A model, a heuristic and a decision support system to solve the scheduling problem of an earth observing satellite constellation

China plans to launch four small optical satellites and four small SAR satellites to form a natural disaster monitoring constellation. Data can be obtained by the constellation in all weather conditions for disaster alert and environmental damage analysis. The scheduling problem for the constellation consists of selecting and timetabling the observation activities to acquire the requested images of the earth surface and scheduling the download activities to transmit the image files to a set of ground stations. The scheduling problem is required to be solved every day in a typical 1-day horizon and it must respect complex satellite operational constraints as well as request preferences, such as visibility time windows, transition time between consecutive observations or downloads, memory capacity, energy capacity, polygon target requests and priorities. The objective is to maximize the rewards of the images taken and transmitted. We present a nonlinear model of the scheduling problem, develop a priority-based heuristic with conflict-avoided, limited backtracking and download-as-needed features, which produces satisfactory feasible plans in a very short time. A decision support system based on the model and the heuristic is also provided. The system performance shows a significant improvement with respect to faster and better scheduling of an earth observing satellite constellation.     

Comparison of experimental designs for simulation-based symbolic regression of manufacturing systems

In this article, an empirical analysis of experimental design approaches in simulation-based metamodelling of manufacturing systems with genetic programming (GP) is presented. An advantage of using GP is that prior assumptions on the structure of the metamodels are not required. On the other hand, having an unknown structure necessitates an analysis of the experimental design techniques used to sample the problem domain and capture its characteristics. Therefore, the study presents an empirical analysis of experimental design methods while developing GP metamodels to predict throughput rates in a common industrial system, serial production lines. The objective is to identify a robust sampling approach suitable for GP in simulation-based metamodelling. Experiments on different sizes of production lines are presented to demonstrate the effects of the experimental designs on the complexity and quality of approximations as well as their variance. The analysis showed that GP delivered system-wide metamodels with good predictive characteristics even with the limited sample data.     

A review of optimization methodologies in support vector machines

Support vector machines (SVMs) are theoretically well-justified machine learning techniques, which have also been successfully applied to many real-world domains. The use of optimization methodologies plays a central role in finding solutions of SVMs. This paper reviews representative and state-of-the-art techniques for optimizing the training of SVMs, especially SVMs for classification. The objective of this paper is to provide readers an overview of the basic elements and recent advances for training SVMs and enable them to develop and implement new optimization strategies for SVM-related research at their disposal.     

基于相对熵的直方图差异与JPEG隐写的定量分析

针对图像隐写分析中常用的直方图特征,给出了一类基于相对熵的直方图差异计算方法,并提出了一种基于相对熵的JPEG隐写的定量分析方法.首先根据两假设检验中的最优检验——似然比检验,分析了相对熵在衡量2个直方图间的距离时的优越性,并给出了2种基于相对熵的直方图差异计算方法.然后,以新的直方图差异特征为基础,训练支持向量回归分析器,以估计隐写对DCT系数的更改比率.针对JSteg和改进的F5隐写的实验结果表明:与其他的直方图差异特征相比,根据所提出的基于相对熵的直方图差异特征所训练的定量隐写分析器具有更高的估计精度和稳定性.     

支持向量机的进化多核设计

为提高支持向量机分类精度,提出一种基于遗传程序设计的进化多核算法.算法中每个个体表示一个多核函数,并采用树形结构进行编码,增强了多核函数的非线性;初始种群由生长法产生,经过遗传操作后得到适合具体问题的进化多核函数.遗传程序设计的全局搜索性能使得算法设计不需要先验知识.与单核函数及其他多核函数的对比实验结果表明,进化多核...     

基于信息支持设备的电路板元器件定位软件开发

信息支持设备是一种便携式的辅助维修设备,具有处理能力强、人机交互界面好、屏幕分辨率高的特点。在此平台上,利用C#语言,针对故障电路板有无PCB格式两种类型电路图进行了元器件定位软件开发,为进一步检查和测试提供有用方位信息,大大缩短了维修时间。     

基于最小二乘支持向量机的机器视觉识别方法

为了解决传统的机器视觉识别技术识别精度低的难题,提出基于粒子群优化最小二乘支持向量机的机器视觉识别方法.首先,对机器视觉采集的图像进行特征提取;然后,利用特征数据建立基于粒子群优化最小二乘支持向量机的识别模型;最后,以红枣缺陷识别作为应用案例以证明该方法的有效性及优越性.分别采用人工神经网络、支持向量机与该方法进行对比...     

基于最大似然线性回归矩阵的说话人识别算法研究

研究了将自适应领域的最大似然线性回归(Maximum likelihood linear regression, MLLR)变换矩阵作为特征进行文本无关的说话人识别算法. 本文引入了基于统一背景模型的MLLRSV-SVM说话人识别算法, 并在此基础上进行高层音素聚类以进一步提高识别性能. 在采用多种信道补偿技术后, 在NIST SRE 2006年1训练语段-1测试语段同信道和跨信道数据库上, 基于MLLR特征的系统与其他最好的系统性能接近并有很强的互补性, 经过简单线性融合可以极大提高识别性能.     

基于LS-SVR的图像噪声去除算法研究

通过对最小二乘支持向量机(Least squares support vector regression, LS-SVR)滤波特性的分析, 给出了LS-SVR用于图像滤波的卷积模板构造方法, 解决了LS-SVR在应用中需要求解的问题, 在此基础上, 提出了基于LS-SVR的开关型椒盐噪声滤波算法. 滤波算法中以Maximum-minimum算子作为椒盐噪声检测器, 利用滤波窗口内非噪声点构成LS-SVR的输入数据, 使用事先构造出的LS-SVR滤波算子, 对滤波窗口进行简单的卷积运算, 实现了被椒盐噪声污染点数据的有效恢复. 实验表明, 本文提出的方法具有较好的细节保护能力和较强的噪声去除能力.     

用兴趣点凸包和SVM加权反馈实现图像检索

针对采用环状颜色直方图的图像检索方法存在的不足,提出一种基于兴趣点凸包的图像特征提取方法,通过对用小波变换检测出的必趣点递归求出它们的凸包,并将每个凸包上的兴趣点按一定的算法安插在相应的桶内,对每个桶求出颜色直方图,利用桶与桶之间的相似度定义两幅图像的相似度.这种特征提取方法可有效抑制兴趣点集合中出现游离兴趣点的情况,结合基于兴趣点的空间离散度和Gabor小波纹理等特征实现图像检索,可有效提高图像检索精度.最后,提出一种新的相关反馈方法,通过利用支持向量机分类结果设置权值来改进移动查询点相关反馈方法.实际图像数据库上的实验表明,引入这种反馈方法后可将图像检索的查准率提高20%左右,查全率提高10%左右.