基于GLS导航的超长跑道OAS面评估方法研究

CAAC had put forward the plan of applying new navigation techniques in 2009. RNP and GLS are the very important part of this plan. These techniques can improve the operational efficiency of the complicated high elevation airport and enhance aircraft operational safety level. However, the operational standards of RNP are higher than the category I(CAT-I) approach. Thus, RNP plus GLS mode is the most effective navigation method for these airports in the future. Using CAT-I approach, it must assess the obstacles of vicinity which close to the track using obstacle assessment surface (OAS) model, so as the safety level satisfies the aviation regulators which are required by the ICAO. But OAS model can’t give out a clear result which the distance between localizer and runway’s threshold is over 4500 meters. The user sometimes can only accept this result by their practiced experience. This may produce unsafe flight results. Up to now, there is no regulator or notice on this problem in CAAC or ICAO. OAS model is the simplified version of CRM model. This project try to reconstruct the CRM model that the distance is less than 4500 meters, and construct the CRM model that the distance is more than 4500 meters. Meanwhile, we also construct the over length runway OAS model. The plot of this research is going to collect recorder’s data of QAR, then analysis these data which don’t satisfy the normal distributing state. The way will use statistic method. At last, the goal of this project is putting the density function of flight deviation along the pre-defined track based on the CAT-I approach.

从2009年起中国民航开始推广航行新技术，其中基于性能的导航（RNP）与卫星导航着陆系统（GLS）是重点应用的技术，以提高复杂机场运行效率和安全裕度。但RNP达不到I类进近的运行标准，因此未来所需要的模式是RNP衔接I类精密进近。使用I类精密着陆系统时，需用OAS对航迹周边的障碍物进行碰撞安全评估，以保障运行水平达到安全要求。但当航向台距跑道入口超过4500米时，OAS仅能提供4500米的评估数据，这给安全运行带来了相当的隐患。中国民航以及国际民航组织对此情况目前尚无明确的法规解释。OAS是CRM的简化模型，本项目拟还原4500米以内的CRM模型，并构建超过4500米的CRM模型及OAS模型。拟用的技术方案就是采集机载QAR中记录的相关航行参数数据，并对飞机沿航迹飞行的偏差不满足正态分布的数据进行数值分析、归纳统计，构建I类精密进近航迹偏差的概率分布函数。

近年来，我国修建了大量高原机场，其跑道长度超出了障碍物评价面（Obstacle Assessment Surface，OAS）规范的上限，而ICAO对在此类情况的规定却较为笼统，有必要论证其安全性；此外，净空保护与城市建设的矛盾日益突出，而现行的OAS却较为保守，故有必要改进障碍物风险评价方法。本项目针对高原地区障碍物评价模式进行研究，重点探讨OAS、碰撞风险模型（Collision risk model，CRM）和精密进近航段航迹偏差数据的分布。该技术可以实现1）提出一种基于CRM等概率线构建OAS面的方法；2）基于CRM等概率线构建障碍物椭圆评价面，可减缓对机场附近障碍物不必要的限制。3）研究精密进近航段的航迹分布，使用高斯混合模型（Gauss Mixture Model，GMM）和最大期望算法（Expectation Maximization,EM），得到拟合度较高的概率密度函数，并基于此建立安全评价模型用以计算障碍物碰撞风险概率。以期望在GLS进近方式下，应用该技术可以实现对机场周围障碍物的碰撞风险进行安全可靠地评价。