四种多变量校准方法在FTIR多组分分析中的性能比较

本文对四种多变量校准方法－－经典最小二乘法（CLS），偏最小二乘法（PLS），卡尔曼滤波法（KFM）以及人工神经网络法（ANN）－－在多组分浓度分析方面的性能进行了比较。选择五种红外谱图严重混叠的大气有机毒物－－1，3－丁二烯，苯，邻二甲苯，氯苯和丙烯醛－－作为分析对象。分别计算各种方法对该5组分体系的平均预测误差MPE和平均相对误差MRE进行比较。结果表明，偏最小二乘法在处理这类问题中是最稳健的方法。     

A molecular network model to describe the nonlinear viscoelastic behavior of polymers

A transient molecular network model is built to describe the nonlinear viscoelasticity of polymers by considering the effect of entanglement loss and regeneration on the relaxation of molecular strands. It is an extension of previous network theories. The experimental data on three thermoplastic polymers (ABS, PVC and PA6) obtained under various loading conditions are used to test the model. Agreement between the theoretical and experimental curves shows that the suggested model can describe successfully the relaxation behavior of the thermoplastic polymers under different loading rates by using relatively few relaxation modes. Thus the micromechanism responsible for strain-rate dependence of relaxation process and the origin of nonlinear viscoelasticity may be disclosed. The project supported by the National Natural Science Foundation of China and Doctorial Fund     

偏最小二乘回归的贝叶斯正则化神经网络集成模型在证券分析预测中的应用

神经网络集成技术能有效地提高神经网络的预测精度和泛化能力,已经成为机器学习和神经计算领域的一个研究热点.利用Bagging技术和不同的神经网络算法生成集成个体,并用偏最小二乘回归方法从中提取集成因子,再利用贝叶斯正则化神经网络对其集成,以此建立上证指数预测模型.通过上证指数开、收盘价进行实例分析,计算结果表明该方法预测精度高、稳定性好.     

基于BP神经网络的企业未来获利能力智能综合评价

分析了相关分析——多指标综合评价法在确定企业未来获利能力方面的优点和不足 ;并在其基础上提出了基于 BP神经网络的多指标综合评价法 ;仿真试验证明了基于 BP神经网络的多指标综合评价法的有效性     

人工神经网络在SARS疫情分析与预测中的应用

讨论人工神经网络在 SARS疫情分析与预测中的应用 .采用三层结构的反向传播网络 ( Backpropagation network,简称 BP网络 ) ,对 SARS在中国的传播与流行趋势及控制策略建立了网络模型 .并利用实际数据拟合参数 ,针对北京、山西的疫情进行了计算仿真 .结果表明 ,该网络模型算法收敛速度较快 ,预测精度很高     

求解无线传感器网络路由问题的蚁群最优化算法及其收敛性

首先将无线传感器网络的路由问题转化成求解最小Steiner树问题,然后给出了求解无线传感器网络路由的蚁群优化算法,并对算法的收敛性进行了证明．最后对找到最优解后信息素值的变化进行了分析．即在限制信息素取值的条件下,当迭代次数充分大时,该算法能以任意接近于1的概率找到最优解,并且当最优解找到后,最优树边上的信息素单调增加,而最优解以外边上的信息素在有限步达到最小值．     

The construction of infinite families of any k-tight optimal and singular k-tight optimal directed double loop networks

The double loop network(DLN)is a circulant digraph with n nodes and outdegree 2.It is an important topological structure of computer interconnection networks and has been widely used in the designing of local area networks and distributed systems.Given the number n of nodes,how to construct a DLN which has minimum diameter?This problem has attracted great attention.A related and longtime unsolved problem is:for any given non-negative integer k,is there an infinite family of k-tight optimal DLN?In this paper,two main results are obtained:(1)for any k≥0,the infinite families of k-tight optimal DLN can be constructed,where the number n(k,e,c)of their nodes is a polynomial of degree 2 in e with integral coefficients containing a parameter c.(2)for any k≥0, an infinite family of singular k-tight optimal DLN can be constructed.     

Improving the neural network method for finite element model updating using homogenous distribution of design points

In developing a neural network technique for a finite element model updating, researchers have been shown that the number of training samples and their quality, significantly affect the accuracy of the NN predication. In this study, based on the genetic algorithm (GA) method, we reduce the number of analyses required to develop the training pairs and reduce the amount of time for training the NN. In the other words, a uniform distribution of design points inside the design space will be obtained by means of this approach. To validate the efficiency of GA sample selection, random generation (RG) method is used for comparison. Two comparisons are made based on a numerical and experimental example. One is updated from 10 degrees of freedom lumped parameters system and the other is updated from a free–free beam using test data. The results indicate that the GA sample selection can reduce the number of training samples without affecting the accuracy of the NN predication. In our present study, also, the advantages of using frequency response function (FRF) data as input to the NN are kept and the drawback of having too many frequency points is overcome by the application of principal component analyses (PCA).     

电力市场中合同电量与竞争电量交易比例的研究

在单边开放的区域电力市场中,合理的合同电量与竞争电量交易比例是保证电力市场有效运行的一个重要环节。竞争电量所占的比例将主要取决于当前发电公司的市场行为。首先使用BP神经网络对电力需求弹性系数进行了预测,然后以长期电力市场均衡为目标函数,考虑贵州电网发电机组的可用容量与负荷预测的误差,以及贵州输电线路的可靠性诸因素,推导出合同电量与竞争电量交易比例,经过与南方区域电力市场目前运营规则规定的交易比例比较,该比例是合理的,可以规避电力市场价格波动等带来的风险。