes the meandering channel formation. The annual migration rate of the stream is consistent with other large-scale natural meandering rivers. Congruences and disparities with the analytical meandering migration model of the present stream (that the meandering path follows the Kinoshita curve with noticeable flatness but no skewness) highlight the complex interplay of local factors in shaping meandering processes, offering valuable insights into both the unique characteristics of the Cuoka streams and the broader principles governing meander formation.
{"title":"Formation of meandering streams in a young floodplain within the Yarlung Tsangpo Grand Canyon in the Tibetan Plateau","authors":"Yunshuo Cheng, Zhiwei Li, Guo‐An Yu, Weiwei Yao, Bang Chen","doi":"10.2166/hydro.2024.171","DOIUrl":"https://doi.org/10.2166/hydro.2024.171","url":null,"abstract":"\u0000 \u0000 A recent discovery of two unique meandering streams near the Yarlung Tsangpo Grand Canyon facilitates the present study. Given the contrasting channel patterns compared with the surrounding bedrock and braided reaches, as well as their recent formation due to dam-induced topographic changes within the valley, this study offers critical insights into the formation and evolution processes of meandering channels. It is found that, first, the prolonged sedimentation process due to the backwater of the mainstream of the floodplain proves a material base for the formation of the meandering river. Proper bank strength provided by the floodplain (stratified layer of root-soil composite and silty clay) contrasts the stream from a braided pattern into a single-threaded pattern, then the alternate bar in the upstream preludes the meandering channel formation. The annual migration rate of the stream is consistent with other large-scale natural meandering rivers. Congruences and disparities with the analytical meandering migration model of the present stream (that the meandering path follows the Kinoshita curve with noticeable flatness but no skewness) highlight the complex interplay of local factors in shaping meandering processes, offering valuable insights into both the unique characteristics of the Cuoka streams and the broader principles governing meander formation.","PeriodicalId":54801,"journal":{"name":"Journal of Hydroinformatics","volume":null,"pages":null},"PeriodicalIF":2.2,"publicationDate":"2024-07-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141826266","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Sanjit Kumar, Giuseppe Oliveto, Vishal Deshpande, Mayank Agarwal, Upaka S. Rathnayake
Forecasting the time-dependent scour depth (dst) is very important for the protection of bridge structures. Since scour is the result of a complicated interaction between structure, sediment, and flow velocity, empirical equations cannot guarantee an advanced accuracy, although they would preserve the merit of being straightforward and physically inspiring. In this article, we propose three ensemble machine learning methods to forecast the time-dependent scour depth at piers: extreme gradient boosting regressor (XGBR), random forest regressor (RFR), and extra trees regressor (ETR). These models predict the scour depth at a given time, dst, based on the following main variables: the median grain size, d50, the sediment gradation, σg, the approach flow velocity, U, the approach flow depth y, the pier diameter Dp, and the time t. A total of 555 data points from different studies have been taken for this research work. The results indicate that all the proposed models precisely estimate the time-dependent scour depth. However, the XGBR method performs better than the other methods with R = 0.97, NSE = 0.93, AI = 0.98, and CRMSE = 0.09 at the testing stage. Sensitivity analysis exhibits that the time-dependent scour depth is highly influenced by the time scale.
{"title":"Forecasting of time-dependent scour depth based on bagging and boosting machine learning approaches","authors":"Sanjit Kumar, Giuseppe Oliveto, Vishal Deshpande, Mayank Agarwal, Upaka S. Rathnayake","doi":"10.2166/hydro.2024.047","DOIUrl":"https://doi.org/10.2166/hydro.2024.047","url":null,"abstract":"\u0000 Forecasting the time-dependent scour depth (dst) is very important for the protection of bridge structures. Since scour is the result of a complicated interaction between structure, sediment, and flow velocity, empirical equations cannot guarantee an advanced accuracy, although they would preserve the merit of being straightforward and physically inspiring. In this article, we propose three ensemble machine learning methods to forecast the time-dependent scour depth at piers: extreme gradient boosting regressor (XGBR), random forest regressor (RFR), and extra trees regressor (ETR). These models predict the scour depth at a given time, dst, based on the following main variables: the median grain size, d50, the sediment gradation, σg, the approach flow velocity, U, the approach flow depth y, the pier diameter Dp, and the time t. A total of 555 data points from different studies have been taken for this research work. The results indicate that all the proposed models precisely estimate the time-dependent scour depth. However, the XGBR method performs better than the other methods with R = 0.97, NSE = 0.93, AI = 0.98, and CRMSE = 0.09 at the testing stage. Sensitivity analysis exhibits that the time-dependent scour depth is highly influenced by the time scale.","PeriodicalId":54801,"journal":{"name":"Journal of Hydroinformatics","volume":null,"pages":null},"PeriodicalIF":2.2,"publicationDate":"2024-07-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141829109","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Matthew Hayslep, Edward Keedwell, R. Farmani, Joshua Pocock
Both minimum night flow (MNF) and pipe failures are common ways of understanding leakage within water distribution networks (WDNs). This article takes a data-driven approach and applies linear models, random forests, and neural networks to MNF and pipe failure prediction. First, models are trained to estimate the historic average MNF for over 800 real-world DMAs from the UK. Features for this problem are constructed from pipe records which detail the length, diameter, volume, age, material, and number of customer connections of each pipe. The results show that 65% of the variation in historic average MNF can be explained using these factors alone. Second, a novel method is proposed to deconstruct the models' predictions into a leakage contribution score (LCS), estimating how each individual pipe in a DMA has contributed to the MNF. In order to validate this novel approach, the LCS values are used to classify pipes based on historic pipe failure and are compared against models directly trained for this. The results show that the LCS performs well at this task, achieving an AUC of 0.71. In addition, it is shown that both LCS and directly trained models agree in many cases on an example real-world DMA.
{"title":"An explainable machine learning approach to the prediction of pipe failure using minimum night flow","authors":"Matthew Hayslep, Edward Keedwell, R. Farmani, Joshua Pocock","doi":"10.2166/hydro.2024.204","DOIUrl":"https://doi.org/10.2166/hydro.2024.204","url":null,"abstract":"\u0000 \u0000 Both minimum night flow (MNF) and pipe failures are common ways of understanding leakage within water distribution networks (WDNs). This article takes a data-driven approach and applies linear models, random forests, and neural networks to MNF and pipe failure prediction. First, models are trained to estimate the historic average MNF for over 800 real-world DMAs from the UK. Features for this problem are constructed from pipe records which detail the length, diameter, volume, age, material, and number of customer connections of each pipe. The results show that 65% of the variation in historic average MNF can be explained using these factors alone. Second, a novel method is proposed to deconstruct the models' predictions into a leakage contribution score (LCS), estimating how each individual pipe in a DMA has contributed to the MNF. In order to validate this novel approach, the LCS values are used to classify pipes based on historic pipe failure and are compared against models directly trained for this. The results show that the LCS performs well at this task, achieving an AUC of 0.71. In addition, it is shown that both LCS and directly trained models agree in many cases on an example real-world DMA.","PeriodicalId":54801,"journal":{"name":"Journal of Hydroinformatics","volume":null,"pages":null},"PeriodicalIF":2.7,"publicationDate":"2024-06-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141346544","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The increasing availability of condition assessment data highlights the challenge of managing data imbalance in the asset management of aging infrastructure. Aging sewer pipes pose significant threats to health and the environment, underscoring the importance of proactive management practices to enhance asset maintenance and mitigate associated risks. While machine learning (ML) models are widely employed to model the complex deterioration process of sewer pipes, they face performance limitations when trained on imbalanced condition grade data. This paper addresses this issue by proposing a novel approach using conditional generative adversarial network (cGAN) for data augmentation. By generating synthetic data for minority classes, the skewed distribution of the sewer dataset is balanced, facilitating more robust and accurate predictive models. The utility of the proposed method is evaluated by training different ML classifiers, including neural network (NN), decision tree, quadratic discriminant analysis, Naïve Bayes, support vector machine (SVM), and K-nearest neighbor. Quadratic discriminant, Naïve Bayes, NN, and SVM classifiers demonstrated improvement. The cGAN-based data augmentation method also outperformed two other data imbalance handling techniques, random under-sampling, and cost-sensitive NN. Consequently, data generated by cGAN can effectively aid asset management by developing proactive classifiers that accurately predict pipes at a high risk of failure.
{"title":"Data augmentation using conditional generative adversarial network (cGAN): applications for sewer condition classification and testing using different machine learning techniques","authors":"Haile Woldesellasse, Solomon Tesfamariam","doi":"10.2166/hydro.2024.135","DOIUrl":"https://doi.org/10.2166/hydro.2024.135","url":null,"abstract":"\u0000 The increasing availability of condition assessment data highlights the challenge of managing data imbalance in the asset management of aging infrastructure. Aging sewer pipes pose significant threats to health and the environment, underscoring the importance of proactive management practices to enhance asset maintenance and mitigate associated risks. While machine learning (ML) models are widely employed to model the complex deterioration process of sewer pipes, they face performance limitations when trained on imbalanced condition grade data. This paper addresses this issue by proposing a novel approach using conditional generative adversarial network (cGAN) for data augmentation. By generating synthetic data for minority classes, the skewed distribution of the sewer dataset is balanced, facilitating more robust and accurate predictive models. The utility of the proposed method is evaluated by training different ML classifiers, including neural network (NN), decision tree, quadratic discriminant analysis, Naïve Bayes, support vector machine (SVM), and K-nearest neighbor. Quadratic discriminant, Naïve Bayes, NN, and SVM classifiers demonstrated improvement. The cGAN-based data augmentation method also outperformed two other data imbalance handling techniques, random under-sampling, and cost-sensitive NN. Consequently, data generated by cGAN can effectively aid asset management by developing proactive classifiers that accurately predict pipes at a high risk of failure.","PeriodicalId":54801,"journal":{"name":"Journal of Hydroinformatics","volume":null,"pages":null},"PeriodicalIF":2.7,"publicationDate":"2024-06-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141345563","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Cascade gates and pumps are common hydraulic structures in the open-canal section of water transfer projects, characterized by high energy consumption and substantial costs, causing it challenging to regulate. By implementing cascade gates regulation to control the hydraulic process, lift distribution of pump stations can be optimized, thus enhancing operational efficiency and reducing energy consumption. However, the selection of control models and parameter optimization is difficult because hydraulic processes is nonlinear, high-dimensional, large hysteresis, strong coupling, and time-varying. This study considers minimum energy consumption of pump station as the regulation objective and employs reinforcement learning (RL) algorithm for the optimization regulation (OR) within a typical canal section of the Jiaodong Water Transfer Project. Our results demonstrate that after regulating, OR can precisely control the water level to achieve the high efficiency lift interval of pump station, enhancing efficiency by 4.12–6.02% compared to previous operation. Moreover, using optimized hyperparameters group, the RL model proves robust under different work conditions. The proposed method is suitable for complex hydraulic process, highlighting its potential to support more effective decision-making in water resources regulation.
{"title":"Research on low-energy consumption automatic real-time regulation of cascade gates and pumps in open-canal based on reinforcement learning","authors":"Tian Gan, Yunzhong Jiang, Hongli Zhao, Junyan He, Hao Duan","doi":"10.2166/hydro.2024.020","DOIUrl":"https://doi.org/10.2166/hydro.2024.020","url":null,"abstract":"\u0000 \u0000 Cascade gates and pumps are common hydraulic structures in the open-canal section of water transfer projects, characterized by high energy consumption and substantial costs, causing it challenging to regulate. By implementing cascade gates regulation to control the hydraulic process, lift distribution of pump stations can be optimized, thus enhancing operational efficiency and reducing energy consumption. However, the selection of control models and parameter optimization is difficult because hydraulic processes is nonlinear, high-dimensional, large hysteresis, strong coupling, and time-varying. This study considers minimum energy consumption of pump station as the regulation objective and employs reinforcement learning (RL) algorithm for the optimization regulation (OR) within a typical canal section of the Jiaodong Water Transfer Project. Our results demonstrate that after regulating, OR can precisely control the water level to achieve the high efficiency lift interval of pump station, enhancing efficiency by 4.12–6.02% compared to previous operation. Moreover, using optimized hyperparameters group, the RL model proves robust under different work conditions. The proposed method is suitable for complex hydraulic process, highlighting its potential to support more effective decision-making in water resources regulation.","PeriodicalId":54801,"journal":{"name":"Journal of Hydroinformatics","volume":null,"pages":null},"PeriodicalIF":2.7,"publicationDate":"2024-06-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141350743","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Hydrological modeling is one of the most complicated tasks in sustainable water resources management, particularly in terms of predicting rainfall. Predicting rainfall is critical to build a sustainable society in terms of hydropower operations, agricultural planning, and flood control. In this study, a hybrid model based on the integration of k-nearest neighbor (KNN), XGBoost (XGB), decision tree (DCT), and Random Forest (RF) has been developed and implemented for forecasting daily rainfall for the first time at Sydney airport, Australia. Daily rainfall, temperature, evaporation, and humidity have been selected as input parameters. Three statistical measurements, namely, root mean square error (RMSE), Coefficient of determination (R2), mean absolute error (MAE), and Normalized Root Mean Square Error (NRMSE) have been utilized in order to check the accuracy of the proposed model. A sensitivity analysis was conducted, and the results indicated that for the purpose of prediction, the temperature, humidity, and evaporation were highly sensitive to the rainfall data. According to the results, the developed hybrid model was capable of predicting daily rainfall with high performance for both training and testing parts with RMSE = 0.124, R2 = 0.999, MAE = 0.007, NRMSE = 0.04 and RMSE = 1.246, R2 = 0.991, MAE = 0.109, NRMSE = 0.339, respectively.
{"title":"Forecasting daily rainfall in a humid subtropical area: an innovative machine learning approach","authors":"M. Mohammed, S. Latif","doi":"10.2166/hydro.2024.016","DOIUrl":"https://doi.org/10.2166/hydro.2024.016","url":null,"abstract":"\u0000 \u0000 Hydrological modeling is one of the most complicated tasks in sustainable water resources management, particularly in terms of predicting rainfall. Predicting rainfall is critical to build a sustainable society in terms of hydropower operations, agricultural planning, and flood control. In this study, a hybrid model based on the integration of k-nearest neighbor (KNN), XGBoost (XGB), decision tree (DCT), and Random Forest (RF) has been developed and implemented for forecasting daily rainfall for the first time at Sydney airport, Australia. Daily rainfall, temperature, evaporation, and humidity have been selected as input parameters. Three statistical measurements, namely, root mean square error (RMSE), Coefficient of determination (R2), mean absolute error (MAE), and Normalized Root Mean Square Error (NRMSE) have been utilized in order to check the accuracy of the proposed model. A sensitivity analysis was conducted, and the results indicated that for the purpose of prediction, the temperature, humidity, and evaporation were highly sensitive to the rainfall data. According to the results, the developed hybrid model was capable of predicting daily rainfall with high performance for both training and testing parts with RMSE = 0.124, R2 = 0.999, MAE = 0.007, NRMSE = 0.04 and RMSE = 1.246, R2 = 0.991, MAE = 0.109, NRMSE = 0.339, respectively.","PeriodicalId":54801,"journal":{"name":"Journal of Hydroinformatics","volume":null,"pages":null},"PeriodicalIF":2.7,"publicationDate":"2024-06-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141360708","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：