Facing the complexity and variability of electronic circuit design, evolvable hardware has gradually become one of the effective methods to design hardware. Algorithm influences the direction and progress of evolution. In order to improve algorithm and solve the current problem that the evolutionary time is long, the convergence speed is slow and the computation is large. This paper proposed an improved stages hybridization genetic algorithm. Using evolvable hardware platform, improved algorithm adopted the stages strategy and simulated the natural population hybridization based on genetic principle. The results show that the improved algorithm has better evolution results than the standard algorithm, whose convergent speed is faster and evolutionary cycle is shorter. Offspring circuits have a higher adaptability. Improved algorithm reduces calculation, and stages parallel strategy shortens time consuming and saves resources. Finally the optimal values of grouping number N and evolutionary generation T are found to improve the algorithm accuracy, which is beneficial to the whole evolution.