基于隐Markov树的设备状态综合诊断模型

为了充分利用小波系数之间的统计相依性以更有效地诊断设备状态,提出了一种基于隐Markov树(HMT)的综合诊断模型。首先通过主成分分析将来自多个传感器的信号转换为主成分,求出各主成分对应的频谱,然后通过比较对已训练的各HMT模型的适应度,运用Bayes决策融合法则得到设备状态综合诊断决策。为了克服HMT模型存在的计算溢出困难,采用尺度变换对EM算法进行了改进。通过两个实例验证了该综合诊断模型具有较高的诊断准确率。     

高频雷达中提取海洋回波信号的新方法

雷达信号经过海面反射时,由于电离层的作用和外部的干扰,必然会有一些噪声夹杂在海洋回波中。噪声是一些无规则的频谱,对分析有用信号是极其不利的,严重时甚至淹没有用的信号。将噪声中的信号提取出来成了一个很关键的问题。基于海洋回波信号是1/f类分形信号的特性,利用不同尺度下小波系数方差来估计噪声方差及分形信号参数,进而从白噪声背景下估计出海洋回波信号。实验结果表明:用新方法处理过的海洋回波的多普勒谱的信噪比提高了10-15 dB。     

η_(e,k)~j算子作用下的小波变换误差分析

在研究拟微分算子及预小波基础上,探讨了jeη,k算子作用下的小波变换,得到了一些新的有用的结果.     

基于置乱变换的DCT域二维数字水印算法

提出了DCT变换域上基于人类视觉特性(HVS)的一种水印算法。该算法先把原始宿主图像进行幻方置乱,然后分块DCT,使原始宿主图像的DCT系数能均匀分布在系数矩阵中。添加水印时,首先对二值水印图像进行2×2分块,然后用伪随机序列对分块后的水印信号进行伪随机排序,最后再把排序后的水印图像进行Arnold置乱,并利用人眼视觉特性(HVS)对嵌入的水印强度进行了调整,把水印信号嵌入到原始图像的DCT系数的低频系数上。使水印的嵌入强度在不引起视觉警觉的基础上得到显著提高,有效地提高了检测正确率,增强了水印的稳健性。     

A Robust IP Packets Filtering Mechanism for Protecting Web Server from DDoS Attacks

Distributed denial of service （DDoS） attacks exploit the availability of Web servers, resulting in the severe loss of their connectivity. We present a robust IP packets filtering mechanism which combines the detection and filtering engine together to protect Web Servers from DDoS Attacks. The mechanism can detect DDoS attacks by inspecting inbound packets with an IP address database, and filter out lower priority IP addresses to preserve the connection for valid users by monitoring the queues status. We use the Netfilter＇s technique, a framework inside the Linux 2.4. X, to implement it on a Web server. Also, we evaluate this mechanism and analyze the influence of some important parameters on system performance. The experimental results show that this mechanism is effective against DDoS attacks.     

基于时间轴小波变换的运动对象分割算法研究

对时间轴一维小波变换的视频运动对象分割算法进行了研究。将视频序列进行时间轴一维小波变换，利用变换后的高频帧信息提取出初步的运动掩模图像；进行数学形态学后处理，以消除各种噪声的影响和对象的不连通性，得到理想的运动掩模图像。实验结果表明该方法分割效果较好，分割出的运动对象可直接应用于基于对象的视频编码，且算法耗时较少。     

The Lifting Scheme Based on the Second Generation Wavelets

0 IntroductionTfhaect w tahvaetl e,t“ twheaovreyle thsas ar mea noypt ip rmoaple rbtaiseess w fhoirch coremspurltesisnintghe,esti mating,and recoveringfunctions”[1]. The traditional wave-let defined as the dyadic translates and dilates of one particularfunction, which relies on the Fourier transform, suitable forregular sampling data. But it is helpless for the application ofli mit region,curve surface and not-even sampling. Then W.Sweldens[1]and other researchers develop some other method…     

Digital Watermarking Algorithm Based on Wavelet Transform and Neural Network

An effective blind digital watermarking algorithm based on neural networks in the wavelet domain is presented. Firstly, the host image is decomposed through wavelet transform. The significant coefficients of wavelet are selected according to the human visual system （HVS） characteristics. Watermark bits are added to them. And then effectively cooperates neural networks to learn the characteristics of the embedded watermark related to them. Because of the learning and adaptive capabilities of neural networks, the trained neural networks almost exactly recover the watermark from the watermarked image. Experimental results and comparisons with other techniques prove the effectiveness of the new algorithm.     

电力负荷预测技术及其发展趋势

介绍了电力系统负荷预测方法的几种分类方法、神经网络负荷预测方法及其在短期负荷预测中的研究情况,也指出了负荷预测技术的发展趋势。     

N-策略M/G/1/∞可修排队系统的队长

可修排队系统是一种重要的排队论模型.讨论服务台可修的N-策略M/G/1/∞排队系统的队长,利用L-S变换和母函数法,得到了任意时刻系统队长的表达式以及系统达到平衡时的队长.