Hourly and Daily Urban Water Demand Predictions Using a Long Short-Term Memory Based Model

Li, Mu; Zheng, Feifei; Tao, Ruoling; Zhang, Qingzhou; Kapelan, Zoran

doi:10.1061/(asce)wr.1943-5452.0001276

Cited by 56 publications

(15 citation statements)

References 36 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…They are the absolute percentage error (APE), the mean absolute percentage error (MAPE), the coefficient of determination ( 2 R ), the Nash-Sutcliffe model efficiency (NSE), and the Kling-Gupta Efficiency (KGE). These five metrics are selected due to their wide applications in assessing the model performance within the water resources domain (Mu et al, 2020). The APE between the th i observation i Y and its corresponding simulation i Y ˆ is defined as…”

Section: Metrics Used For Performance Evaluationmentioning

confidence: 99%

Real-time foul sewer hydraulic modelling driven by water consumption data from water distribution systems

et al. 2021

Self Cite

View full text Add to dashboard Cite

One way to address many issues (e.g., illicit inflows) within foul sewer systems (FSSs) is via real-time hydraulic models. However, a bottleneck within real-time FSS modelling is the lack of spatio-temporal manhole inflow data. To address this problem, this paper proposes a new method to develop real-time FSS models driven by water consumption data from associated water distribution systems (WDSs) that often have a proportionally larger number of sensors. Within the proposed method, the FSS manholes are integrated with the WDS water consumption nodes based on their underlying physical connections. An optimization approach is subsequently proposed to identify the transfer factor k between nodal water consumption and FSS manhole inflows based on historical observations. These identified k values combined with the acquired real-time nodal water consumption data from the WDS equipped with a dense network of sensors drive the FSS real-time modelling. The proposed method is applied to two real FSSs and results show that it can produce sewer flow and manhole water depth simulations matching well with observations at the monitoring locations with averaged R 2 , NSE and KGE (Kling-Gupta efficiency) around 0.99, 0.88 and 0.92, respectively. It is anticipated that real-time models developed by the proposed method can be useful for the efficient FSS management and operation.

show abstract

Section: Metrics Used For Performance Evaluationmentioning

confidence: 99%

Real-time foul sewer hydraulic modelling driven by water consumption data from water distribution systems

et al. 2021

Self Cite

View full text Add to dashboard Cite

show abstract

“…The main limitation of statistical models is that they must have a predetermined structure [29], making it difficult to find one mathematical function that would work well on different data [30]. Furthermore, statistical models often fail to effectively deal with complex data relationships; their prediction accuracy also decreases with an increase in the amount of data [31]. Other methods should be employed when dealing with big and complex data [32].…”

Section: Models For Predicting Water Demandmentioning

confidence: 99%

“…To evaluate each model's performance, the following three metrics were employed: mean squared error (MSE), mean absolute error (MAE), and coefficient of determination (R 2 ). These metrics were considered due to their wide use in studies on water demand prediction [5,[30][31][32].…”

Section: Model Testingmentioning

confidence: 99%

Water Demand Prediction Using Machine Learning Methods: A Case Study of the Beijing–Tianjin–Hebei Region in China

Shuang

Zhao

2021

Water

View full text Add to dashboard Cite

Predicting water demand helps decision-makers allocate regional water resources efficiently, thereby preventing water waste and shortage. The aim of this study is to predict water demand in the Beijing–Tianjin–Hebei region of North China. The explanatory variables associated with economy, community, water use, and resource availability were identified. Eleven statistical and machine learning models were built, which used data covering the 2004–2019 period. Interpolation and extrapolation scenarios were conducted to find the most suitable predictive model. The results suggest that the gradient boosting decision tree (GBDT) model demonstrates the best prediction performance in the two scenarios. The model was further tested for three other regions in China, and its robustness was validated. The water demand in 2020–2021 was provided. The results show that the identified explanatory variables were effective in water demand prediction. The machine learning models outperformed the statistical models, with the ensemble models being superior to the single predictor models. The best predictive model can also be applied to other regions to help forecast water demand to ensure sustainable water resource management.

show abstract

“…Long short term memory (LSTM) (Yu et al 2019;Mu et al 2020) model (Eqs. 9 and 10) is an improved recurrent neural network (RNN) model, so the generalization ability of LSTM is superior to that of RNN.…”

Section: Bidirectional Ensemble Learning Long Short Term Memory (Bellstm)mentioning

confidence: 99%

Multi-Model Coupling Water Demand Prediction Optimization Method for Megacities Based on Time Series Decomposition

Liu

Sang

Chang

et al. 2021

Water Resour Manage

View full text Add to dashboard Cite

The water supply in megacities can be affected by the living habits and population mobility, so the fluctuation degree of daily water supply data is acute, which presents a great challenge to the water demand prediction. This is because that non-stationarity of daily data can have a large influence on the generalization ability of models. In this study, the Hodrick-Prescott (HP) and wavelet transform (WT) methods were used to carry out decomposition of daily data to solve the non-stationarity problem. The bidirectional long short term memory (BLSTM), seasonal autoregressive integrated moving average (SARIMA) and Gaussian radial basis function neural network (GRBFNN) were developed to carry out prediction of different subseries. The ensemble learning was introduced to improve the generalization ability of models, and prediction interval was generated based on student's t-test to cope with the variation of water supply laws. This study method was applied to the daily water demand prediction in Shenzhen and cross-validation was performed. The results show that WT is superior to HP decomposition method, but maximum decomposition level of WT should not be set too high, otherwise the trend characteristics of subseries will be weakened. Although the corona virus disease 2019 (COVID-19) outbreak caused a variation in water supply laws, this variation is still within the prediction interval. The WT and coupling models accurately predict water demand and provide the optimal mean square error (0.17%), Nash-Sutcliffe efficiency (97.21%), mean relative error (0.1), mean absolute error (3.32%), and correlation coefficient (0.99).

show abstract

Hourly and Daily Urban Water Demand Predictions Using a Long Short-Term Memory Based Model

Cited by 56 publications

References 36 publications

Real-time foul sewer hydraulic modelling driven by water consumption data from water distribution systems

Real-time foul sewer hydraulic modelling driven by water consumption data from water distribution systems

Water Demand Prediction Using Machine Learning Methods: A Case Study of the Beijing–Tianjin–Hebei Region in China

Multi-Model Coupling Water Demand Prediction Optimization Method for Megacities Based on Time Series Decomposition

Contact Info

Product

Resources

About