Pipeline failure prediction in water distribution networks using evolutionary polynomial regression combined with K-means clustering

Abstract

This paper presents a new approach for improving pipeline failure predictions by combining a datadriven statistical model, i.e. Evolutionary Polynomial Regression (EPR), with k-means clustering. The EPR is used for prediction of pipe failures in case iron pipes based on length, diameter and age of pipes as explanatory factors. Individual pipes are aggregated using their attributes of age, diameter and soil type to create homogenous groups of pipes. The k-means clustering is employed to partition input data into a number of clusters for individual EPR models. The proposed approach was demonstrated by application to a water distribution network in the UK. The prediction accuracy was evaluated using a cross-validation technique. Results show the proposed approach is able to significantly reduce the error of pipe failure predictions especially in the case of a large number of failures. The prediction models were used to calculate the failure rate of individual pipes for rehabilitation planning.

Comparison of Statistical and Machine Learning Models for Pipe Failure Modeling in Water Distribution Networks

“…Three statistical models, including Linear Regression, Poisson Regression, and EPR are used to estimate the number of expected failures in pipe groups. These models are selected because they produce explicit polynomial expressions, which provide a high level of correlation between input variables and the dependent variable [11,14]. Linear Regression is an extension of regression analysis that includes independent variables as explanatory in a predictive equation.…”

“…+ β j X j , X j are the independent variables, β j are the coefficients to be estimated, and k i,t is the number of failure events. EPR is a hybrid regression method that combines conventional regression techniques and genetic programming, producing a range of equations in trade-off between the number of polynomial terms and accuracy [14]. EPR consists of two main stages: (1) the exploration of the best model structure using a multiobjective genetic algorithm and (2) the estimation for parameters for an assumed model structure using the least-squares method [5].…”

“…The required inputs are the data samples, the number of k clusters, and the stopping condition. The formulation of the clusters is based on the principle of maximizing intracluster similarity and minimizing intercluster similarity [14]. Therefore, the algorithm starts from an initial distribution of clusters' centroids, such that they are as distant as possible from each other, and determines which centroid is closest to each data point.…”

“…The models are trained on 70% of the available data and tested on the remaining 30%. The K-fold cross-validation technique is used to minimize the risk of overfitting [14]. This technique divides the dataset randomly into k-partitions and, at each step, it uses one partition for testing and the rest for training [33].…”

“…Statistical models use available historical breakage data to identify pipe failure patterns [13]. These models are capable of linking failure patterns to the pipe descriptive variables (e.g., diameter, age, and length) [14]. Particularly, Alvisi and Franchini [15] used two probabilistic models, namely the Weibull Exponential model and the Weibull Proportional Hazard model, to predict the number of failures in a pipe over an observation period using pipe diameter, age, and length.…”