基于T/4分数间隔的盲均衡算法研究

为了克服波特间隔常数模算法和T/2（T为字符间隔）分数间隔常数模算法收敛速度慢、稳态误差大的缺点,提出了基于T/4分数间隔的常数模盲均衡新算法,推导了4倍过采样多速率系统和1输入4输出多信道系统模型的公式。该算法采用T/4分数间隔均衡器采样,其采样频率大于奈奎斯特频率,从而避免了因欠采样引起的频谱混叠,均衡器可以更有效地对失真信道进行补偿。与波特间隔、T/2分数间隔均衡器相比,该算法收敛速度快、稳态误差小。多径水声信道盲均衡的仿真实验,验证了该算法的性能。     

一种新的基于身份的盲签名方案

盲签名使得用户可以得到签名人的签名而不让签名人知道具体被签的消息是什么,它在诸如电子现金、电子拍卖、电子选举等场合具有广泛的应用.给出了一种新的基于身份的盲签名方案,该签名方案的安全性是基于CDH问题.     

国产与进口头孢吡肟治疗急性细菌性感染的临床疗效及安全性比较

目的：评价国产头孢吡肟治疗急性细菌性感染的有效性和安全性.方法：采用多中心,随机、双盲对照临床试验设计,选用142例18～65岁的病人,可评价疗效病例139例,随机分为试验组（国产头孢吡肟）70例和对照组（进口头孢吡肟）69例.呼吸系统感染采用国产和进口头孢吡肟,均为2g.而泌尿系统感染采用国产和进口头孢吡肟为每次1g,分别加入50～100ml 0.9%氯化钠注射液或5%葡萄糖注射液中,每日2次静点,疗程均为7～14d.结果：试验组与对照组的痊愈率分别为50%和41%,有效率分别为93%和91%,细菌阳性率分别为89%与87%,细菌清除率分别为100%,98%,不良反应发生率分别为4%与9%,两组比较,无显著性差异（P〉0.05）.结论：国产头孢吡肟治疗急性细菌性感染安全有效,与进口头孢吡肟疗效相当.     

一种安全有效的盲签名体制

为了满足电子招投标系统的安全需求和提高它的运行效率，在分析椭圆曲线公钥密码、自认证公钥、Nyberg-Rueppel签名方案和位委托协议的基础上，设计了一种安全有效的盲签名方案，并对该方案的正确性、有效性和安全性进行了必要分析和证明．     

基于恒模算法的Hopfield神经网络盲多用户检测算法的研究

将恒模算法与Hopfield神经网络的能量函数相结合,提出了基于连续Hopfield神经网络的盲多用户信号检测算法。     

基于动量项前馈神经网络盲均衡算法

针对基于前馈神经网络的盲均衡算法中，BP优化算法具有收敛速度慢、易陷入局部极小的缺点，提出了一种新的盲均衡算法，该算法结合动量项前馈神经网络与传统恒模盲均衡算法的优点，将以前权值的调节量用于当前权值的修改过程，降低了算法对于误差曲面局部极值点的敏感性。仿真结果表明，该算法可有效抑制网络陷入局部极小，防止振荡，加快盲均衡器的收敛速度。     

基于误差信号自相关函数的变步长恒模盲均衡算法研究

提出了一种自适应变步长恒模盲均衡算法，利用剩余误差信号的自相关函数估计值作为控制步长的因子来自适应改变步长的大小，克服了恒模算法存在的固有缺陷。理论分析和仿真实验均表明该算法与恒模算法相比，在收敛过程中加快了收敛速度，收敛后又能减小对均衡器参数的误调，具有更小的稳态剩余误差。     

Yarn Properties Prediction Based on Machine Learning Method

Although many works have been done to constructprediction models on yarn processing quality, the relationbetween spinning variables and yam properties has not beenestablished conclusively so far. Support vector machines(SVMs), based on statistical learning theory, are gainingapplications in the areas of machine learning and patternrecognition because of the high accuracy and goodgeneralization capability. This study briefly introduces theSVM regression algorithms, and presents the SVM basedsystem architecture for predicting yam properties. Model.selection which amounts to search in hyper-parameter spaceis performed for study of suitable parameters with grid-research method. Experimental results have been comparedwith those of artificial neural network(ANN) models. Theinvestigation indicates that in the small data sets and real-life production, SVM models are capable of remaining thestability of predictive accuracy, and more suitable for noisyand dynamic spinning process.     

Analysis of Petri net model and task planning heuristic algorithms for product reconfiguration

Reconfiguration planning is recognized as an important factor for reducing the cost of manufacturing reconfigurable products, and the associated main task is to generate a set of optimal or near-optimal reconfiguration sequences using some effect algorithms. A method is developed to generate a Petri net as the reconfiguration tree to represent two-state-transit of product, which solved the representation problem of reconfiguring interfaces replacement. Relating with this method, two heuristic algorithms are proposed to generate task sequences which considering economics to search reconfiguration paths effectively. At last, an objective evaluation is applied to compare these two heuristic algorithms to other ones. The developed reconfiguration task planning heuristic algorithms can generate better strategies and plans for reconfiguration. The research finds are exemplified with struts reconfiguration of reconfigurable parallel kinematics machine （RPKM）.     

ON-LINE PREEMPTIVE SCHEDULING ON UNIFORM MACHINES

1 IntroductionWe consider the problem of sChedallng a 1ist (J1, J2,' .) of on-line jobs preemPtive1y onm unifornily related machines. In suCh a setting, the jobs arrive one by one. Job Jj becomesknown (with its etistence and its processing time Pj ) only When job Jj--1 has already beenscheduled, which gives rise to the name online schedallng. Job processing can be preemPtive,i.e., the processing of any job can be interruPted and resumed later. The machines are paralleI,which allows any …