基于改进神经网络的干点软测量

针对传统的神经网络收敛判断以模型的拟合精度为指标造成训练时间过长和过拟合等缺点,提出了一种改进神经网络(M-ANN).M-ANN将样本分成训练样本和校验样本,并提出了过拟合判据参数.通过训练样本采用误差反传算法对网络进行训练,训练过程中以模型对校验样本的预测性能为指标,通过过拟合判据参数的计算自适应地在获得具有最佳预测性能模型时终止网络训练.同时,针对影响初馏塔塔顶石脑油干点的因素众多且呈高度非线性的特征,应用M-ANN建立初顶石脑油干点软测量模型,获得模型的预测相对误差平方和均值比传统神经网络模型降低了27.5%.     

钴湿法冶炼萃取过程质量指标的软测量

利用基于物料衡算关系的机理模型、基于萃取平衡关系的半经验模型以及基于径向基函数-偏最小二乘法的萃余液的pH值预测模型,提出了一种钴湿法冶炼萃取过程质量指标软测量建模方法.机理模型用以描述过程的整体特性;半经验模型利用拟牛顿变尺度方法对其中未知参数进行估计,并用以逼近机理模型中的萃取平衡关系;基函数-偏最小二乘法则用来预测萃余液pH值.将所建立的软测量模型应用于实际钴湿法冶炼生产过程中,结果表明,该方法具有良好的预测精度.     

基于ELM的非线性PLS软测量方法研究

提出了一种基于极限学习机的非线性偏最小二乘软测量方法,有效地处理过程非线性、数据共线性以及在线快速建模等问题.该方法保留线性偏最小二乘外部模型以去除变量间的相关性,采用极限学习机建立内部模型来准确描述系统的非线性关系,与其它方法相比,该方法建模速度快,并具有良好的泛化能力.通过在苯乙烯反应器仿真及穿孔能耗预测应用结果表明,该方法建立的模型精度、泛化能力均优于传统方法,验证了其可行性和有效性.     

时延系统中T-LSSVR动态软测量建模方法研究

针对工业过程中由于系统存在延时导致软测量模型难以建立、模型精度偏低等问题,提出将系统延时(T)与最小二乘支持向量回归机(LSSVR)相结合,构建一种基于T-LSSVR的动态软测量建模方法;该方法在建模过程中利用互相关函数与一阶广义差分算法辨识得到“静态响应延时”和“动态响应延时”,通过软测量手段对变量进行预测以实现辅助变量对主导变量的最佳估计。对某化工企业具有此类双延时性质的系统进行实验,实验结果表明该建模方法在动态和稳态数据预测方面都有良好的预测效果。     

改进果蝇优化LSSVM超声波萃取产物浓度软测量

针对超声波天然产物萃取过程中产物浓度难以在线检测的问题,提出了一种改进果蝇优化最小二乘支持向量机(Least Squares Support Vector Machine,LSSVM)的超声波萃取产物浓度软测量建模方法。首先将混沌优化与迭代步长动态调节方法相融合,提出了一种混沌动态步长改进果蝇优化算法(Chaos Dynamic Step Fluit Fly Optimization Algorithm,CDSFOA),该算法引入动态调节因子对步长动态更新,并利用混沌优化实现各变量之间映射等操作,能够有效提高果蝇优化算法的收敛精度和收敛速度,然后利用CDSFOA对LSSVM进行参数寻优,构建最优CDSFOA-LSSVM软测量模型,最后利用超声波斛皮素萃取实验数据进行验证。结果表明,提出的模型不仅有较好的学习和泛化能力,而且具有良好的预测精度,可为超声波天然产物萃取工艺优化提供理论指导。     

装载机载重测量的支持向量机软测量建模方法

在阐述支持向量机(SVM)和最小二乘支持向量机(LS-SVM)的原理和算法并对两者的特点进行比较后,为装载机载重测量建立了基于LS-SVM的软测量模型,并从核函数选择以及核参数确定两方面阐述了LS-SVM软测量建模的过程,最后与RBF函数网络以及BP网络的软测量建模结果进行对比.仿真分析结果表明,LS-SVM同时兼顾了精度和泛化能力两方面的性能,其最大泛化误差仅为6.863 8×10-6,是适合装载机载重软测量的建模方法.     

工业过程软测量实现技术

文章研究用RBF神经网络建立实际生产过程模型,研究了RBFN的训练信息β值和σ值的选取,并改进了其选取方法及其在线学习方法。与改进前相比模型性能有明显提高。现场运行也表明所建立的模型简单、精度高,能满足实际生产要求。     

基于ABC-LSSVM的铝热连轧板凸度软测量建模

在铝热连轧板形控制中,板凸度是铝板带的重要指标之一。为了准确预测铝热连轧板凸度,提出了一种基于人工蜂群(ABC)和最小二乘支持向量机(LSSVM)的板凸度软测量模型。由于LSSVM的精度和泛化能力取决于模型参数的选择,故引入ABC进行参数优化。根据某厂1+4铝热轧现场采集的数据验证软测量模型的预测性能,并与GA-LSSVM模型和Marquardt模型做比较,仿真结果表明:建立的ABC-LSSVM板凸度软测量模型参数优化速度快、结构简单,并且具有较高精度。     

基于光纤传感技术的油液污染度软测量研究

采用基于Mie理论的激光散射法测量油液污染度时,油液污染度与监测参数——入射光强、散射光强、出射光强以及颗粒参数之间存在着复杂的非线性关系,给油液颗粒污染度的准确测量带来困难。利用支持向量机优良的非线性映射和强大的泛化能力,建立了一个基于最小二乘支持向量机的油液污染度软测量模型,给出了相应的系统结构和算法;仿真和实际运行结果表明基于LS-SVM的油液污染度软测量模型具有较高的估算精度与泛化能力,为油液污染度的在线测量提供了一种简单、可靠的新方法。     

偏最小二乘回归模型在织物厚度预测中的应用

织物厚度是织物服用性的重要影响因素之一,通过分析影响织物厚度的组织参数,建立了织物厚度预测的偏最小二乘回归模型,得到了很好的拟合结果。     

Radial point interpolation based finite difference method for mechanics problems

A radial point interpolation based finite difference method (RFDM) is proposed in this paper. In this novel method, radial point interpolation using local irregular nodes is used together with the conventional finite difference procedure to achieve both the adaptivity to irregular domain and the stability in the solution that is often encountered in the collocation methods. A least‐square technique is adopted, which leads to a system matrix with good properties such as symmetry and positive definiteness. Several numerical examples are presented to demonstrate the accuracy and stability of the RFDM for problems with complex shapes and regular and extremely irregular nodes. The results are examined in detail in comparison with other numerical approaches such as the radial point collocation method that uses local nodes, conventional finite difference and finite element methods. Copyright © 2006 John Wiley & Sons, Ltd.     

A point interpolation meshless method based on radial basis functions

A point interpolation meshless method is proposed based on combining radial and polynomial basis functions. Involvement of radial basis functions overcomes possible singularity associated with the meshless methods based on only the polynomial basis. This non‐singularity is useful in constructing well‐performed shape functions. Furthermore, the interpolation function obtained passes through all scattered points in an influence domain and thus shape functions are of delta function property. This makes the implementation of essential boundary conditions much easier than the meshless methods based on the moving least‐squares approximation. In addition, the partial derivatives of shape functions are easily obtained, thus improving computational efficiency. Examples on curve/surface fittings and solid mechanics problems show that the accuracy and convergence rate of the present method is high. Copyright © 2002 John Wiley & Sons, Ltd.     

基于径向基函数的点云岛屿孔洞自动修复

研究了利用点云获得的模型的孔洞修复,针对目前主要通过人工修复带有岛屿面片的孔洞耗时较长的问题,提出了一种基于径向基函数(RBF)自动修复岛屿孔洞的方法。该方法首先利用最小权重三角化法修复模型主体上的孔洞,其次计算模型主体上孔洞与岛屿面片的相关性,利用模型主体上孔洞和与其相关岛屿面片周围点来计算径向基函数,最后将粗修复后细分的点调整到径向基函数描述的曲面上。实验表明,与其他方法相比,该方法能快速、准确地修复缺陷模型。     

基于神经网络的混凝土断裂参数灰色软测量

断裂能和断裂区长度作为骨料的重要参数无法通过常规方法进行检测,基于神经网络提出了一种新的软测量方法,采用灰关联分析作为软测量建模中辅助变量选择的工具,对土木工程中常用的混凝土弯曲梁的断裂能和断裂区长度进行预测,结果表明,该方法具有较高精度,可满足实际需要。     

A meshfree Hermite‐type radial point interpolation method for Kirchhoff plate problems

A meshfree computational method is proposed in this paper to solve Kirchhoff plate problems of various geometries. The deflection of the thin plate is approximated by using a Hermite‐type radial basis function approximation technique. The standard Galerkin method is adopted to discretize the governing partial differential equations which were derived from using the Kirchhoff's plate theory. The degrees of freedom for the slopes are included in the approximation to make the proposed method effective in enforcing essential boundary conditions. Numerical examples with different geometric shapes and various boundary conditions are given to verify the efficiency, accuracy, and robustness of the method. Copyright © 2005 John Wiley & Sons, Ltd.     

误差非线性函数的变步长FxLMS算法研究

滤波-x最小均方(Filtered-x Least Mean Square,FxLMS)算法是主动噪声控制的经典算法,其存在收敛速度与稳态误差不可兼得的问题,解决方法之一是采用变步长FxLMS算法.总结了现有的基于误差非线性函数的变步长模型,并将其应用于FxLMS算法以改善算法性能.用三种常见的噪声作为参考输入信号进行...     

A hybrid radial boundary node method based on radial basis point interpolation

The hybrid boundary node method (HBNM) retains the meshless attribute of the moving least squares (MLS) approximation and the reduced dimensionality advantages of the boundary element method. However, the HBNM inherits the deficiency of the MLS approximation, in which shape functions lack the delta function property. Thus in the HBNM, boundary conditions are implemented after they are transformed into their approximations on the boundary nodes with the MLS scheme.This paper combines the hybrid displacement variational formulation and the radial basis point interpolation to develop a direct boundary-type meshless method, the hybrid radial boundary node method (HRBNM) for two-dimensional potential problems. The HRBNM is truly meshless, i.e. absolutely no elements are required either for interpolation or for integration. The radial basis point interpolation is used to construct shape functions with delta function property. So unlike the HBNM, the HRBNM is a direct numerical method in which the basic unknown quantity is the real solution of nodal variables, and boundary conditions can be applied directly and easily, which leads to greater computational precision. Some selected numerical tests illustrate the efficiency of the method proposed.     

Analysis of meshless local radial point interpolation (MLRPI) on a nonlinear partial integro-differential equation arising in population dynamics

In this paper the meshless local radial point interpolation (MLRPI) method is applied to simulate a nonlinear partial integro-differential equation arising in population dynamics. This PDE is a competition model in which similar individuals are competing for the same resources. It is a kind of reaction–diffusion equation with integral term corresponding to nonlocal consumption of resources. In MLRPI method, it does not require any background integration cells so that all integrations are carried out locally over small quadrature domains of regular shapes, such as circles or squares in two dimensions and spheres or cubes in three dimensions. The point interpolation method is proposed to construct shape functions using the radial basis functions. A one-step time discretization method is employed to approximate the time derivative. To treat the nonlinearity, a simple predictor–corrector scheme is performed. Also the integral term, which is a kind of convolution, is treated by the cubic spline interpolation. The numerical studies on sensitivity analysis and convergence analysis show that our approach is stable. Finally, two numerical examples are presented showing the behavior of the solution and the efficiency of the proposed method.     

A meshless local radial point collocation method for free vibration analysis of laminated composite plates

In this paper, a meshless local radial point collocation method based on multiquadric radial basis function is proposed to analyze the free vibration of laminated composite plates. This method approximates the governing equations based on first-order shear deformation theory using the nodes in the support domain of any data center. Natural frequencies of the laminated composite plates with various boundary conditions, side-to-thickness ratios, and material properties are computed by present method. The choice of shape parameter, effect of dimensionless sizes of the support domain on accuracy, convergence characteristics are studied by several numerical examples. The results are compared with available published results which demonstrate the accuracy and efficiency of present method.