基于B样条曲线节点插入算法的拟合优化

在逆向工程中对叶片进行非接触式测量,为了更好得到叶片型线的数据,需要用曲线对有序的点云进行拟合.为了满足精度要求,提出了用节点插入算法来对拟合出B样条曲线进行优化.该方法首先通过最小二乘法拟合出曲线,然后根据误差和曲率信息以及节点插入算法不断插入节点,所得结果与上一次进行对比,直到达到拟合要求.实验表明,该方法考虑到局部曲率及误差,得到的拟合曲线逼近效果更好.     

测量数据点的高精度B样条曲线拟合算法

为提高曲线重构的效率,提出了关键点提取算法,用于三维测量型值点的B样条曲线逼近.根据离散曲率分析提取具有曲率极值的型值点作为初始关键点,再根据初始关键点和型值点的参数值构建的节点矢量,确保最小二乘矩阵满秩,用最小二乘法反算控制顶点.通过Hausdorff距离衡量逼近曲线与型值点间的逼近偏差,设定偏差阈值和多点调整算法,确定新增关键点的位置区间,根据形状指数分析找到新增关键点的精确位置,通过不断迭代找到满足逼近允差要求的最终关键点和控制顶点.实例验证表明,同一逼近允差前提下,新算法在迭代计破算时间、迭代次数及最终所得控制顶点个数等方面优于其他方法.     

基于压缩控制顶点的B样条曲线逼近技术

为压缩逆向工程中重构物体轮廓的B样条曲线的控制顶点,提出一种基于特征点提取及改进粒子群算法的B样条曲线逼近技术。利用等弦长法计算离散点的曲率,基于曲率分析提取出离散点列的特征点并构造初始逼近曲线。在误差最大处插入新的插值点,构造新的逼近曲线。重复该过程,当增加插值点不能较大幅度地降低逼近误差时,利用改进的粒子群算法优化曲线控制顶点的位置,由此得到最终的B样条逼近曲线。实例验证表明,相比于其他方法,该方法能更有效地降低控制顶点个数,迭代效率高,逼近效果好。     

基于特征点自动识别的B样条曲线逼近技术

提出一种实用的用三次B样条曲线逼近稠密且带噪声的二维断面数据点列的算法。剔除数据点列中的重合点并对其进行均匀弧长重采样处理后,利用相邻点拟合圆弧的方法来近似计算各数据点的离散曲率值,并根据相邻点之间的离散曲率符号变化情况及相近点之间的曲率值和曲率差分关系自动识别出断面数据中绝大多数的特征点(拐点、折痕点、曲率极值点)。构造插值于特征点的B样条曲线,并在逼近误差最大处插入新的插值点。重复这一过程,直到逼近误差小于预先给定值,从而得到最终插值点列并构造相应的B样条曲线。试验结果表明,所构造的曲线节点数目及其分布合理,能够很好地反映原始断面数据点列中的细小特征部分。该算法具有速度快、逼近精度高等特点,可广泛应用于二维断面数据的曲线重建。     

圆柱螺旋线的NURBS逼近表示

针对圆柱螺旋线不能用多项式或有理多项式精确表达的缺陷,提出一种二次非均匀有理B样条逼近算法.通过组合平面圆弧和轴向二次非均匀有理B样条直线构成初始非均匀B样条螺旋曲线;为了减小逼近误差,根据权因子影响非均匀有理B样条曲线形状的特性,提出了修正de Boor细分算法中新控制点轴向坐标的逼近细分算法.实例表明,该逼近算法不仅简单,而且能够稳定地逼近原曲线以满足给定的误差要求,并且为构造非均匀有理B样条螺旋曲面和螺旋体提供了数据.     

基于UG曲线光顺拟合二次开发算法

为了解决涡轮叶片曲面不光顺而造成的数控加工刀轨生成失败的问题,本文针对叶片曲面截面线进行光顺拟合,提出了一种三次非均匀B样条曲线整体光顺逼近算法。该算法建立了一个由最小二乘法、型值点曲率和、型值点曲率变化和三项组成的目标函数,并利用高斯消元法求出最优控制顶点坐标序列。该算法可以使曲线在给定的逼近误差范围内,实现型值点曲率和降低的同时,曲率变化也更加均匀。最后将算法程序嵌入UG5.0二次开发程序中实现自动光顺拟合模块的开发,并通过与已存在的算法进行光顺效果对比实验,验证了本文算法更有效。     

基于遗传算法的单螺杆钻具型线优化设计

以螺杆钻具廓形曲线为研究对象,通过遗传算法对其进行光顺处理。利用三次B样条曲线拟合算法进行数值分析,得到螺杆钻具廓形曲线;通过对B样条曲线的降次处理得到拟合曲线的曲率;检测待光顺的型值点,以样条曲线型值点的变化量作为优化变量,通过制定遗传策略,加快算法收敛速度,完成光顺算法。经过验证:该算法准确有效并符合约束条件。     

一种基于插值拟合的复杂刀具轨迹在线平滑压缩算法

提出了一种基于插值拟合的在线复杂刀具轨迹平滑压缩算法,该算法依据主导点的选取策略,对原始数据点进行离线预处理,然后进行主导点的在线插值拟合以及非主导点的误差检测,进而生成一条满足拟合精度要求的B样条曲线。主导点依据离散数据点的曲率阈值、曲率极大值、曲线拐点,以及分段Bezier曲线逼近拟合后的误差最大值点进行选取。在具有C2连续性的分段Bezier曲线逼近拟合前,需要利用长度均分策略,提取长度突变点作为新增的主导点,以保证拟合的准确性。对主导点进行B样条插值拟合后,利用轮廓误差跟随法对非主导点到拟合曲线的误差进行检测。该方法与牛顿迭代法相比,其计算速度更快且能提高算法效率。仿真结果表明,提出的算法可对复杂刀具轨迹进行平滑压缩,且误差检测的精度能够满足要求。     

散乱数据点云型面特征分析算法的研究与应用

提出一种产品散乱数据点云型面特征快速分析算法.该算法基于R*-tree动态空间存取模型,能快速、准确地获取点云局部型面参考数据,并采用最小面积包围矩形法对其进行参数化;以自由曲面逼近产品局部型面,估算数据点云型面曲率值;将点云型面曲率映射为RGB颜色值,以曲率云图反映产品型面特征分布.该算法运行效率高,产品数据适应性强,型面特征分析结果准确.     

三维自由曲线的相似性比较算法

针对B样条表达的空间自由曲线形状的相似性评价问题,提出了基于曲率特征的相似性评价算法。首先在自由曲线上等弧长均匀取点,并计算各点的曲率,得到曲线的曲率分布;然后以曲率分布作为描述空间曲线的几何特征,采用EMD(Earth Mover’s Distance)算法计算两个曲率分布的距离,衡量对应曲线的相似程度。算法以提取曲率特征的方法将三维空间曲线的相似性比较问题转化为二维分布的距离计算,简单有效地度量空间曲线形状的相似性。为检验算法的效果,以VS2010为集成开发环境对不同类型的B样条空间曲线进行大量的试验,结果表明,提出的相似性比较算法可行有效,且对旋转和取点顺序有较好的鲁棒性,能很好地反映空间曲线的相似程度。     

从人工配送到高速配送中心的转变

最近Liberty Hardware公司更新了它的西海岸配送中心（DC），使其从人工操作变成一个高度自动化的高速配送中心，在这个过程中可以降低40％的劳动时间，并且仅用以往1／3的时间就可以把产品直接交付给客户。     

The isotopic distribution conundrum

Although access to high-resolution mass spectrometry (MS), especially in the field of biomolecular MS, is becoming readily available due to recent advances in MS technology, the accompanied information on isotopic distribution in high-resolution spectra is not used at its full potential, mainly because of lack of knowledge and/or awareness. In this review, we give an insight into the practical problems related to calculating the isotopic distribution for large biomolecules, and present an overview of methods for the calculation of the isotopic distribution. We discuss the key events that triggered the development of various algorithms and explain the rationale of how and why the various isotopic-distribution calculations were performed. The review is focused around the developmental stages as briefly outlined below, starting with the first observation of an isotopic distribution. The observations of Beynon in the field of organic MS that chlorine appeared in a mass spectrum as two variants with odds 3:1 lie at the basis of the first wave of algorithms for the calculation of the isotopic distribution, based on the atomic composition of a molecule. From here on, we explain why more complex biomolecules such as peptides exhibit a highly complex isotope pattern when assayed by MS, and we discuss how combinatorial difficulties complicate the calculation of the isotopic distribution on computers. For this purpose, we highlight three methods, which were introduced in the 1980s. These are the stepwise procedure introduced by Kubinyi, the polynomial expansion from Brownawell and Fillippo, and the multinomial expansion from Yergey. The next development was instigated by Rockwood, who suggested to decompose the isotopic distribution in terms of their nucleon count instead of the exact mass. In this respect, we could claim that the term "aggregated" isotopic distribution is more appropriate. Due to the simplification of the isotopic distribution to its aggregated counterpart, Rockwood was able to use the convolution for the calculation of the "aggregated" isotopic distribution. Convolution methods are computationally efficient and economic in their memory usage. We spend a section on the work introduced by Rockwood during the 1990s. Due to recent breakthroughs in mass spectrometric technology and the widespread high-resolution instruments (e.g., FTICR-MS, FTOrbitrap-MS, and TOF-MS) that provide high-resolution, isotope-resolved, accurate mass data, there is an emerging need for algorithms that can calculate isotopic distributions for large biomolecules. The number of recent publications on this topic does witness this trend. The new methods are mostly based on complex mathematical developments such as, for example, cellular automata (Meija and Caruso [2004]. J Am Soc Mass Spectrom, 15(5):654-658), dynamic programming (Snider [2007]. J Am Soc Mass Spectrom, 18:1511-1515), and hierarchical models (Li et al. [2008] J Am Soc Mass Spectrom, 19:1867-1874). We also comment on the ideas to use Punnet squares and Pascal's triangle to introduce the concept of the isotopic distribution for educational and didactic purposes.     

On the stress distribution in prestressed extrusion dies under non-uniform distribution of internal pressure

Cold extrusion die stress distribution is normally calculated on the assumption that there is a uniform distribution of internal pressure, e.g. by application of the Lamé equations. With FE-analysis this assumption can be overcome by the introduction of arbitrary boundary conditions. However, as little information existed about realistic distribution of radial and axial stresses in container and die this possibility was not used very much. Only after a recent investigation of Bay [1] into friction and pressure distribution in forward extrusion an FE-calculation of stresses in a prestressed extrusion die under non-uniform distribution of radial pressure seemed to be promising. The results show very clearly that no negative effects on the stress state—e.g. increase of stress peaks—may be expected in case of combined continuous decrease of radial pressure in the container and discontinuous increase at the die entry.     

以工位为中心的生产物流配送优化

为解决准时生产配送模式下拣货效率低、配送成本高的问题,提出了以工位为中心的生产物流配送思想,从而优化制造业总装车间生产物流配送.以总配送时间最小为目标函数,建立车间配送优化调度模型,采用遗传算法求解该模型.模拟算例表明,该模型有效且可行,为制造企业生产物流配送优化提供了可参考的模型和算法.     

Analysis of sprocket load distribution

A study was made to determine the load distribution for an elastic roller chain on an elastic sprocket. The equations developed for the distribution include, as a subset, those for a rigid roller chain on a stationary rigid sprocket. The effect of friction and sprocket rotation on load distribution is discussed. Results are presented for a 20-tooth sprocket with a 180° angle of roller chain wrap, for various tight-to-slack side tension ratios of both driver and driven sprockets, with various spring constant ratios.     

制造业分销系统的集成性研究

研究分销系统与企业信息化类软件在功能和数据上的集成,最后提出分销企业综合信息管理平台,并对平台进行详细的介绍。     

基于数字孪生的配送管理系统研究

为提升货物配送的全生命周期管理和配送过程的服务优化,将数字孪生理念融入配送管理,提出和研究了基于数字孪生理念的配送管理系统;分析了系统的应用需求,建立了系统的架构和功能结构,提出了配送过程数字孪生体的构建内容与方法,并设计构建了配送过程的基本数字孪生体、装箱规划的数字孪生体模型和配送过程优化的数字孪生体模型,建立了面向...