FRFT域能量谱模拟电路故障特征提取方法

故障特征提取是模拟电路故障诊断的关键技术之一,为了提高故障特征的可诊性,提出 1种基于分数阶傅里摘叶变换(Fractional Fourier Transform,FRFT)域能量谱的模拟电路故障特征提取方法。首先,采集测试节点电压信号并将其映射到不同的 FRFT域空间中(p从 0变化到 1);然后,计算所有 FRFT域空间中的能量谱峰值并将其作为故障特征;最后,将归一化后的特征用于训练最近邻分类器进行诊断验证。与现有的 FRFT故障特征提取方法相比,该方法减少了计算量,且提取的特征能够在所有 FRFT域中更全面地反映不同故障响应信号的细微差异,有利于提高故障特征的可分性。在仿真和物理电路上进行了验证,实验结果表明:所提方法能提高故障诊断准确率,且时间复杂度有明显改善。     

单频GPS载波相位相对定位研究

利用 GPS载波相位作为观测量 ,研究了采用最小二乘法和 FARA法相结合进行整周模糊度搜索和相对定位的方法 ,并探讨了影响相对定位精度的主要因素和解决途径 ,进行了相关的实验研究。实验结果表明 :在短基线的条件下 ,采用该方法进行 GPS载波相位测量的静态相对定位精度可优于 1cm。     

整周模糊度动态快速求解

初始整周模糊度的求解是利用GPS载波相位进行测量时的关键问题，本文提出了一种初始整周模糊度在航快速解算方法，该方法由低阶Kalman滤波器和基线约束条件来实现，利用Kalman滤波器动态递推的特性可以实现模糊度的动态解算，由基线约束条件可以有效地剔除不合理模糊度组合，文中对所提出的算法进行了理论推导，并进行了计算机仿真和实测数据分析，仿真和实验结果表明，该算法计算量小，快速准确，适合整周模糊度的快速动态求解。     

基于分数阶模糊函数延时对消的海面运动弱目标检测

研究了分数阶模糊函数的时移特性,根据此特性提出了基于分数阶模糊函数延时对消的海面运动弱目标检测方法.通过对IPIX实测数据验证表明,该方法在增加目标与杂波FRFT峰值差、提高信杂比以及检测概率等方面都明显优于仅对回波作分数阶模糊函数.设置适当的恒虚警检测门限,该检测方法能够达到更好的检测效果.     

冲击噪声环境下基于任意阵列流形的空时二维DOA估计方法

 针对冲击噪声环境中信号源的高分辨二维波达方向(DOA)估计问题,提出了一种适合任意阵形的基于联合对角化分数低阶空时(ST)相关矩阵的二维DOA估计方法。该方法先对分数低阶空时矩进行采样构建伪快拍数据矩阵,然后对两个时延分数低阶空时矩阵进行联合对角化处理。其无需二维谱峰搜索和参数配对,在低信噪比下对强冲击噪声具有更好的抑制作用,适合存在一维角度兼并的情况。仿真结果证明了该方法的有效性。     

Integer phase ambiguity estimation in next-generation geodetic Very Long Baseline Interferometry

Next-generation Very Long Baseline Interferometry (VLBI) system designs are aiming at 1 mm global position accuracy. In order to achieve this, it is not only necessary to deploy improved VLBI systems, but also to develop analysis strategies that take full advantage of the observations taken. Since the new systems are expected to incorporate four independent radio frequency bands, it should be feasible to resolve phase ambiguities directly from post-correlation data, providing roughly an order of magnitude improvement in precision of the delay observable. As the unknown ambiguities are of integer nature, it is discussed here how they the can be resolved analytically using algorithms which have been developed for Global Navigation Satellite System (GNSS) applications. Furthermore, it will be shown that ionosphere contribution and source structure effects, so-called core-shifts, can be solved simultaneously with the delay, which is the main geodetic observable for follow-on analysis. In order to verify the proposed algorithm, simulated observations were created using parameters from actual design studies. It is shown that, even in the case of low signal-to-noise ratio observations, reliable phase ambiguity resolution can be achieved and it is discussed how the integer ambiguity recovery depends on the number of observations and signal-to-noise ratio.     

基于模糊分数阶PI~λD~μ的柔性结构振动主动控制

为防止翅翼结构因干扰而产生振颤,以柔性悬臂梁结构为被控对象,提出一种参数自整定的分数阶比例积分微分控制器,在利用遗传算法取得优化参数后,通过模糊规则第2次对控制器参数进行在线整定,以达到实时最优控制的目的.在Simulink软件环境下进行了数值仿真,结果表明:设计的分数阶比例积分微分控制器具有记忆性,动态表征能力强.与常规比例积分微分控制器相比,冲击激励下的系统响应稳态时间缩短0.9s;简谐激励下的系统响应衰减幅度从88%提升到96%;随机激励下系统响应抖动减小,提高了稳定性.     

Inertial aided BDS triple-frequency integer ambiguity rounding method

The integer ambiguity resolution (AR) of carrier phase is significant for Global Navigation Satellite System (GNSS) precise positioning. However, in kinematic case, single-epoch AR methods based on alone GNSS are usually not reliable due to the instable pseudorange accuracy. Moreover, the computation of classical AR method Least Squares Ambiguity Decorrelation Adjustment (LAMBDA) is large. Thus, the inertial measurement unit (IMU) is introduced, a new inertial-aided AR method that directly rounds the float ambiguity of BeiDou triple-frequency combined observations, which is characterized by long wavelength, low carrier-phase noise and ionospheric delay, is proposed. The mathematical model of the new method is derived first. Then the impacts of the carrier-phase noise, ionospheric delay and inertial navigation system (INS) position error on the AR success ratio of combined observation are analyzed through probabilistic approach. Based on above investigation, the combinations (0, ?1, 1), (1, 4, ?5) and (4, ?2, ?3) are selected to resolve the original ambiguity. A vehicular integrated navigation test is performed to demonstrate the proposed method. The results show that the average AR success ratios of the three selected combinations, whose float ambiguity errors are 0.041, 0.146, 0.279 cycle respectively, are above 97.25% without regard to low-elevation C05. With respect to positioning accuracy based on our AR method when compared with IE software, the east, north, up error RMS of position are 0.042, 0.024, 0.069 m, respectively. In terms of the AR recover after the BeiDou signals outage, as long as 62 s BeiDou signal complete outage, all the ambiguities of all satellites could be re-fixed immediately. Besides, during the 90 s signals partial outage, the AR is not influenced by the position error, since the float ambiguity errors are all below half-cycle. The research of this contribution demonstrates the effectiveness of the proposed new method, which indicates it is applicable to kinematic positioning, even in BDS degraded and denied environments.     

基于虚拟基线变换和RBFNN的宽频段DOA估计

提出利用虚拟基线变换的方法解决了频率高端的相位模糊,克服由于相位模糊性造成的方向特征的多值模糊问题,从而简化了从方向特征到波达方向的映射关系。提取无模糊的方向特征,训练RBF神经,进行宽频段波达方向估计。理论分析和试验表明,该解模糊方法计算量小,能有效消除圆阵列的相位模糊,适用阵元数大于等于5的任意均匀圆阵列天线（6元除外）;利用解模糊后的特征建立的DOA估计模型结构简单,性能大大优于直接建模和相关干涉仪法,具有优越的性能和工程实用性。