Machine Learning and Urban Drainage Systems: State-of-the-Art Review

Abstract

In the last decade, machine learning (ML) technology has been transforming daily lives, industries, and various scientific/engineering disciplines. In particular, ML technology has resulted in significant progress in neural network models; these enable the automatic computation of problem-relevant features and rapid capture of highly complex data distributions. We believe that ML approaches can address several significant new and/or old challenges in urban drainage systems (UDSs). This review paper provides a state-of-the-art review of ML-based UDS modeling/application based on three categories: (1) operation (real-time operation control), (2) management (flood-inundation prediction) and (3) maintenance (pipe defect detection). The review reveals that ML is utilized extensively in UDSs to advance model performance and efficiency, extract complex data distribution patterns, and obtain scientific/engineering insights. Additionally, some potential issues and future directions are recommended for three research topics defined in this study to extend UDS modeling/applications based on ML technology. Furthermore, it is suggested that ML technology can promote developments in UDSs. The new paradigm of ML-based UDS modeling/applications summarized here is in its early stages and should be considered in future studies.

Modeling the 2D Inundation Simulation Based on the ANN-Derived Model with Real-Time Measurements at Roadside IoT Sensors

“…) by the proposed SM_EID_VIOT model at the four VIOT grids around the specific IoT sensors are treated as the corresponding simulations via the following equation: (10) Thus, the corresponding errors at the IoT sensors could be quantified based on the difference between the simulations and observations at the previous time steps during the rainstorm using Equations ( 8) and (9). Since the proposed SM_EID_VIOT model estimates the inundation depths at the VIOT grids with the measurements at the road IoT sensors, the errors of the estimated inundation depths at the current time are considered in quantifying the error of the inundation-depth estimates at the VIOT grids within the regional error correction.…”

“…Despite the hydraulic numerical models being applied in the flood simulation, their reliability and accuracy might be affected by the uncertainties in the requirement of sufficient observations, the complex model structures, hydrological/hydraulic features, and extensive computation time [3,4,8]. Recently, the artificial intelligence (AI) models have been comprehensively employed in flood-induced inundation based on machine learning (ML) techniques [9]. Of the relevant ML approaches commonly used, the two types of natural network (NN) methods, the convolutional NN (CNN) and artificial NN (ANN) models, are both more efficiently established for describing the nonlinear mathematic relationships by configuring the linear multi-layer network with all possible predictor variables under the multiple training algorithm.…”

“…With computation power increasing, the CNN-based and ANN-derived models are widely applied to the relevant hydrological/hydraulic analysis, especially in flood-related simulations/forecasts [1,3,4,[9][10][11][12][13][14]. In detail, the CNN-based model is derived via the neutral network, comprising the convolution, pooling, and fully connected layers, requiring extensive 2D data (i.e., gridded data) as datasets (e.g., the image and videos) for the model training and application [3,4,15]; accordingly, the CNN model can efficiently provide the single model output from the grid-format model inputs.…”

“…In detail, the CNN-based model is derived via the neutral network, comprising the convolution, pooling, and fully connected layers, requiring extensive 2D data (i.e., gridded data) as datasets (e.g., the image and videos) for the model training and application [3,4,15]; accordingly, the CNN model can efficiently provide the single model output from the grid-format model inputs. That is to say, the CNN-based model is advantageous concerning the 2D flood simulations for predicting or estimating the spatiotemporal flood-related variates (e.g., the inundation depths and corresponding area) [9]. The CNN-derived models have been proven to be efficiently implemented in the 2D flood forecast with the gridded inputs (such as the grid-based radar precipitation); however, without the mixed data, including the single dataset (e.g., at-site water level) and grid-format data under high-performance computing servers [3], it might cause the extensive computation time in the model training and application.…”

“…As a result, in this study, the 2D inundation simulation intends to be accomplished by combining the gridded inundation depths estimated by the ANN-derived model for different locations. Moreover, it is well known that the roadside IoT sensors have recently been implemented in the real-time measurement of the hydrological variables (e.g., water levels), and they are proven to be advantageous to water conservation computerization [1,9,21]. Therefore, to efficiently estimate the inundation depths at all ungauged locations (named virtual IoT sensors, VIOTs) with the real-time measurements at the roadside IoT sensors for delineating the flooding zones, this study aims to model the 2D inundation simulation based on the ANN-derived model; in detail, the relationship of the estimated inundation depths at each ungauged location with the real-time measurements at the roadside IoT sensors (VIOTs) could be established based on the ANN-derived model (called the SM_EID_VIOT model).…”