Evolutionary and swarm intelligence algorithms on pavement maintenance and rehabilitation planning

Abstract

ABSTRACT Maintenance and Rehabilitation (M&R) scheduling is one of the vital aspects of a pavement management system (PMS). This study aims to establish accurate M&R plans for a large-scale pavement network. To this intent, parameters affecting pavement deterioration were identified from the literature, then Random Forest Regression was employed to determine the effective features for pavement deterioration modelling. An accurate pavement deterioration function was generated by applying significant features. The most robust metaheuristic and evolutionary algorithms were selected and adjusted to solve the M&R scheduling optimisation problem, including the Water Cycle Algorithm (WCA), Arithmetic Optimisation Algorithm (AOA), Differential Evolutionary (DE), Ant Colony Optimisation (ACO), Particle Swarm Optimisation (PSO), and Genetic Algorithm (GA). The performance of the mentioned algorithms was compared to help researchers and decision-makers to select the appropriate algorithm for M&R scheduling optimisation. WCA and AOA showed to have the best performance among the adapted algorithms. Compared to AOA, DE, ACO, PSO, and GA, WCA's objective function was calculated to be 45%, 74%, 74%, 77% and 83% less, while its M&R cost was cheaper by 13%, 16%, 27%, 19%, and 18%, respectively.