SENSITIVITY ANALYSIS OF GENETIC ALGORITHM BASED NETWORK OPTIMIZATION SYSTEM WITH MULTIPLE OBJECTIVES

摘要

The implementation of Genetic Algorithm (GA) based Network Optimization System (NOS) provides a state-of-practice optimization tool for the preservation and rehabilitation of pavements, based on application of intelligent computation and the classic NOS. To have more understanding of the influence of input parameters on results of GA based NOS, a sensitivity analysis is undertaken in the paper. A total of four inputs (period length, current conditions, standard condition and action unit cost) are evaluated by studying the effect of each input on three output measures (total cost, total performance rating and objective value) resulting in 13 simulations.