With the high economic growth in China, the field of civil aviation transportation has witnessed a rapid development. But the efficiency of current centralized air-traffic management is limited, so billions of yuan was wasted by flight delay and fuel waste every year. It is a significant problem to ensure the aircraft keep safe separation as the solutions must meet the requirements of the feasibility and timeliness, but the existing algorithms can’t balance them in actual scenarios. In this paper, an algorithm based on stochastic optimal control is proposed to improve the search capability. The experimental studies which using real UAVs shown that the proposed algorithm provides robustness against uncertainties in the system and is suitable for real applications.