An improved greedy correction method is advanced for overcome the flaw of greedy transform method adopted by hybrid genetic algorithm (HGA). And based on steady state reproduction strategy, the choice method of random selection is advanced. These new methods are combined with genetic algorithm to propose a high-efficient hybrid genetic algorithm (IHGA), and new algorithm was used to solve large-scale 0-1 knapsack problem. By many simulation experiments, IHGA algorithm is compared with HGA algorithm, and how the mutation probability affect the performance of the new algorithm has been studied. The experimental results show that the new algorithm has higher convergent speed and better optimization capability.