高精度数字全息显微衍射层析成像

采用样品转动结构的光学衍射层析技术存在"苹果核频谱缺失"的问题,为此,本文提出傅里叶衍射映射正约束迭代算法以恢复缺失频谱,实现高精度折射率层析成像.首先基于光学衍射层析成像原理,采用带有样品转动装置的数字全息显微成像系统记录下360°视角内各个角度的显微全息图,计算得到每个角度的复振幅像分布;然后运用傅里叶衍射映射和正约束迭代相结合的衍射层析成像算法,计算得到样品内部折射率的三维分布.实验结果表明,该方法能有效地恢复缺失频谱,高精度地测量微小样品内部的折射率三维分布.

基于改进慢特征分析的CSTR故障诊断方法与实验平台

针对传统慢特征分析(SFA)无法充分解析连续搅拌釜式反应器(CSTR)非线性特性问题,提出一种改进的慢特征分析故障诊断方法——随机傅里叶SFA(RFSFA),并开发了相应的仿真实验平台。该方法引入随机傅里叶映射技术实现过程变量的非线性变换,进而利用慢特征分析建立非线性统计监控模型。为了避免模型随机参数的影响,应用贝叶斯推理理论构建了集成学习模型。为验证该方法的有效性,设计了一个CSTR故障模拟与算法测试实验平台,包括正常工况模拟、故障工况模拟、故障检测等多个子系统。测试结果表明,RFSFA方法具有比传统SFA方法更好的故障检测性能,所开发的实验平台易于操作,开放性好,能够很好地验证算法的有效性。

Permutation-substitution image encryption scheme based on a modified chaotic map in transform domain

A new chaotic image encryption scheme based on permutation and substitution in the Fourier domain is presented.Fractional Fourier Transform(FRFT)is used before the encryption scheme to get a large degree of randomization.The permutation is achieved by Baker map and the substitution by a key-related-to-plain-image algorithm based on the modified Logistic map.Modification of the Logistic map is developed to increase the space of the encryption key,and hence increase the security.The key of the encryption algorithm dependents on the plain image,and thus,the cipher image is sensitive to both the initial key and the plain image to resist known-plaintext and chosen plaintext attacks.The key space is large and hence the algorithm can effectively resist brute-force attacks.The proposed scheme is examined using different performance evaluation metrics and the results prove that the proposed scheme is highly secure,and it can effectively resist different attacks.