一种结合多重插补的ADS-B数据筛选算法

为了提高广播式自动相关监视(Automatic Dependent Surveillance-Broadcast,ADS-B)报文质量,解决由于地面站多重覆盖、硬件设施配置、运行状态等原因引起的报文在数据重复性、数据完整性和实时性上存在的问题,采用多条件直接起始法建立航迹,采用多项约束条件筛选进入优选的报文,并基于专家评级法对报文中表征数据项完整性的权重参数进行估计,结合报文的位置精度和完好性参数综合计算得出ADS-B数据质量指标并将其作为挑选依据。对于ADS-B报文质量极度不理想的特殊情况,使用多重插补(Multiple Imputation,MI)算法对其进行补全处理,综合以上步骤挑选出高质量和高可靠性的ADS-B更新点迹。仿真计算证明该方法可有效剔除重复覆盖造成的重复报文和网络延迟造成的历史数据,防止航迹点回跳,提高数据质量和监视效率。目前,该方法已实际应用于地面站和传输链路条件均不理想的场合,大量实践结果表明该方法可用于保证输出高实时性的、周期性的、稳定平滑的、可靠的ADS-B数据。

An ADS-B data filtering algorithm combined with multiple imputation

To improve the quality of automatic dependent surveillance-broadcast(ADS-B) reports and solve the problems of data repeatability,data integrity,and real-time performance caused by multiple coverages of the ground stations,hardware configuration,operation status,and other reasons,the direct start method with multi-constraint is used to establish the track,and multiple constraints are used to filter these message to be optimized.The weight parameters of the integrity of data items in the message are estimated based on expert rating method,and the ADS-B data quality index is obtained based on the position accuracy and integrity parameters of the message.For the special case that the quality of the ADS-B report is extremely bad.The multiple imputation(MI) methods are used to complete it,and above steps are combined to select the ADS-B update track with high quality and reliability.The simulation results show that the method can effectively eliminate the duplicate messages caused by repeated coverage and the historical data caused by network delay,prevent the track point from jumping back,and improve the data quality and monitoring efficiency.At present,the method has been applied in the case that the ground station and transmission link conditions are not ideal.A large number of practical results demonstrate that the method can be used to ensure the output of highly real-time,periodic,stable,and reliable ADS-B data.