Optimum Genetic Algorithm Structure Selection in Pavement Management

Pavement management encompasses a wide range of tasks from data collection and data processing to maintenance management and life cycle cost analysis. This study concentrates on maintenance management at a project level and a network level. A decision support system provides optimum maintenance actions over time to enhance the performance of a pavement network and prolong its life span. A heuristic method i.e., Genetic Algorithm is applied to tackle this optimization problem. Due to complexity of the problem, applying an optimum Genetic Algorithm structure results in significant enhancements in the Genetic Algorithm procedure and saves computation time which has not been received enough attention to date by researchers. An experimental design is conducted to investigate the optimum Genetic Algorithm structure for solving a pavement maintenance problem. Since the current Genetic Algorithm software is not suitable to run the experiment, an accurate spreadsheet program has been developed for this purpose to conduct the experiment. Two types of objective functions have been applied: single objective functions (minimizing cost) vs. multiple objective functions (minimizing cost and maximizing benefit).