This paper proposes an adaptive genetic algorithm of symbol array chromosome coding method based on adaptive mutation probability and penalty function method, in order to solve the problems of dimension limitation and schema deceptive for the technical preparation procedure optimization of constraint aircraft. For the technical preparation procedure optimization of a certain-type aircraft, it took the total time consumption at different stages of technical preparation as the objective function and the total reliability of technical preparation as constraint conditions. The genetic algorithm evolution process of static variation probability and adaptive variation probability was compared. The algorithm can effectively used for constrain process optimization with large scale calculation and strong deception, and acquire optimal technology preparation process plan. Keywords: genetic algorithms; adaptive probability