基于自适应IMM算法的蛇形机动目标加速度估计研究

针对防空导弹拦截高速蛇形机动反舰导弹的难题,基于导引头测量的视线角速度、弹目相对速度等信息,开展了蛇形机动目标加速度估计算法的研究,最终建立了基于交互多模型的扩展卡尔曼滤波算法,实现了对蛇形机动目标的加速度估计。考虑传统加速度方差计算方法受限于先验值的缺陷,根据残差对加速度方差的计算方法进行了改进。数字仿真证明:提出的自适应交互多模型算法能有效提高蛇形机动目标加速度的估计精度。     

双重未知干扰解耦的多传感器系统偏差校正与状态估计

具有未知干扰输入的随机系统状态估计问题广泛存在于控制、通信、信号处理和故障诊断等领域,但目前的研究成果大多局限于单传感器动态离散系统。针对状态方程含有未知干扰、量测方程含有未知偏差的多传感器系统状态估计问题开展了研究,提出了一种双重未知干扰解耦下的最小方差无偏估计滤波器。首先,建立量测偏差通用演化模型;然后,解耦偏差演化模型中的未知输入,设计最小方差无偏估计器对量测干扰偏差进行估计;之后,利用估计出的量测偏差来校正动态系统测量值;最后,根据量测偏差校正后的系统模型设计最优状态观测器,获得具有最小方差无偏的状态估计。径向飞行控制系统的例子验证了所提方法的有效性,与相关方法的仿真对比反应了所提方法的优越性。     

不同温度下丁羟包覆层的横向弛豫时间与拉伸性能的相关性

为了研究不同温度下丁羟包覆层的横向弛豫时间与拉伸性能的相关性,开展了核磁共振和拉伸应力-应变性能测定试验。单因素方差分析表明温度对横向弛豫时间有显著影响;试验温度从30℃升到90℃,横向弛豫时间呈线性增大;90℃升到130℃,横向弛豫时间先减小后增大。30～90℃,升温使包覆层的拉伸强度下降,断裂伸长率升高;在100℃较90℃强度得到了提高,断裂伸长率稍有降低;在100～130℃时,受复杂化学反应和分子热运动共同影响,断裂伸长率迅速增加,强度降低。断裂伸长率、拉伸强度均与横向弛豫时间存在较好的相关性,利用该关系可以预测丁羟包覆层在不同横向弛豫时间下的拉伸性能。     

利用小波方差进行原子钟频率稳定度的估计

离散小波变换可以在不同尺度上分解时间序列,而不同尺度的波动性可用小波方差来表征。从小波方差的定义入手,系统地归纳了基于极大重叠离散小波变换(MODWT)的小波方差估计方法,及其等效自由度(EDF)的实用计算方法。最后利用一个实测算例进行计算分析,并与相应的重叠阿伦方差、重叠哈达玛方差进行比较,通过实验分析可以看出小波方差可有效消除原子钟信号非线性和非平稳性的影响,通过选择适当的小波基函数,如D4、D6小波,其方差可以像哈达玛方差一样,减少调频闪变噪声和调频随机游走噪声的泄露,适用于原子钟频率稳定度的表征。     

An optimized strategy for inter-satellite links assignments in GNSS

Inter-Satellite Links (ISLs) can effectively enhance the Global Navigation Satellite System (GNSS) performance, such as communication and ranging performance. These performances would have been improved by disposing sparsely of ground stations and ISLs. Here, we demonstrate a topology scheduling strategy based on Signed Variance and Blossom algorithm (SV-Blossom) to promote the topology performance of GNSS. We also analyze the effect of this strategy on the robustness of the ranging information quality of the ISLs assignments system. For the domestic and overseas satellite communication Time Delay (TD) and Position Dilution of Precision (PDOP), by analyzing the relation of satellites, a set of the ISLs cost models is established, and the cost matrix is obtained. The final ISLs assignments result is obtained by processing the link cost matrix with the Blossom algorithm in the graph theory. In order to verify the rationality of the SV-Blossom strategy, a typical BDS-3 simulation scene with 10080 min is built to test these performances of ISLs. The final results show that the ranging information quality of the SV-Blossom strategy is greatly improved compared with recently published strategies, and the proposed links assignments mechanism has good robustness.     

基于自适应KF动态虚拟陀螺数据融合算法的研究

微机电系统(MEMS)陀螺与光纤陀螺相比，传感器的精度较低。为了提高MEMS陀螺的精度，通过组合多个相同陀螺实现虚拟陀螺的功能，同时提高虚拟陀螺的静态和动态性能。通过分析陀螺的Allan方差，并考虑陀螺之间的相关性，建立陀螺的测量模型;使用自回归(AR)模型建立预测模型，对卡尔曼滤波(KF)算法进行优化;搭建多MEMS陀螺仪硬件平台，获取数据并实时计算，融合多陀螺数据输出最优估计值，使用高精度转台分别在静态和动态条件下测试滤波效果。实验结果表明:静态条件下虚拟陀螺误差的方差可降低为单个陀螺的1/94，动态条件下降低为单个陀螺的1/18。基于自适应KF的虚拟陀螺可以显著提高精度。     

基于响应曲面法的成型CBN砂轮高频感应钎焊温度均匀性研究

成型砂轮在高频感应钎焊过程中存在温度分布不均匀导致钎焊质量无法满足使用要求，这是长期困扰超硬磨料砂轮高频感应钎焊技术的难题。针对该问题，提出温度均匀性的表征方法，以温度均匀度和平均温度为响应值，基于有限元仿真数据获得了响应曲面法模型。基于该模型，采用方差分析成型砂轮感应钎焊温度均匀性的影响因素显著程度由大到小依次为加热间隙、感应电流、导磁体长度。以平均温度值和温度均匀度为优化目标，基于响应曲面法优化线圈结构和工艺参数，开展了感应加热试验，证明响应曲面法模型优化结果的误差在6.94%以内。高频感应钎焊的成型立方氮化硼（cubic boron nitride，CBN）砂轮宏观形貌显示，钎料在成型面各处铺展的一致性好，表明钎焊过程中在成型面的温度分布具有较好的均匀性。     

基于三维原子磁强计的核磁共振陀螺仪实验优化

核磁共振陀螺仪内嵌三维原子磁强计是实现核磁共振陀螺仪小型化的一种有效途径,故介绍了一种基于三维原子磁强计的核磁共振陀螺仪.对影响三维原子磁强计性能的重要参数进行了优化,得到了较优的x轴、yy轴和z轴磁强计信号标度因子,进而实现了更灵敏的三维原子磁强计.在三维磁场闭环锁定6000 s后,测得核磁共振陀螺仪的角度随机游走和零偏稳定性分别为0.038(°)/h1/2和0.94(°)/h.     

Tight integration of BDS-3/BDS-2/GPS/Galileo observations considering the new overlapping DISBs and its application in obstructed environments

Tight integration can enhance the model strength and positioning performance by considering the characteristic of differential inter-system bias (DISB), especially in obstructed environments. However, limited work emphasizes the comprehensive analysis of five-frequency DISBs between BDS-3 and other systems considering the receiver type, receiver configuration, and antenna type. In addition, the overlapping DISBs between BDS-3 and BDS-2 are also in great demand for further investigation since they are often regarded as one system. In this study, one DISB-float model is introduced to estimate the DISBs, and one DISB-fixed model and one DISB-free model are formulated to enhance the model strength of tight integration. Four dedicated datasets were collected to estimate the DISBs, which are also comprehensively analyzed considering the receiver type, receiver configuration, and antenna type. The results show that the DISBs between BDS-3 and other systems are rather stable over a certain period and are related to the receiver type and receiver configuration, whereas are not related to the antenna type. More interestingly, the B1I code DISB between BDS-3 and BDS-2 exhibits significant magnitude with a mean value of ?1.44 m for the baseline composed of two different receivers. In this case, the B1I code DISB must be considered and the tight integration between BDS-3 and BDS-2 considering its calibration can improve the positioning performance. Besides, the tight integration of the DISB-fixed model can significantly improve the positioning accuracy between multiple GNSS. Compared to the loose integration, the improvement of 60.6 %, 56.6 %, and 61.2 % can be obtained in the E, N, and U directions, when only two satellites are available for each system. In real obstructed environments, the tight integration of the DISB-free model can also improve the positioning performance in terms of positioning availability and accuracy, as well as the ambiguity resolution performance.     

Cramér-rao lower bound for pulsar rotation parameters estimation with X-ray pulsar observation data

In this paper, the Cramér-Rao Lower Bound (CRLB) for estimating the rotation parameters of pulsars using X-ray pulsar observation data is deduced, and the calculation equation is presented. In order to verify the correctness of the deduced equation, we use the X-ray pulsar observation data to estimate pulsar rotation parameters, and obtain the root mean square error (RMSE) of the estimated pulsar rotation parameters through conducting repeated experiments. The experimental results suggest that when the observation time increases, the RMSE gradually approaches the estimated CRLB, and that when the observation time is 2.4 × 10⁶ s, the error between the RMSE of pulsar frequency estimation and the CRLB stays at 10^?11 order of magnitude. This verifies that the CRLB expression deduced in this paper is the theoretical lower bound for estimating pulsar rotation parameters. The deduced CRLB in this paper helps determine the minimum variance estimator for pulsar rotation parameter estimation using X-ray pulsar data, providing a benchmark for the comparison between the minimum variance estimator and other estimators.