一种高精度气动光学星光偏折预测方法  

A High Precision Aero-Optical Starlight Deviation Prediction Method

作　　者：刘宇鑫 王新龙[1] 丁伟 胡晓东[3] Liu Yuxin;Wang Xinlong;Ding Wei;Hu Xiaodong(School of Astronautics,Beihang University,Beijing 100083,China;Beijing Institute of Control&Electronic Technology,Beijing 100038,China;AVIC Xi’an Flight Automatic Control Research Institute,Xi’an 710065,China)

机构地区：[1]北京航空航天大学宇航学院,北京100083 [2]北京控制与电子技术研究所,北京100038 [3]航空工业西安飞行自动控制研究所,西安710065

出　　处：《航空兵器》2023年第6期109-116,共8页Aero Weaponry

基　　金：国家自然科学基金(61673040);重点基础研究项目(2020-JCJQ-ZD-136-12);航空科学基金项目(20170151002);天地一体化信息技术国家重点实验室基金项目(2015-SGIIT-KFJJ-DH-01)。

摘　　要：基于星光折射的新型天文导航方法具有自主性强、导航精度高的特点,在航空航天领域受到了广泛关注。然而,当星光穿过高超声速飞行器外的复杂流场时会发生气动光学效应,导致星光传播方向发生偏折,引起星光矢量测量误差进而影响天文导航的精度。传统基于拟合模型的星光偏折预测法利用数据拟合星光偏折角模型用于预测补偿,这种方法忽略了星光偏折的物理过程,对用于拟合的数据精度以及数据量要求高,能取得的精度有限。针对这一问题,本文提出了一种基于解析计算结果的高精度气动光学星光偏折预测方法。通过分析气体通过锥面激波后的流动规律,基于PSO-BP网络构建了流场的密度代理模型。通过分析星光在流场中发生偏折的机理,综合考虑了激波以及激波后流场对星光传播的影响,建立了一种精确的星光偏折角模型。进一步结合密度代理模型与星光偏折角模型设计了气动光学星光偏折预测方法,以实现对星光偏折的高精度预测与补偿。仿真结果表明,相比于传统的数据拟合模型预测法,所提方法能对星光偏折角实现亚角秒级别的高精度预测,能补偿80%以上的气动光学效应所导致的星光偏折,并且能大幅简化预测模型的构建过程,具有较高的适用性和工程应用价值。The novel celestial navigation method based on starlight refraction has attracted widespread attention in the aerospace field due to its strong autonomy and high navigation accuracy.However,aero-optical effect occurs when starlight passes through the complex flow field outside hypersonic aircraft,resulting in deflection of the starlight propagation direction.This deflection causes measurement errors in starlight vectors and subsequently affects the accuracy of celestial navigation.The traditional starlight deviation prediction method based on fitting models uses data to fit the starlight deviation model for prediction.This method ignores the physical process of starlight deviation,and thus requires high accuracy and data volume for fitting,resulting in limited prediction accuracy.Aiming at this problem,a high precision aero-optical starlight deviation prediction method based on analytical computation results is proposed.By analyzing the flow characteristics of gas passing through the conical shock wave,a density proxy model of the flow field is constructed based on the PSO-BP network.By analyzing the mechanism of starlight deflection in the flow field and considering the influences of shock wave and the flow field after shock wave on starlight propagation,an accurate model of starlight deflection angle is established.Furthermore,in order to achieve high precision prediction and compensation of starlight deflection,an aero-optical starlight deflection prediction method based on density proxy model and starlight deflection model is proposed.Simulation results demonstrate that compared to the traditional data fitting model prediction method,the proposed method can achieve higher accuracy of starlight deflection angle at the sub angular second level,compensate for over 80%of starlight deviation caused by aero-optical effects,and significantly simplifies the construction process of the prediction model,thereby exhibiting high applicability and engineering value.

关 键 词：星敏感器 气动光学效应 星光偏折 密度代理模型 神经网络 高超声速飞行器 

分 类 号：TJ760[兵器科学与技术—武器系统与运用工程]