受限空间欠采样条件下航空发动机关键截面流场重构方法研究进展

由于受整机环境下几何特征对探针位置的限制以及3维流动等因素影响，通过有限的测点数据准确获取整机关键测试 截面参数非常困难。为解决这一难题，综合关键截面流场重构项目的研究成果，针对航空发动机内强3维、强不均匀流场，介绍了 一种受限空间欠采样条件下航空发动机复杂流场重构技术。基于航空发动机内流场近周期分布的特征，利用离散的探针数据通 过“多波束近似”的方法来精确重构发动机内部的周向流场。相关技术在重构多级压气机出口截面总压分布、燃烧室出口热斑分 布、高压涡轮出口截面总温分布的测试中进行了初步试验验证，结果表明：重构结果与真实试验结果的误差分别小于0.1%、0.5%、 0.3%。与传统均布或等节距探针布局方案及均值数据处理方法相比，采用新的探针布局方案及流场重构方法可以实现航空发动 机压气机出口总压场、燃烧室及高压涡轮出口总温场的精确重构。     

一种新的宽带LFM信号波束形成方法研究

为了提高电子对抗无源侦察系统的测向、定位以及成像性能,针对宽带LFM信号提出一种基于延迟相乘去斜的波束形成方法。该方法巧妙地利用LFM信号频率调制样式的特点,通过时域延迟相乘预处理,将LFM信号的阵列输出数据转化为具有窄带阵列结构的相位延迟形式,最后利用窄带波束形成方法形成波束。新方法只适用于LFM信号,针对性很强,极大地减小了系统设计的复杂度。由计算机仿真实验分析可知,该方法较传统的方法有较好的稳健性,突破传统阵列间距的限制,提高了系统性能,在高分辨测向与成像领域中具有很好的应用前景。     

插值逼近结合FDTD法快速计算目标宽角度RCS

FDTD(Finite Difference Time Domain)法结合多项式插值逼近和样条函数插值逼近快速计算了三维目标的宽角度RCS(Radar Cross Section).引入插值逼近方法可以节省计算时间.在整个入射角度范围内选定若干个入射角,对不同的入射角,分别用FDTD法计算得到外推面上各点的切向电磁场值,进而得到这些场值随入射角度变化的插值函数,然后用插值函数计算出全入射角度范围内外推面上各点的切向电磁场值,最后通过近远场变换得到宽角度RCS.计算结果表明,在只有少数几个插值节点的情况下该方法就能很好地逼近FDTD法的精确计算结果,节省了计算时间.      

压力面气膜冷却数值模拟

针对典型高压涡轮叶型平面叶栅压力面气膜冷却,采用数值模拟方法,比较分析了不同冷气进口方式对计算结果的影响,并对冷气入射角度等参数对叶栅流场和性能的影响规律进行了研究。结果表明,在没有考虑二次流流动影响情况下,平面叶栅中通过管道给定冷气进口和直接在叶片表面给定冷气进口这两种方式对气膜冷却数值模拟结果影响很小,能量损失系数的差别仅为1%左右。冷气入射增加了叶栅损失,但能量损失系数与冷气入射角度并不是简单的单调关系,在入射角度从15°到60°的变化范围内,能量损失系数存在最小值,对应冷气入射角度在30°左右。     

Second post-Newtonian approximation of scalar-tensor theory of gravity

Deep space laser ranging missions like ASTROD I (Single-Spacecraft Astrodynamical Space Test of Relativity using Optical Devices) and ASTROD, together with astrometry missions like GAIA and LATOR will be able to test relativistic gravity to an unprecedented level of accuracy. More precisely, these missions will enable us to test relativistic gravity to 10^-7–10^-9${10}^{-7}''–''{10}^{-9}$ of the size of relativistic (post-Newtonian) effects, and will require second post-Newtonian approximation of relevant theories of gravity. The first post-Newtonian approximation is valid to 10^-6${10}^{-6}$ and the second post-Newtonian approximation is valid to 10^-12${10}^{-12}$ in terms of post-Newtonian effects in the solar system. The scalar-tensor theory is widely discussed and used in tests of relativistic gravity, especially after the interests in inflation models and in dark energy models. In the Lagrangian, intermediate-range gravity term has a similar form as cosmological term. Here we present the full second post-Newtonian approximation of the scalar-tensor theory including viable examples of intermediate-range gravity. We use Chandrasekhar’s approach to derive the metric coefficients and the equation of the hydrodynamics governing a perfect fluid in the second post-Newtonian approximation in scalar-tensor theory; all terms inclusive of O(c^-4)$\mathrm{O}(c^{-4})$ are retained consistently in the equations of motion.     

复合正交级数神经网络模型在飞控中的应用

针对现有神经网络逼近快变非线性函数能力低、结构复杂等问题,提出了复合正交多项式模型和一种改进的正匀多项式神经网络模型。这两种模型综合了正交级数和神经网络的结构特点,便于工程实现。理论研究和仿真计算表明,使用此模型构造快变非线性函数时,可使神经网络的节点数成倍减少,神经网络的训练次数成倍降低。飞行器非线性控制的应用表现,此方法可用来解决一般非线性系统的控制问题。     

滚珠丝杠与直线滚动导轨在测量中的应用

介绍了以滚珠丝杠与滚动导轨副为主体的机械伺服系统在大型轮式零件尺寸综合测量中的应用 ,并分析了传动系统刚度对测量准确度的影响     

高频角振动台变负载的模型跟随自适应控制系统设计

以高频角振动台为研究对象,考虑负载变化对系统响应的影响,根据高频角振动系统确定参考模型,给出了线性模型跟随控制(LMFC)和自适应模型跟随控制(AMFC)系统的设计.仿真结果表明:在负载变化时AMFC的控制效果明显优于LMFC.     

Influence of sensor and actuator errors on impulsive satellite formation control methods

This paper investigates how sensor and actuator errors are impacting formation control accuracy and propellant consumption of a two-satellite formation in a low Earth orbit. Realistic relative navigation errors are implemented, based on the results from the PRISMA mission, as well as realistic actuator uncertainty and actuator constraints. Two impulsive control methods are investigated. The first method is based on a controller that is implemented onboard PRISMA and the second method uses linear programming to arrive at a model predictive controller. The control methods are tested in a simulation environment and are subjected to orbital perturbations and realistic sensor errors and actuator errors. Both control methods are able to maintain the desired relative geometry of a projected circular orbit in the presence of the errors. The PRISMA control method demonstrates lower propellant consumption, while the model predictive controller shows better control accuracy. The results show that, based on the used scenario, sensor errors dominate both the formation control accuracy and propellant consumption. The versatility of the model predictive controller is demonstrated in a challenging formation control scenario including formation maintenance and formation reconfiguration tasks.