一种考虑飞行轨迹不确定性的碰撞风险评估方法

With the development of intelligent and unmanned transportation, the automation management of the transportation system has become an inevitable trend, and collision risk is of paramount consideration in traffic management automation. In this paper, we propose a collision risk evaluation method, which takes collision probability as a safety-evaluation index, and hence, can control collision risks. We developed the proposed methodology in a simple form that consists of three steps for data preprocessing, spatiotemporal discretization, and probability calculation. First, we perform coordinate unification and time-alignment for the given trajectory data. Then, we divide the trajectory into many small segments according to the interpolation interval and divide the uncertainty cuboid into many grid cells according to the given safe distance. Finally, we calculate the instantaneous collision probability according to the probability distribution model, and then, we obtain the collision probability of the entire flight period. Simulation results reveal that the method has convergence for the interpolation interval. We also discuss the effects of safe distance on the calculation result of collision probability. The proposed method has a wide range of application scenarios, which can be applied to traffic management decision support tools of various types of vehicles as long as an appropriate safe distance is adopted.