基于差异化估计模型的航电系统地面验证飞参数据预处理方法

The pretreatment method of flight data on differentiating estimation models for avionics ground validation

飞机航电系统地面验证,是航电系统设计完成后和交付装机前的重要检验环节,需要通过数据激励航电设备模拟飞行过程中多种使用场景,从而验证设备级、系统级、全机级的功能响应正确与否。通过飞参数据挖掘和深度分析,实现飞参数据在航电系统地面验证中的利用,是改进飞机性能提高飞机设计水平,探索航电系统验证专业的目标和方向。文章结合飞参数据在航电系统近地告警子系统地面验证中的实践,介绍基于差异化估计模型的飞参数据预处理方法,包括特征数据筛选、冗余数据清理、缺失数据插补等系列流程,经过数据分析、数据预处理、模型校验、整合输出,获得可用、完整、简洁的高质量验证激励数据,有效提升航电系统验证的质量和效率。

The aircraft avionics system ground validation is the important step for aircraft designing and intergration.In order to validate the performance of devices and systems,it uses incentive data to simulate the flight scenes.In recent years,flight data mining and analysis are gradually becoming effective methods for improving aircraft designing levels.This article discusses the pretreatment method of flight data based on differentiating estimation models,for the Ground Proximity Warning System(GPWS)validation.It includes characteristic data selection,redundance dataclear and missing data interpolation.Through this method and above steps,we obtained a series of available activating data,so we can validate the GPWS effectively.