基于交互多模型（IMM）的航空器冲突探测算法

航空器之间的冲突预测预警是空中交通管制活动中的重要组成部分。针对广播式自动相关监视(ADS-B)航迹的监视精度低、目标跟踪模型与目标运动模型匹配速度较慢以及碰撞概率预测困难的问题,提出一种基于交互多模型(IMM)的航空器冲突概率预测算法。利用高斯分布积分的近似解析解来计算两航空器瞬时冲突概率。根据飞机匀速运动模型和转弯的动力学模型,仿真计算一段时间内两机冲突概率的变化。结果表明,可以通过交互式多模型(IMM)处理ADS-B数据,交互式多模型(IMM)相比其他模型计算结果具有一定时延,但计算所得的两航空器冲突概率预测值更精确。通过仿真对比各运动模型预测的两航空器冲突概率计算结果,证明了该算法的实用性。

Interactive Multi-Model (IMM) Based Aircraft Conflict Prediction Algorithm

Conflict prediction and warning between aircraft is an important part of air traffic control activities. To address the problems of low surveillance accuracy of broadcast automatic correlation surveillance (ADS-B) trajectory and slow matching of target tracking model with target motion model as well as the difficulty of collision probability prediction, an interactive multi-model (IMM) is proposed based aircraft conflict probability prediction algorithm. An approximate analytical solution of the Gaussian distribution integral is used to calculate the instantaneous collision probability of two aircraft. The change in the conflict probability of the two aircraft over a period of time is simulated based on the aircraft uniform motion model and the dynamics model of the turn. It is shown that the ADS-B data can be processed by the interactive multi-model (IMM), which has a certain time delay compared with other models, but the calculated conflict probability prediction values of the two aircraft are more accurate. The practicality of the algorithm is demonstrated by comparing the two-aircraft conflict probability calculation results predicted by each motion model through simulation.