Abstract: Based on the multi-objective optimization problems in the aircraft conceptual parameter design, improved multi-objective genetic algorithms were proposed. The "Pareto front "of Multi-objective optimization (MO) was tried to seek by using inherent parallelism of genetic algorithms (GA) while emphasizing the Pareto optimal conception. Different improved genetic algorithms were applied to a same two-objective optimization arterial airliner conceptual parameter optimization design, where both the ratio of lift to drag in cruise and the useful load fraction were asked to be maximized. Through general analysis and comparison to solutions of different algorithms, it can conclude that the Pareto optimal results of the non-dominated sorting genetic algorithm(NSGA) are better than the improved vector-evaluated genetic algorithm(VEGA) and the random-weight genetic algorithm(RWGA), while the results of VEGA and RWGA are equal.