Multi-Gene Genetic Programming Regression Model for Prediction of Transient Storage Model Parameters in Natural Rivers

Abstract

A Transient Storage Model (TSM), which considers the storage exchange process that induces an abnormal mixing phenomenon, has been widely used to analyze solute transport in natural rivers. The primary step in applying TSM is a calibration of four key parameters: flow zone dispersion coefficient (Kf), main flow zone area (Af), storage zone area (As), and storage exchange rate (α); by fitting the measured Breakthrough Curves (BTCs). In this study, to overcome the costly tracer tests necessary for parameter calibration, two dimensionless empirical models were derived to estimate TSM parameters, using multi-gene genetic programming (MGGP) and principal components regression (PCR). A total of 128 datasets with complete variables from 14 published papers were chosen from an extensive meta-analysis and were applied to derivations. The performance comparison revealed that the MGGP-based equations yielded superior prediction results. According to TSM analysis of field experiment data from Cheongmi Creek, South Korea, although all assessed empirical equations produced acceptable BTCs, the MGGP model was superior to the other models in parameter values. The predicted BTCs obtained by the empirical models in some highly complicated reaches were biased due to misprediction of Af. Sensitivity analyses of MGGP models showed that the sinuosity is the most influential factor in Kf, while Af, As, and α, are more sensitive to U/U*. This study proves that the MGGP-based model can be used for economic TSM analysis, thus providing an alternative option to direct calibration and the inverse modeling initial parameters.

Identification Framework of Contaminant Spill in Rivers Using Machine Learning with Breakthrough Curve Analysis

“…In addition, Table 5 gives the estimated TSM parameters of each sub reach according to the streamflow scenarios. Notably, reasonable range values were calculated when compared with TSM parameters reported in previous studies [38,39,46]. Moreover, the Froude number (Equation 30) of all streamflow scenarios represents that only subcritical flows were generated.…”

“…To overcome this limitation, empirical equations for TSM parameters have recently been derived from a meta-analysis of river mixing tracer tests [39,47]. From these equations, the TSM parameters can be estimated using easily measurable hydraulic and geometry variables.…”

“…Thus, in this study, Principal Component Regression (PCR)-based empirical equations for TSM parameters were used to estimate TSM parameters. Equation 3gives the equation, while Table 1 gives the derived power [39]:…”

“…For this objective, the Transient Storage zone Model (TSM) was used as a contaminant transport model to generate the contaminant spill scenarios. This model has been successfully used to reproduce the breakthrough curve (BTC), which is a time-concentration curve of the contaminant that represents the mixing processes with advection, shear dispersion, and storage effect in the river [35][36][37][38][39]. The second objective was to develop Machine Learning (ML) models for the identification of both spill location and mass of the contaminant source in rivers.…”

Multi-Gene Genetic Programming-Based Identification of a Dynamic Prediction Model of an Overhead Traveling Crane

“…Genetic programming for symbolic regression using a nonlinear least squares method for parameter estimation was studied in [ 20 ] and shown to improve performance on a wide array of symbolic regression problems. Applications of genetic programming in the development of prediction models include [ 21 ] developing a model predictive control based on a model identified by genetic programming; [ 22 ] the identification of a prediction model for the time dependent total creep in concrete; and the development of transient storage models [ 23 ].…”