混合变量系统基于MCMC的自适应重要抽样法

 系统关键故障的发生,会导致系统处于各种离散性能降级状态.针对传统的基于马尔可夫链蒙特卡罗(Markov chain Monte Carlo,MCMC)的自适应重要抽样法只适用于连续变量系统的不足,提出考虑混合变量的基于MCMC的自适应重要抽样法,以支持系统性能可靠性的高效仿真.该方法首先将由关键故障产生的不同失效域组成失效空间,并通过初始样本点在失效空间中随机游走构造马尔可夫链模拟样本;然后综合考虑连续变量与离散变量,利用核密度估计构建混合核抽样密度函数;再根据该密度函数进行重要抽样仿真并计算系统的性能可靠性;最后对该方法的仿真效率进行理论分析.通过电液舵机(Electro-Hydrostatic Actuator,EHA)案例对方法的正确性和仿真效率进行验证.     

无失效数据下液体火箭发动机的E-Bayes可靠性分析

为了解决航空航天产品、造价昂贵的产品及高可靠性产品的可靠性评估,本文在无失效数据情况下,对液体火箭发动机进行了可靠性分析.应用E-Bayes(Expected-Bayesian)估计法,在有关研究的基础上,针对液体火箭发动机热试车过程中可能出现的无失效数据情况,确定了失效概率的先验分布,并给出了E-Bayes估计和多层...     

高超声速飞行器气动力/热参数辨识研究综述

飞行器气动力和气动热参数辨识是高超声速飞行器设计的关键技术之一。笔者对高超声速飞行器的气动力和气动热参数辨识技术进行了综述。介绍了飞行器气动力、气动热参数辨识的基本原理与主要方法,气动力、气动热参数辨识技术在高超声速飞行器研发中的应用情况与发展趋势。同时也简要介绍了中国空气动力研究与发展中心（CARDC）在气动力和气动热参数辨识研究方面的研究概况。     

飞机刹车系统方案设计及仿真分析

构建飞机刹车系统方案设计环境,以提高飞机刹车系统方案设计阶段的设计效率和质量。研究刹车装置参数和防滑控制系统参数的快速确定方法,通过迭代的设计思想使结果符合设计要求;估算刹车装置性能和刹车系统的动态特性。在MATLAB／Simulink中建立刹车系统模型,通过MATLAB引擎调用的方式进行刹车系统的动力学分析;以某无人机刹车系统为设计实例进行其刹车系统的方案设计,动力学仿真中刹车系统工作良好,与实际工作状况基本吻合。结果表明：本文提供的飞机刹车方案设计方法快速有效。     

飞行器气动参数估计不确定度评价方法研究

为了有效地利用飞行器气动参数估计结果,必须同时给出估计结果的不确定度,因此研究了不确定度评价方法.当试验样本较多时,样本标准差具有明确的统计含义,比分散度能够更好地描述参数估计的不确定度.对于单个试验样本,基于不确定度椭球导出的C-R(Cramer-Rao)界是参数估计不确定度最好的理论预测,但利用飞行实测数据得到的C-R界普遍比样本标准差小.鉴于此,通过在低频有色噪声的基础上构造白噪声的方法,得到了一种C-R界修正方法,并通过仿真算例验证了修正方法的正确性.最后,将C-R界修正方法应用于飞行实测数据,得到的修正C-R界与利用多次飞行试验参数估计结果计算的样本标准差比较一致,表明该修正方法能够较好地给出参数估计的不确定度区间.     

Satellite lithium-ion battery remaining useful life estimation with an iterative updated RVM fused with the KF algorithm

Lithium-ion batteries have become the third-generation space batteries and are widely utilized in a series of spacecraft. Remaining Useful Life (RUL) estimation is essential to a spacecraft as the battery is a critical part and determines the lifetime and reliability. The Relevance Vector Machine (RVM) is a data-driven algorithm used to estimate a battery’s RUL due to its sparse feature and uncertainty management capability. Especially, some of the regressive cases indicate that the RVM can obtain a better short-term prediction performance rather than long-term prediction. As a nonlinear kernel learning algorithm, the coefficient matrix and relevance vectors are fixed once the RVM training is conducted. Moreover, the RVM can be simply influenced by the noise with the training data. Thus, this work proposes an iterative updated approach to improve the long-term prediction performance for a battery’s RUL prediction. Firstly, when a new estimator is output by the RVM, the Kalman filter is applied to optimize this estimator with a physical degradation model. Then, this optimized estimator is added into the training set as an on-line sample, the RVM model is re-trained, and the coefficient matrix and relevance vectors can be dynamically adjusted to make next iterative prediction. Experimental results with a commercial battery test data set and a satellite battery data set both indicate that the proposed method can achieve a better performance for RUL estimation.     

A fully automated framework for helicopter rotor blades design and analysis including aerodynamics,structure, and manufacturing

This study describes an integrated framework in which basic aerospace engineering aspects (performance, aerodynamics, and structure) and practical aspects (configuration visualiza-tion and manufacturing) are coupled and considered in one fully automated design optimization of rotor blades. A number of codes are developed to robustly perform estimation of helicopter config-uration from sizing, performance analysis, trim analysis, to rotor blades configuration representa-tion. These codes are then integrated with a two-dimensional airfoil analysis tool to fully design rotor blades configuration including rotor planform and airfoil shape for optimal aerodynamics in both hover and forward flights. A modular structure design methodology is developed for real-istic composite rotor blades with a sophisticated cross-sectional geometry. A D-spar cross-sectional structure is chosen as a baseline. The framework is able to analyze all realistic inner configurations including thicknesses of D-spar, skin, web, number and ply angles of layers of each composite part, and materials. A number of codes and commercial software (ANSYS, Gridgen, VABS, PreVABS, etc.) are implemented to automate the structural analysis from aerodynamic data processing to sec-tional properties and stress analysis. An integrated model for manufacturing cost estimation of composite rotor blades developed at the Aerodynamic Analysis and Design Laboratory (AADL), Aerospace Information Engineering Department, Konkuk University is integrated into the framework to provide a rapid and dynamic feedback to configuration design. The integration of three modules has constructed a framework where the size of a helicopter, aerodynamic performance analysis, structure analysis, and manufacturing cost estimation could be quickly investigated. All aspects of a rotor blade including planform, airfoil shape, and inner structure are considered in a multidisciplinary design optimization without an exception of critical configuration.     

基于改进Cycle-GAN的光流无监督估计方法CSCD

卷积神经网络为光流的计算提供了一种新的方式,但作为一种数据驱动技术,用于训练网络的大规模光流真值在现实世界中不易获取。为了解决这个弊端,基于Cycle-GAN的循环对抗机制,提出了一种光流无监督估计方法。首先,引入双判别器机制在生成器生成的光流样本的底层和高层特征上进行鉴别,迫使生成器提高光流生成的精度。其次,引入Spynet作为教师网络,在生成器训练前期对其进行指导,防止网络陷入模式崩塌。最后,改进损失函数,提出了光流一致性损失和轮廓一致性损失函数,进一步提升光流估计精度。实验结果表明,与现有的先进算法相比,提出的方法达到与有监督算法相同的精度水平。     

一种基于模态参数实时辨识方法的参数时变航天器控制方法

针对一种因挠性结构转动引起模态参数变化的航天器研究了一种基于模态参数辨识的控制方法。首先以一种刚柔耦合复杂航天器为对象,建立分析航天器的动力学模型。然后,基于该模型,采用一种基于改进递归预测器的子空间辨识(RPBSID)法模态辨识方法估计系统状态量。最后,基于辨识状态量采用无模型控制方法进行控制,通过基于MATLAB软件的仿真实验,验证了方法的有效性。     

天线组阵中相位差的递推最小二乘估计与滤波

相位差的估计精度是影响深空网天线组阵中信号合成性能的关键因素。利用天线组阵S UMPLE或SIMPLE相关算法中相位差估计值之间的几何关系,提出了相位差变化率的最小二乘 估计,并在此基础上提出了相位差的递推最小二乘滤波方法。经过理论分析和仿真实验表明 ,所提出的方法能够大大提高相位差的估计精度;对于给定的例子,估计精度提高了接近两 个数量级。相位差经滤波后,天线阵的信号合成性能显著增强,从而大大提高对深空通信中 微弱信号的数据接收能力。