复杂产品开发项目调度问题的模糊优化算法

对具有不确定时间参数的复杂产品开发项目调度问题,提出一种有效的模糊优化调度算法-基于预测的模糊BoP项目调度算法.首先,用模糊数表示不确定的时间参数,并构造相应的模糊数运算方法,对适合于确定性调度问题的BoP算法进行扩展,使其能处理模糊性时间参数.其次,修正了BoP算法中子项目调度方法,提高了算法的调度性能,降低了计算复杂度.大量的数值仿真实验表明,与基于启发式规则的调度算法相比,模糊BoP算法更适合于具有不确定时间参数的复杂产品开发项目调度问题.     

基于混沌时间序列的变电站设备温度预测

变电站输变线路和设备的温度变化能够反映其老化、负载过高等引起的安全隐患.通过对变电站设备温度数据的非线性分析和预测,实现对设备的有效预警,将避免事故引起的巨大损失.对变电站已测温度数据建立时间序列,利用小数据量法验证变电站设备温度时间序列的混沌特性.研究基于RBF神经网络的混沌时间序列预测并与神经网络预测进行对比,单步预测与多步预测结果均优于神经网络预测.仿真结论证明了基于神经网络的混沌时间序列预测方法的有效性.     

Hierarchical classification of protein function with ensembles of rules and particle swarm optimisation

This paper focuses on hierarchical classification problems where the classes to be predicted are organized in the form of a tree. The standard top-down divide and conquer approach for hierarchical classification consists of building a hierarchy of classifiers where a classifier is built for each internal (non-leaf) node in the class tree. Each classifier discriminates only between its child classes. After the tree of classifiers is built, the system uses them to classify test examples one class level at a time, so that when the example is assigned a class at a given level, only the child classes need to be considered at the next level. This approach has the drawback that, if a test example is misclassified at a certain class level, it will be misclassified at deeper levels too. In this paper we propose hierarchical classification methods to mitigate this drawback. More precisely, we propose a method called hierarchical ensemble of hierarchical rule sets (HEHRS), where different ensembles are built at different levels in the class tree and each ensemble consists of different rule sets built from training examples at different levels of the class tree. We also use a particle swarm optimisation (PSO) algorithm to optimise the rule weights used by HEHRS to combine the predictions of different rules into a class to be assigned to a given test example. In addition, we propose a variant of a method to mitigate the aforementioned drawback of top-down classification. These three types of methods are compared against the standard top-down hierarchical classification method in six challenging bioinformatics datasets, involving the prediction of protein function. Overall HEHRS with the rule weights optimised by the PSO algorithm obtains the best predictive accuracy out of the four types of hierarchical classification method.     

Investigation of viscosity effect on droplet formation in T-shaped microchannels by numerical and analytical methods

Both numerical and analytical models have been developed to explore the viscosity effect of the continuous phase on drop formation at a T-shaped junction in immiscible liquids. The effects of the generalized power law coefficient, the power law exponent and the yield stress on the mechanism of drop breakup, final drop size and frequency of drop formation are studied by using the numerical three-dimensional volume of fluid model. Droplets coalescence in Bingham fluids is observed in the beginning transient period. The effect of yield stress on drop extension is also discussed. Predictions of drop size by using an analytical force balance show satisfactory agreement with simulation results for Newtonian and power law fluids with different viscosity ratios. The approximation error associated with the analytical model for Bingham fluids is also acceptable. This analytical model can greatly shorten the prediction time as compared with the numerical model, which is helpful for on-line control.     

基于改进BP网络的当前非法出入境活动发展趋势预测

提出了基于BP神经网络的当前非法出入境活动发展趋势预测方法,以某年相关原始数据为依据,通过Matlab编程对当前非法出入境活动的发展趋势进行了预测和分析,为有关部门的正确决策与规划提供一定的指导和依据.     

高炉铁水硅含量预报的ICA-SVM建模方法

提出了一种基于改进的动态独立分量分析（independent component analysis, ICA）和支持向量机(support vector machine, SVM)的高炉铁水硅含量预报模型建模方法．采用动态ICA方法对样本数据进行特征提取,消除生产工艺参数之间的相关性．在此基础上,再使用目前计算复杂性较小的最小二乘SVM算法建立高炉铁水硅含量预报的动态递推模型,并引入了遗传算法以优化模型性能．以某钢厂高炉实际生产数据进行了应用实验,并与现有的时间序列分析、人工神经网络和基本SVM建模方法进行了对比．实验统计结果表明,本文方法显著提高了铁水硅含量的预测命中率．     

RCGNN：一种基于实数编码的遗传神经网络预测方法

将遗传算法与神经网络相结合,提出一种实数编码、自适应选择、算术交叉、高斯变异、爬山操作的改进遗传BP神经网络RCGNN,利用遗传算法对神经网络权值和阈值进行优化。以时间序列预测的实例进行编程计算表明,用遗传算法进行网络训练,其收敛速度快,最终总误差最小,预测准确率高。对算法中参数进行的相应研究表明,增加爬山操作次数能很好地提高网络训练的速度,同时使误差下降快;爬山操作越多,收敛速度越快,最终误差越小,但计算运行时间也会增加。     

基于相空间重构的汇率预测研究

汇率在宏观经济政策、商业经营和个人决策制定上的作用越来越重要,使其成为了研究的热点。根据混沌动力系统的相空间延迟坐标重构理论,基于支持向量机的强大的非线性映射能力,提出了一种基于支持向量机回归的超短期汇率预测方法,并建立了模型,对美元港币的即时汇率进行了实证计算,且与BP神经网络模型进行了比较。结果表明,所建立的模型能很好地跟踪即时汇率的变化趋势,预测精度比较高且算法运行速度比BP神经网络模型快得多。     

基于灰色预测模型的个人建房分析

灰色模型具有所需数据少、预测精度高和无需先验信息的特点.本文通过建立GM(1,1)模型对某区的农村个人建房进行预测,为相关职能部门提供科学的决策依据.结果表明灰色预测模型精度较高、预测误差较小、简捷实用.     

A method for rapidly assessing and refining simple solvent treatments in molecular modelling. Example studies on the antigen-combining loop H2 from FAB fragment McPC603

Different simple solvent models have been implemented in anextended simulated annealing process (ESAP), developed by Higoet al. [(1992) Biopolymers, 32, 33–43] and proven to beable to predict ab initio the conformation of the antigen-combiningloop H2 from FAB McPC603. Hie rationale used here provides auseful new method for testing solvent models in general. Thedifferent solvent models comprise a high dielectric constant,a screened coulomb potential, a dummy water model and a statisticalcontinuum treatment of the solvent effect in which the reactionfield and the solvent accessible area is accounted for. To assessthe effect of the solvent, we tested the ability of simulationsto retain the experimental conformation of loop H2. We comparedthe different structures obtained at the end of the annealingprocess in terms of root mean square deviation (r.m.s.d.), forboth the backbone and for all atoms, root mean square (r.m.s.)fluctuation, solvent accessible surface area (ASA), hydrogenbonding network and - plot distribution. The relationship betweenthe r.m.s.d. and the internal energy of a structure is alsoevaluated in terms of precision and another possible methodfor obtaining the best conformation is discussed. The accuracyof modelling the coarse effect of the solvent and the similaritiesof the resulting structures with respect to the X-ray referencestructure are examined. The possible choice of one of thesesolvent models in the structure determination of an unknownloop structure is discussed