设计并制作了一种基于多模-细芯-多模光纤（MMF-TCF-MMF）结构的双参数测量传感器，在此结构中同时产生马赫-曾德尔干涉（MZI）与表面等离子体共振（SPR），来实现对折射率和温度的同时测量。该传感器中的两段多模光纤（MMF）作为输入输出耦合器，细芯光纤（TCF）镀上一层金属银膜作为传感臂。实验结果表明，MZI和SPR在1.333~1.412 RIU范围内折射率灵敏度分别为-70.373 nm/RIU和4888.77 nm/RIU，在20~80 ℃范围内温度灵敏度分别为73.48 pm/℃和-0.3215 nm/℃。采用灵敏度矩阵实现了双参数测量，可解决折射率传感器温度的交叉敏感。该传感器成本低、灵敏度高、制作简单，在生物化学医疗检测领域有良好的应用前景。

提出并制备了一种温度自补偿的光纤表面等离子体共振（SPR）折射率传感器，用于实现温度和折射率（RI）的同时测量。该传感器利用细芯光纤（TCF）与多模光纤（MMF）构成MMF-TCF-MMF双通道级联结构，通道1涂覆银膜实现折射率传感，通道2涂覆复合膜（Ag-ITO）以及聚二甲基硅氧烷（PDMS）膜来实现温度传感，进而达到温度补偿的目的。当折射率和温度变化时，两个共振波长将发生变化，通过监测两个共振波谷的波长偏移可以同时测量折射率和温度。实验结果表明，两个通道在1.333~1.357 RIU（RIU为折射率单位）范围内的折射率灵敏度分别为3141.85 nm/RIU和0 nm/RIU，40~80 ℃范围内的温度灵敏度分别为-0.07 nm/℃和-1.74 nm/℃。该传感器具有体积小、灵敏度高、易制作等优点，在生物医学、环境监测等领域具有一定的应用价值。

地形匹配导航是一种重要的利用地形和地物特征进行导航的技术。为了进一步提高地形匹配算法的精度，提出了一种基于形态学增强型方向梯度直方图（EHOG）的地形匹配算法。该算法采用形态学闭运算对干涉合成孔径雷达（InSAR）获取的实时高程图（REM）进行预处理，获取EHOG特征；将地形匹配转换为HOG特征描述子的匹配，以特征向量间的欧氏距离作为相似性度量。匹配过程中，采用粗细精结合的三步优化匹配搜索策略，提升了算法实时性。实验结果表明，与未增强的HOG算法、传统的梯度互相关算法相比，所提匹配算法具有更好的匹配精度和抗噪性，同时具有较强的鲁棒性和实用性，能够很好地应用于InSAR地形匹配。

To overcome the invalid phase and phase jump phenomenon generated during the phase unwrapping, a phase error correction method based on the Gaussian filtering algorithm and intensity variance is proposed in this paper. First, a threshold of fringe intensity variance is set to identify and clear the phase in the invalid region. Then, the Gaussian filtering algorithm is employed to correct the phase order at the fringe junction, and then the absolute phase is corrected. Finally, the phase correction experiments of different geometric objects are carried out to verify the feasibility and accuracy of the proposed method. The method proposed in this paper can be extended to the correction of absolute phase error obtained by any coding method.

针对铝合金结构服役过程中的疲劳裂纹监测问题,提出了一种基于光纤光栅传感阵列的疲劳裂纹实时监测方法。首先,结合扩展有限元技术获得裂纹扩展过程的应变场分布特征,进行传感器布局设计;其次,通过疲劳加载实验获取加载次数、裂纹长度对波长峰峰值的响应特征,采用三参数指数法进行疲劳扩展a-N(其中N为加载次数,a为裂纹长度)曲线拟合;在实验数据基础上,采用梯度提升回归树算法搭建疲劳损伤预测模型,并将其与线性回归、支持向量两种方法进行对比,选取可释方差、平均绝对误差、平均方差、决定系数和五折交叉验证等多种指标进行模型评估。结果表明,梯度提升回归树算法对疲劳裂纹长度的预测效果最佳。

为了有效剔除无效目标，降低雷达虚警概率，提出一种基于雷达环境噪声统计和目标功率分布特性的动态阈值计算方法。运用直方图统计方法估计每个距离门的噪声水平并对其进行峰值搜索找到局部峰值点，同时根据不同距离门雷达信号衰减的强弱，动态分配阈值增益，得到自适应峰值检测阈值。然后将局部峰值点作为参考单元的信号，并将其输入到检测方法中求解检测门限。实验结果表明，所提方法能够有效识别雷达在实际复杂应用场景下的目标。

To achieve the real-time detecting and localization of hypervelocity impact events, a monitoring and localization system was designed based on fiber Bragg grating (FBG) sensor network. First, the simulation model was built to study the damage evolution and wave propagation process. Subsequently, based on the response mechanism of FBG to strain, the corresponding high frequency demodulation system was designed. Furthermore, a hypervelocity impact experiment was performed to verify the effectiveness of the designed system. Finally, combined with the diamond sensor array localization algorithm, the accurate position of hypervelocity impact source can be achieved.

In this paper, to reduce the non-linear phase error caused by gamma effect of the digital projector, a novel phase difference extraction method is proposed. First, through multi-layer discrete wavelet decomposition (DWT), the components of the original phase difference at different frequencies are extracted. Subsequently, the effective components are taken by the orthogonal matching pursuit (OMP) algorithm, and the phase difference is reconstructed. According to the upper and lower envelopes of the original phase difference extracted by cubic spline interpolation, the residual threshold of OMP is set. According to the results of three-dimensional (3-D) reconstruction experiment on a standard sphere, the proposed method is capable of reducing the errors in the original phase difference and achieving a better imaging result.