Abstract: The development of a new kind of thermal protection material which combines non-ablation, thermal insulation, light-weight and reliability is the core technology for the development of new hypersonic vehicle. The interfaces between different materials are the most serious areas of thermal mismatch. The design of the gradient layers is the key to the realization of the composite material. To improve the performances of the composite material, deterministic optimization of transitional form of material components of the gradient layers and the material size parameters was made by using simplex method and neighborhood cultivation multi-objective genetic algorithm. Six Sigma method was further used to optimize the performances of the composite material considering uncertainty of material properties and load conditions and taking the above optimal solution as initial value. The optimization result shows that the optimization strategy can effectively improve the optimization efficiency while obtaining the optimal material performance parameters that meet the requirement of material reliability, which is especially important for calculating complex optimization problems.