The proper arrangement of taxiways is one of the most important jobs for airport operations. Based on simulation technique, the problem that latency time was hard to be presented by mathematics analytic expressions is solved in this paper. By transforming operating limitations into constraint conditions, taxiway optimization model aims at the shortest weighting taxiing time is established and the genetic algorithm is used in solving the model. To improve the efficiency, proper deletions of scheme set are conducted by the rules. At last, the validity of the simulation model and the algorithm are illustrated by an example.