基于迭代滤波的上升段不确定参数辨识  

Uncertain parameter estimation of ascent trajectory based on iteration-filtering algorithm

作　　者：赵志刚 李伟杰[1] 沈作军[1] ZHAO Zhi-gang;LI Wei-jie;SHEN Zuo-jun(School of Aeronautics Science and Engineering, BUAA, Beijing 100191, China)

机构地区：[1]北京航空航天大学航空科学与工程学院,北京100191

出　　处：《飞行力学》2018年第3期60-64,共5页Flight Dynamics

摘　　要：针对扩展卡尔曼滤波进行高超声速飞行器在线参数辨识时无法准确辨识参数的不确定性问题,提出了在线实时参数修正的迭代滤波辨识方法。首先,推导了包含侧滑角和推力的三自由度动力学方程;其次,提出基于第一次滤波辨识结果在线实时修正空速并进行第二次滤波的迭代滤波辨识策略;然后,采用一阶马尔科夫过程描述不确定性参数模型,并将不确定性参数增广到系统状态中,建立了基于扩展卡尔曼滤波理论的参数辨识模型。仿真结果表明,这种辨识方法可以有效解决在风场条件下大气密度不确定性无法准确辨识的问题。To solve the problem that the extended Kalman filter can not accurately identify the uncertainty of parameters when identifying the hypersonic vehicle parameters online,an iterative-filtering algorithm based on online real-time parameter correction is proposed. Firstly,the 3 D dynamic equation including the side slip angle and thrust is derived. Secondly,an iterative-filtering identification algorithm based on the first filter identification results is presented to correct the airspeed in real time and make the second filtering. Then,a parameter identification model based on extended Kalman filter is established by using the first-order Markov process to describe the uncertain parameter model and augment the uncertain parameters into the system state. Simulation results show that the identification algorithm can effectively solve the problem that the uncertainty of air density can not be accurately identified under the wind conditions.

关 键 词：扩展卡尔曼滤波 参数辨识 高超声速飞行器 一阶马尔科夫链 

分 类 号：V411.8[航空宇航科学与技术—航空宇航推进理论与工程]