基于拉丁超立方抽样及响应面的结构模糊分析

在对边值模糊分析方法和优化边值模糊分析方法研究的基础上,将拉丁超立方抽样和响应面方法引入结构模糊分析,提出一种基于拉丁超立方抽样及响应面的模糊分析方法。讨论这种方法的数学运算,并与Monte Carlo方法进行比较。最后基于这种方法,对一个用隐式方程描述的有限元悬臂梁模型进行模糊分析,得到变形模糊结果。     

An improved three-point method based on a difference algorithm

The reason for the rounding error in a measurement system under the condition of a limited sampling period is analysed, and a three-point method based on the difference algorithm (D3P method) is proposed. The effect of the rounding error is reduced by differential approximation of the roundness error and reconstruction of the transfer function. The generalized three-point method (G3P method), the three-point method based on the first-order difference algorithm (I-D3P method) and the three-point method based on the second-order difference algorithm (II-D3P method) are compared and analysed by simulation. The results show that the accuracy of the D3P method is greatly improved compared with that of the G3P method under the restricted sampling period and that the D3P method retains high precision in the case of failed error separation of the G3P method. An experimental system was constructed to realize error separation, and the experimental results show that the D3P method has higher accuracy than the G3P method over a small sampling period.     

基于间隙热源法的车轮踏面制动摩擦温升研究

针对传统模拟车轮踏面制动摩擦热流边界的加载方式存在的不足,如均布热源法计算精度较低和移动热源法计算效率较低,提出基于间隙热流加载法的车轮踏面制动摩擦温升有限元计算方法。以重载列车紧急制动为研究对象,通过与以往的移动热源法和均布热源法进行对比,研究新方法的计算精度和计算效率。计算结果显示:在计算精度方面,新方法不仅可以达到移动热源法的计算精度,还能体现均布热源法无法模拟的因闸瓦和踏面的接触-脱离现象导致的车轮踏面温度时间历程曲线波动现象。在计算时间方面,新方法的计算效率与均布热源法的相当,但新方法的计算速度比移动热源法的快4.3倍。     

Accuracy analysis of Stewart platform based on interval analysis method

A Stewart platform is introduced in the 500 m aperture spherical radio telescope(FAST) as an accuracy adjustable mechanism for feed receivers.Accuracy analysis is the basis of accuracy design.However,a rapid and effective accuracy analysis method for parallel manipulator is still needed.In order to enhance solution efficiency,an interval analysis method(IA method) is introduced to solve the terminal error bound of the Stewart platform with detailed solution path.Taking a terminal pose of the Stewart platform in FAST as an example,the terminal error is solved by the Monte Carlo method(MC method) by 4 980 s,the stochastic mathematical method(SM method) by 0.078 s,and the IA method by 2.203 s.Compared with MC method,the terminal error by SM method leads a 20% underestimate while the IA method can envelop the real error bound of the Stewart platform.This indicates that the IA method outperforms the other two methods by providing quick calculations and enveloping the real error bound of the Stewart platform.According to the given structural error of the dimension parameters of the Stewart platform,the IA method gives a maximum position error of 19.91 mm and maximum orientation error of 0.534°,which suggests that the IA method can be used for accuracy design of the Stewart platform in FAST.The IA method presented is a rapid and effective accuracy analysis method for Stewart platform.     

Free vibration analysis of straight-line beam regarded as distributed system by combining Wittrick-Williams algorithm and transfer dynamic stiffness coefficient method

We developed a method for analyzing the free vibration of a structure regarded as a distributed system, by combining the Wittrick-Williams algorithm and the transfer dynamic stiffness coefficient method. A computational algorithm was formulated for analyzing the free vibration of a straight-line beam regarded as a distributed system, to explain the concept of the developed method. To verify the effectiveness of the developed method, the natural frequencies of straight-line beams were computed using the finite element method, transfer matrix method, transfer dynamic stiffness coefficient method, the exact solution, and the developed method. By comparing the computational results of the developed method with those of the other methods, we confirmed that the developed method exhibited superior performance over the other methods in terms of computational accuracy, cost and user convenience.     

弹塑性力学问题的无网格法分析

提出弹塑性力学问题的无网格局部Petrov—Galerkin(meshless local Petrov—Galerkin，MLPG)方法，这是一种真正的无网格方法。这种方法采用移动最小二乘近似函数作为试函数，并且采用移动最小二乘近似函数的权函数作为加权残值法的权函数，本质边界条件用罚因子法施加。文中采用Newton—Raphson法进行计算。计算实例表明．局部Petrov—Galerkin方法是一种很有效的求解弹塑性力学问题的方法。     

基于代理模型的变复杂度方法在板料成形优化中的应用

板料成形优化中存在一些缺陷问题,如一步法与优化算法相结合时精度低、增量法与优化算法直接结合时效率低以及通过传统构造代理模型的方法构造高精度的代理模型时需要大量增量法模拟结果。为了充分发挥一步法和增量法各自的优点,提出采用变复杂度方法通过较少次数的试验样本点数据先在一步法和增量法间建立一个差值补偿响应面模型。通过一步法和差值补偿响应面模型构造新的试验样本点,在新构样本点和原有用增量法计算的样本点数据基础上建立移动最小二乘法代理模型,建立的移动最小二乘法代理模型精度和效率都较高。利用粒子群优化算法对移动最小二乘法代理模型进行优化求解。将该方法应用到了汽车某内板的成形性优化中,优化的结果显著地提高了板料的成形性。算例表明,该方法具有较高的精度和较强的工程实用性。     

随机结构分析的加权残值方法

基于矩法和加权残值法两者的特点 ,提出了随机结构分析的一种新方法。该法是将加权残值法和概率论中的矩法相结合 ,对随机结构问题进行力学分析和计算。此法简便、快捷 ,且能获得结构各个参数的随机性对结构反应的影响。通过梁和板两算例 ,证明了方法的可行性和有效性。     

Three dimension profile extrusion simulation using meshfree method

A meshfree method based on reproducing kernel approximation and point collocation is presented for analysis of 3D profile extrusion processes. The point collocation method is a true meshfree method without the employment of a background mesh. We show that, in a point collocation approach, the implementation of program does not need the background mesh, which is a very time consuming process in a Galerkin method for integration. Among the points of the meshfree model, there is no connecting information. A mesh quality control method with mold interpenetration checking based on the Delaunay Bowyer-Watson algorithm is introduced to produce the topological relations, which enable us to show meshfree simulation results using the same procedure as in the finite element method. A C-shape profile extrusion application example is presented and compared with the finite element method to verify the proposed method.     

新型灰关联分析方法研究及其在数控机床主轴故障识别中的应用

针对传统灰关联分析方法存在的问题,引入动态分辨系数和因子权重系数,提出新型灰关联分析方法。与传统灰关联分析方法相比,该方法具有两个优点:一是降低对人为确定分辨系数和权重系数的依赖性;二是提高识别结果的可靠性和准确性。最后将该方法应用于数控机床主轴故障识别中,并与传统灰关联分析方法和神经网络识别结果进行对比分析。结果显示,新型灰关联分析方法识别结果更准确可靠。     

浅谈保证装配精度的装配方法

产品是由许多零件组合而成的。尺寸精度反映装配中各相关零件的尺寸精度与装配精度的关系;相对位置精度反映装配中各有关零件的位置精度与装配精度的关系。根据不同的机械、不同的生产类型条件,选用不同的装配工艺方法。装配工艺方法归纳起来有4种:互换装配法、选配装配法、修配装配法和调整装配法。     

可靠性分析的最大可能点摄动法

讨论随机变量间不相互独立的可靠性计算问题。基于摄动法和可靠性理论,针对JC法在计算结构可靠度时不能直接处理不相互独立的可靠性计算问题,提出了一种扩展的JC法即最大可能点摄动法,该方法继承JC法收敛快、精度高等特点,同时也克服了JC法对相关变量的求解过程中需要正交变换的要求。编写相应计算程序,实例计算结果与蒙特卡洛数值模拟结果的对比验证了最大可能点摄动法的正确性、有效性。     

线性时滞系统区域极点配置的数值方法

介绍了一种应用于线性时滞系统的区域极点配置的数值方法,它实质上是应用于线性定常无时滞系统的传统极点配置方法的推广应用.由于时滞系统的特征方程是超越方程,故有无穷多个特征值.该方法的目标是通过进行数值方法的不断迭代,来不断整合反馈增益阵,最终把最右端的一批极点配置到指定的区域中去.通过引入对带有输入迟延和状态迟延的线性系统进行配置的实例来对这一方法进行说明.实践证明,该配置方法能取得较好的配置效果.     

基于流形-奇异值熵的滚动轴承故障特征提取

提出一种基于流形-奇异值熵的滚动轴承时频故障特征提取方法。首先,在HHT(Hilbert-Huang transform,简称HHT)时频分析基础上,应用二维流形方法提取信号流行成分以达到对轴承故障特征进行降维和提取敏感参量的目的;然后,定义了奇异值熵来定量衡量不同故障状态下流行成分的差异;最后,将流形奇异值向量与概率神经网络相结合,有效实现了轴承故障样本分类。与一般的考虑欧式空间全局范围最优值的主分量(principal component analysis,简称PCA)方法及以向量为研究对象的一维流形方法不同,该方法直接以二维信息为研究对象,避免了一维流形算法需将二维信息转化为向量带来的信息损失,与PCA方法相比更能发现隐藏在高维数据流形结构中的局部数据特征。工程信号分析验证了该方法的有效性,为准确提取滚动轴承故障特征提供了一种可靠手段。     

模拟带材轧制过程的条元法

本文介绍了模拟带材轧制过程三维应力与变形的一个新的工程实用方法——条元法。本方法对于四辊轧机冷轧带材单位轧制压力、两向单位摩擦力在变形区内的分布以及前、后张力横向分布的计算结果与实验结果吻合较好,证明用本方法模拟带轧材轧制过程是可行的。与三维差分法、三维有限元法及三维边界元法相比,本方法具有计算简便、经济和实用的特点。     

计算应力强度因子的无网格-直接位移法

目前计算裂纹尖端应力强度因子的无网格法一般均采用。积分方法,但由于该方法为间接求解,降低了求解精度与求解效率。文中采用无网格—伽辽金方法,选取带有扩展基的奇异基函数,以精确计算裂纹尖端位移场,并借鉴有限元法中计算应力强度因子的直接位移法,提出一种计算含裂结构裂纹尖端应力强度因子的新方法,即无网格—直接位移法。数值计算结果表明,该方法具有简捷、高效的特点,可以准确计算裂纹尖端应力强度因子。     

疲劳分析中变幅载荷的循环计数方法

将压力容器承受的实际变幅载荷历史分解为恒幅块载荷谱的方法称为循环计数方法。介绍了ASME规范推荐的雨流计数法、最大-最小计数法和欧盟标准的蓄水池法。     

An adaptive stochastic resonance method based on grey wolf optimizer algorithm and its application to machinery fault diagnosis

Stochastic resonance (SR) is widely used as an enhanced signal detection method in machinery fault diagnosis. However, the system parameters have significant effects on the output results, which makes it difficult for SR method to achieve satisfactory analysis results. To solve this problem and improve the performance of SR method, this paper proposes an adaptive SR method based on grey wolf optimizer (GWO) algorithm for machinery fault diagnosis. Firstly, the SR system parameters are optimized by the GWO algorithm using a redefined signal-to-noise ratio (SNR) as optimization objective function. Then, the optimal SR output matching the input signal can be adaptively obtained using the optimized parameters. The proposed method is validated on a simulated signal detection and a rolling element bearing test bench, and then applied to the gear fault diagnosis of electric locomotive. Compared with the conventional fixed-parameter SR method, the adaptive SR method based on genetic algorithm (GA-SR) as well as the well-known fast kurtogram method, the proposed method can achieve a greater accuracy. The results indicated that the proposed method has great practical values in engineering.     

变胞机构理论与应用研究概述

阐述了变胞机构的特点及其结构学、运动学、动力学、控制、柔性建模及设计等若干理论问题。介绍了矢量法、矩阵法、指数积方法等几种运动分析方法,比较了Newton/Euler法、Lagrange方法、Kane方法、Roberson/W ittenburg方法、高斯最小约束原理法等几种动力学建模方法的优缺点,提出了变胞机构的几种控制方法和柔性变胞机构建模中存在的几个关键性的问题。最后概述了变胞机构的应用领域,展望了变胞机构的应用前景。     

异常载荷剔除中幅值与梯度门限的联合应用

幅值门限法对于大误差及超数量级的异常数据具有良好的识别性,而梯度门限法对于小幅值的奇异点检测能力强。针对这一特点,分析比较了幅值门限法与梯度门限法在剔除异常载荷数据过程中的优劣特性,提出联合应用这两种方式来实现对奇异信号的处理,并以装载机传动系为例给出了异常载荷数据剔除的过程及结果,避免了基于假定样本数据服从某种典型概率分布,利用标准差剔除异常数据的方法存在的固有局限性,可以快速、准确地剔除不同概率分布下的样本异常数据。