基于径向基神经网络的航空发动机气路故障诊断

基于径向基神经网络对民用高涵道比航空发动机风扇、增压级、高压压气机、高压涡轮、低压涡轮5大气路部件的效率降低故障进行诊断。采用Gasturb进行故障训练样本和测试样本库的生成,诊断结果显示,采用径向基神经网络进行航空发动机气路故障诊断的计算时间短、精度较高,不仅能定性的定位故障部位,而且在大多数情况下可以定量的给出该部件的性能衰退程度。某些情况下诊断结果与测试样本不尽一致,但都是方程的合理解,这是因为航空发动机的数学模型是一个多解的复杂方程,一个总性能的衰减对应着多组部件性能衰退的组合。随噪声幅值加大,诊断精度变差,同时研究发现诊断精度受噪声影响的敏感系数在不同的噪声幅值水平下是不同的。     

Best estimates of RBF-based meshless Galerkin methods for Dirichlet problem

In this paper, Galerkin methods based on the radial basis functions to deal with the partial differential equations are discussed. The best error estimates for this method are obtained.     

基于RBF神经网络的Ni-TiN镀层耐蚀性能预测研究

采用脉冲电沉积方法在40Cr钢表面制备Ni-TiN复合镀层,并以TiN粒子浓度、电流密度以及占空比为输入层,以Ni-TiN复合镀层腐蚀量为输出层,建立RBF神经网络模型,对镀层腐蚀量进行预测研究,最后利用扫描电镜观察不同工艺参数下镀层表面形貌.结果表明,RBF神经网络对镀层腐蚀量有较强的预测能力,其预测值与实验值相对误差最小仅为0.73;;SEM分析表明,当TiN粒子浓度10 g/L,电流密度5 A/dm2,占空比60;时,Ni-TiN复合镀层经腐蚀后表面较为平整,腐蚀坑较少,耐腐蚀性能较好.     

Integrated radial basis functions‐based differential quadrature method and its performance

In this paper, indirect radial basis function networks (IRBFN) proposed by Nam and Tranh (Neural Networks 2001; 14 (2):185–199; Appl. Math. Modelling 2003; 27 :197–220) are incorporated into the differential quadrature (DQ) approximation of derivatives. For simplicity, this new variant of RBF‐DQ approach is named as iRBF‐DQ method. The proposed approach is validated by its application to solve the one‐dimensional Burger's equation, and simulate natural convection in a concentric annulus by solving Navier–Stokes equations. It was found that as compared to the benchmark data, the iRBF‐DQ approach can provide more accurate results than the original RBF‐DQ method. Copyright © 2006 John Wiley & Sons, Ltd.     

基于遗传RBF网络的惯性测量组合模拟电路故障诊断

为解决惯性测量组合模拟电路的诊断不易定位到元件级故障的问题,提出了一种基于遗传RBF网络的智能诊断方法。该方法首先利用RBF神经网络快速准确识别故障的能力,以RBF的训练均方误差为遗传算法的适应度函数,依靠遗传算法强大的全局寻优能力实现故障特征选择。在特征选择的过程中,同时记录使训练均方误差达到最小的最优RBF网络,然后直接利用特征选择过程中训练好的最优RBF网络诊断故障,而无需利用特征选择后的训练数据对RBF网络进行再训练,简化了诊断步骤,同时增强了网络的抗干扰能力。仿真结果表明,该方法能有效去除冗余特征,准确诊断惯性测量组合模拟电路的故障,并有较好的抗噪能力,证明了该方法的有效性和可行性。     

A tutorial on support vector machine-based methods for classification problems in chemometrics

This tutorial provides a concise overview of support vector machines and different closely related techniques for pattern classification. The tutorial starts with the formulation of support vector machines for classification. The method of least squares support vector machines is explained. Approaches to retrieve a probabilistic interpretation are covered and it is explained how the binary classification techniques can be extended to multi-class methods. Kernel logistic regression, which is closely related to iteratively weighted least squares support vector machines, is discussed. Different practical aspects of these methods are addressed: the issue of feature selection, parameter tuning, unbalanced data sets, model evaluation and statistical comparison. The different concepts are illustrated on three real-life applications in the field of metabolomics, genetics and proteomics.     

人工神经网络在溶解度预测方面的应用

溶解度作为一项重要的物化指标,一直是化学学科的研究重点。然而,通过实验测量获得数据耗时费力,因此,科研人员建立了多种理论方法来进行估算,其中,人工神经网络因其能够关联复杂的多变量情况而受到广泛关注。本文综述了人工神经网络在物质溶解度预测方面的应用,介绍了应用最广泛的3种神经网络(BP神经网络、小波神经网络、径向基神经网络)的模型结构、预测方法和预测优势,探讨了神经网络的不足以及改进方法。文章最后对神经网络在物质溶解度预测方面的发展前景进行了展望。与其他方法相比,人工神经网络技术在物质溶解度预测方面具有预测结果精确度高、操作简单等特点,具有广阔的应用前景,但输入变量选择、隐含层节点数确定、避免局部最优等问题还需逐步建立系统的理论指导。     

Propagation of premixed laminar flames in 3D narrow open ducts using RBF-generated finite differences

Laminar flame propagation is an important problem in combustion modelling for which great advances have been achieved both in its theoretical understanding and in the numerical solution of the governing equations in 2D and 3D. Most of these numerical simulations use finite difference techniques on simple geometries (channels, ducts, ...) with equispaced nodes. The objective of this work is to explore the applicability of the radial basis function generated finite difference (RBF-FD) method to laminar flame propagation modelling. This method is specially well suited for the solution of problems with complex geometries and irregular boundaries. Another important advantage is that the method is independent of the dimension of the problem and, therefore, it is very easy to apply in 3D problems with complex geometries. In this work we use the RBF-FD method to compute 2D and 3D numerical results that simulate premixed laminar flames with different Lewis numbers propagating in open ducts.     

小潜艇供电系统全桥DC-DC开关电源智能优化设计

小型潜艇水下潜行采用蓄电池供电,通过全电力供电系统推进。以蓄电池为直流电源,在推进系统正常运转的状况下,开关电源电压输出值保持恒定。传统的PID控制器很难对全桥DC-DC开关电源进行精确控制,因此,将粒子群算法(PSO)、BP神经网络与径向基函数(RBF)神经网络与传统PID控制相结合,提出带有RBF神经网络辨识的PSO-BP-PID控制方法。通过神经网络在线自学习对PID三个参数在线调整,最终实现系统恒电压输出控制。仿真结果得出：带有RBF神经网络辨识的PSO-BP-PID控制算法可以很大的改善系统控制效果,同时使系统具有更好的在线调整能力。