分形——数学与后数学 因此,我们需要做的就是找出究竟是哪些数学被用来创造了宇宙,那么我们将能够了解我们是怎么来的,我们又将去往何方。因为我们正试图辨别环境的模式,特别是当它们涉及到生物圈时,我们需要发现自然用来将物理结构放入空间的数学。 这样的任务需要使用几何学,因为根据定义,这一数学分支专门关注空间中的结构的特性,量度和关系。几何学对于宇宙的组织而言具有如此的根本性,以至于在伽利略的觉悟之前很久,柏拉图就认为,“几何学在创世之前就已存在。” 直到 1975 年,普通公众仍然只熟悉欧几里德的几何学原理,它们总结在那 13 卷大约成书于公元前 300 年左右的古老希腊文本,《 欧几里德原理 》中。这就是我们大多数人在学校里学到的几何学,我们用它在绘图纸上画出各种结构,如立方体、球体、锥体等等。欧几里德几何使我们得以预测天体的运动,建造宏伟的建筑和园林,甚至建造各种飞船和尖端武器。 然而,欧几里德几何的数学公式并不能马上用于自然界中。比如说,使用标准欧几里德几何的完美形状,你能创造出一棵怎样的树呢?回想一下你在幼儿园画过的树,一个圆圈坐在一根细长的矩形顶上。毫无疑问,你的幼儿园老师承认它是树的某种表达,但是它无论如何也描述不出树到底是什么,就像用火柴棒摆出的小人描述不了真正的人一样。 在欧几里德几何和圆规的帮助下,你可以画出一个完美的圆。但你没法用欧几里德几何去画一棵完美的,或者至少是一棵真实的树。欧几里德几何也同样无法画出像甲虫、山、云、或者其他任何自然界中那些常见形体的结构。当需要描绘生命的结构时,欧几里德几何就相形见绌了。那么,我们到哪里去找柏拉图和伽利略所说的那种数学,那种可以描述自然界固有的设计原理的数学呢? 大约 90 年前,一位年轻的法国数学家加斯顿·朱利叶发表了一篇论文,报告了他关于迭代函数的研究工作。这篇论文为我们提供了一个线索。他所用的是一个相对简单的公式,只使用乘法和加法,无限地重复下去。要实际地把他的数学公式所编码的图象可视化,朱利叶将不得不解出该公式上百万次的迭代结果,这个过程会花掉他几十年的时间。因此,尽管朱利叶在数学意义上已经构想出了一个分形,但他实际上从来也没有看到过。 只有到了 1975 年,当朱利叶的公式在计算机的帮助下求出结果之后,其深远的意义才得以显现。法裔美籍数学家波努瓦·芒德勃罗在 IBM 计算实验室中分析了混沌系统的模式,他第一个观察到了这种朱利叶只能想象的东西。面对着由分形公式所产生的具有惊人的美丽、充满生机、并且无限复杂的图像,芒德勃罗充满了敬畏。他第一个观察到,分形图像具有重复的自相似模式,无论在何种尺度下研究时均是如此。他越是放大图像,这些结构看起来就越相同。 内在于分形图像的混沌复杂性之中的,是不断重复、相互嵌套的模式。那种国际流行玩具,手绘俄罗斯嵌套娃娃,为分形的重复图像本质提供了一个粗略的观念。每个更小版本的娃娃都与它外边嵌套的那个较大的娃娃相似,但并不一定完全相同。芒德勃罗引入了 自相似 这一名词来描述他在这种新的数学当中所观察到的对象,他把这种数学称为 分形几何 。 图 11-2. 俄罗斯嵌套娃娃代表了分形的重复图像。 芒德勃罗在他分形图像的复杂性中,看到了各种类似于自然界中常见形状的生动模式,如昆虫,贝壳和树木。在历史上,科学多次记录了在自然界结构中的不同尺度上出现的自相似组织模式。然而,在芒德勃罗引入分形几何学之前,这些自相似的模式都被视为仅仅是奇妙的巧合。 分形几何学强调的是整体结构中的模式与其各部分中的模式之间的关系。回想一下前文所说的关于海岸线的例子和关于枝叶、树枝和树干的例子。自相似的模式在自然界中随处可见,特别多见于人体的结构中。例如,在人的肺脏中,气道沿大支气管分支的模式在小的支气管,甚至更小的细支气管的气道分支模式中不断地重复。循环系统中的动脉和静脉血管,以及人体的周围神经网络也都显示重复的,自相似的分支模式。 由于分形几何是真正的自然界设计原理,生物圈本身在其组织的每个层面都显示出相互嵌套的自相似模式。因此,当我们观察并发现了一个组织在较高或者较低水平上的结构模式时,我们就可以像使用地图一样地使用分形原理。分形可以帮助我们洞悉该组织在任何其他水平上的模式。在生物圈中,人类进化的分形模式可以内在地显示出某种与自然界的组织在其他水平上的结构所经历的进化自相似的模式。 恩斯特·海克尔是与达尔文同时代的著名胚胎学家, 1868 年,他在不经意间首次报道了进化过程中自相似分形模式的端倪。海克尔出版了一套现在已经闻名天下的显微图像,它们比较了若干物种和人类的胚胎发育阶段。他指出,所有脊椎动物胚胎,包括人类胚胎在内,都通过了一系列类似的结构阶段。海克尔提出,各种有机体在通过他们的早期发育阶段时,实际上重新追踪了它们的祖先进化的每一个阶段。 海克尔的理论,隐晦地定义为 个体发育重演系统发育 ,其字面意思是“发育是某种对祖先的重演。”不幸的是,这个狂热的海克尔在推广他的想法时,篡改了他的图片,使胚胎的早期阶段看上去比它们实际上更为相似。 尽管他的报告有瑕疵,但人类胚胎在最终获得人形之前的确发生了一系列形变。在这些转变当中,人类的胚胎采取了一系列有序的自相似结构模式,在其中它很像脊椎动物进化早期阶段的那些胚胎。 发育中的人类胚胎形状从一个酷似鱼类的胚胎变形为类似两栖动物的胚胎。然后它继续变形,采纳了爬行动物的胚胎外观,然后是哺乳动物的胚胎外观,最后才获得了人形。通过沿袭其生物圈祖先的胚胎阶段演变,人类胚胎为分形性自相似提供了一个动态实例。 Fractals—Math and Aftermath Consequently, all we need to do is findout which mathematics was used to create the Universe and we will be able tounderstand how we got here and where we are bound. Because we are trying todiscern environmental patterns, specifically as they relate to the biosphere,we need to discover the math Nature used to put physical structure into space. Such a mission invokes the use of geometry because, by definition, this branch of mathematicsis specifically concerned with the properties, measurement and relationships ofstructure in space. Geometry is so fundamental to the organization of the Universethat long before Galileo’s realization, Plato concluded, “Geometry existedbefore creation.” 4 Until 1975, the general public was onlyfamiliar with the principles of Euclidean geometry, summarized in thethirteen-volume ancient Greek text, The Elementsof Euclid, written around 300 b.c.e. This is the geometrymost of us learned in school to plot structures such ascubes and spheres and cones onto graph paper. Euclidian geometry has enabled us to projectthe movement of heavenly bodies, construct great edifices and gardens, and evenbuild spaceships and sophisticated weapons. However, the mathematical formulae of Euclidiangeometry are not readily applicable to Nature. For example, what kind of tree can you createusing the standardized perfect forms of Euclidean geometry? Think back to thetree you drew in kindergarten, a circle sitting atop an elongated rectangle.Your kindergarten teacher, no doubt, recognized it as a representation of atree, but in no way does it describe what a tree really is, no more than astick figure describes a human. WithEuclidean geometry and a compass, you can draw a perfect circle. But you cannotuse Euclidean geometry to draw a perfect or, at least, a realistic tree. Norcan Euclidian geometry describe the structure of a beetle, a mountain, a cloud,or any other familiar patterns found in Nature. Euclidean geometry falls shortwhen it comes to describing the structure of life. So where do we find the typeof mathematics referred to by Plato and Galileo, the math that describes thedesign principles inherent in Nature? We wereoffered a clue about 90 years ago when a young French mathematician namedGaston Julia published a paper on his work with iterated functions. His was a relatively simple equation that used only multiplicationand addition, repeated ad infinitum . To actually visualize the imageencoded in his mathematical formula, Julia would have had to solve millions ofiterations of the formula, a process that would have taken him decades. Therefore,even though he conceived of a fractal in mathematical terms, Julia neveractually saw one. Theprofound implications of Julia’s formula were only revealed when his equationwas solved with the aid of computers in 1975. Benoit Mandelbrot, a French–Americanmathematician who analyzed patterns in chaotic systems at an IBM computing lab,was the first person to observe what Julia could only imagine. Mandelbrot wasawestruck by the strikingly beautiful organic and infinitely complex imagesgenerated by fractal formulae. He was the first toobserve that fractal images possessed repeated self-similar patterns,regardless of the scale on which they were examined. The more he magnified theimages, the more the structure appeared to be the same. Inherent within the chaotic complexity offractal images is the presence of ever-repeating patterns, nested within oneanother. The internationally popular toy, hand-painted Russian nesting dolls,provides a rough idea of the nature of a fractal’s repetitive images. Eachsmaller version of the doll is similar to, but not necessarily an exact versionof, the larger doll in which it is nested. Mandelbrot introduced the term self-similar to describe such objects that heobserved in the new math, which he called fractalgeometry. Withinthe complexity of his fractal images, Mandelbrot observed vivid patterns that resembleshapes common in Nature, such as insects, seashells and trees. Historically,science had frequently documented the presence of self-similar organizationalpatterns at different scales of Nature’s structure. However, until Mandelbrotintroduced fractal geometry, these self-similar patterns were deemed to bemerely curious coincidences. Fractal geometry emphasizes therelationship between the patterns in a whole structure and the patterns seen inits parts. Recall the examples of the coastline and of the twigs, branches andtree trunks cited earlier. Self-similar patterns are found throughout Natureand especially within the structure of the human body. For example in the humanlung, the pattern of branching along the large bronchus air passages isrepeated in the branching structure of the smaller bronchi and even smallerbronchiole passages. Arterial and venous vessels of the circulatory system aswell as the body’s network of peripheral nerves also display repetitive,self-similar branching patterns. Because fractal geometry is truly thedesign principle of Nature, the biosphere inherently reveals nestedself-similar patterns at every level of its organization. Consequently, as weobserve and become aware of patterns at higher or lower levels of anorganization’s structure, we can use fractals in the same way we would usemaps. Fractals can help us gain insight into the organization at any otherlevel. In the biosphere, the fractal pattern of human evolution can inherentlydisplay a self-similar pattern of evolution experienced by structures at otherlevels of Nature’s organization. Ernst Haeckel, a famous embryologist andcontemporary of Darwin, inadvertently reported the first inkling of aself-similar fractal-like pattern in evolution in 1868. Haeckel published a nowfamous sequence of microscopic images that compares the stages of embryonicdevelopment of a number of species with that of the human. He noted that allvertebrate embryos, including the human embryo, pass through a series ofsimilar structural stages. Haeckel argued that, in transitioning through theirearly development, organisms actually re-trace every stage of theirevolutionary ancestry. Haeckel’s theory, cryptically defined as ontogeny recapitulates phylogeny, literallymeans “development is a replay of ancestry.” Unfortunately, when promoting hisideas, an overzealous Haeckel fudged his drawings to make the early stages ofembryos appear more alike than they actually are. Regardlessof his flawed presentation, human embryos do morph through a variety of shapesbefore acquiring human form. In these transitions, the human embryo assumes asequential series of self-similar structural patterns wherein it resembles embryos from earlier stages of vertebrate evolution. Thedeveloping human embryo shape-shifts from one that resembles a fish embryo toone that resembles an amphibian embryo. It continues morphing until it takes onthe appearance of a reptilian embryo and, later, that of a mammal beforefinally assuming a human shape. Evolving through the embryonic stages of its biosphericancestors, human embryos offer a dynamic example of fractal-likeself-similarity.
控制树状分形网络上的运输效率 吴斌 章忠志 摘要 :陷阱问题是众多其它动力学过程中的一个基本机制,有效地控制陷阱过程(尤其是陷阱效率或运输效率)是复杂系统上陷阱问题研究的一个中心课题。因此,研究陷阱问题的控制方法具有重要的理论意义与实际价值。本文提出了一类有向分形网络,研究了该类网络上的陷阱问题,集中研究了陷阱点固定在中心节点这一特殊情形。所提出的有向分形网络可以从之前的无向分形网络按如下方式扩展得到:将原来无向网络的每边条看作具有不同边权的两条有向边,每条有向边的权值通过单个参数控制。根据该有向分形网络的自相似结构,利用重正化群技术,得到了与陷阱过程有关矩阵的所有特征值及其重数,其中特征值是通过一个精确的递推关系式给出的。通过所得的关于特征值的递推关系,计算了最小特征值和平均陷阱时间( ATT )的表达式。这里的 ATT 是指游走者首次到达陷阱点的期望时间,它是衡量陷阱效率的一个主要指标,近似等于最小特征值的倒数。结果表明: ATT 行为完全由权参数控制:通过调节边权参数, ATT 可以是系统规模的亚线性、线性、或超线性函数。本项研究为控制分形网络上的运输效率提供了一种的有效方法。 相关结果已在 The Journal of Chemical Physics 上正式发表。 文章发表的 PDF 版本: Controlling the efficiency of trapping in treelike fractals.pdf
分形和自相似性是自然界中的普遍现象,近年来,一些学者先后在短信通信、股票交易和人体的生理活动上发现了人类行为的分形特征,我们尝试从时间序列和复杂网络的角度挖掘图书借阅行为中十分存在分形特征。文章前不久被Physica A接受,详见附件。 Fractal analysis on human dynamics of library loans Chao Fan, Jin-Li Guo, Yi-Long Zha Physica A Volume 391, Issue 24, 15 December 2012, Pages 6617–6625 Abstract In this paper, the fractal characteristic of human behaviors is investigated from the perspective of time series constructed with the amount of library loans. The values of the Hurst exponent and length of non-periodic cycle calculated through rescaled range analysis indicate that the time series of human behaviors and their sub-series are fractal with self-similarity and long-range dependence. Then the time series are converted into complex networks by the visibility algorithm. The topological properties of the networks such as scale-free property and small-world effect imply that there is a close relationship among the numbers of repetitious behaviors performed by people during certain periods of time. Our work implies that there is intrinsic regularity in the human collective repetitious behaviors. The conclusions may be helpful to develop some new approaches to investigate the fractal feature and mechanism of human dynamics, and provide some references for the management and forecast of human collective behaviors. Keywords Human dynamics; Time series analysis; Long-range dependence; Complex network; Visibility graph 文章PDF: PHYSA_13934_proof.pdf http://www.sciencedirect.com/science/article/pii/S0378437112006231 我们采用的数据是两所图书馆的借阅量,以及借阅的间隔时间。用重标极差法计算了以借阅量为观测值构成的时间序列的Hurst指数和非周期循环长度,发现人类行为具有长期正相关性和持续性,记忆效应对借阅行为有强烈影响,并与时间标度有关。群体用户的分形特征表现较为明显,而个体用户的时间序列中则有一定的波动性;并且不同的用户群之间,以及同一个数据集中的不同用户之间表现出了显著的个体差异。 通过可视算法将人类行为的时间序列和复杂网络结合在一起,计算了由时间序列转化得到的复杂网络的拓扑参数,发现群体用户的网络具有无标度特征、小世界效应和等级结构,而个体用户的网络则只具有以上部分性质。可以认为,人类的重复性行为发生的时间序列中各个观测值之间存在潜在的密切联系,特别是对于日常生活中的某些重要时刻。我们还发现只有部分的个体行为网络具有分形结构和自相似的特征。此外,本文的分析也对于找寻时间序列和复杂网络之间的关系、网络属性之间的关系以及网络分形结构的起源具有一定的借鉴意义。 注:中文内容中部分结论是笔者硕士论文中的一部分,没有写进这篇英文版本中,也欢迎同行批评指正! 《从图书借阅看人类群体和个体行为的动力学机制》,樊超,上海理工大学,2011年。
李超勇博友在《 想象力是有限的,人脑可能根本不能认识mind! 》中回复: 尝试一下:用可列无穷多个点,看能不能填满一个线段? 很受启发,我原来认为脑的状态是有限的,实际上是用现在的计算机来类比,把脑的状态离散化了。按照彭罗斯的说法,现实的物体很多是不可计算的,不能把脑用现在的计算机来类比或者仿真。在有限范围的相空间中,实际的脑的状态可能是无限的。这里提到了有限中的无限,需要仔细解释一下。在一个有限空间范围内,比如一个圆内部,面积是有限的,但里面的点,可以有的坐标位置是无限的。 在宇宙天体学(cosmology)里,普遍地认为我们所处的空间是有限无界的,而且目前空间仍然在不断膨胀扩大,各个星系间互相远离。通常用一个气球来解释,用气球表面表示现有的三维空间,气球二维表面积是有限的,现有的三维空间也是有限的。各个星球是气球上的点。把气球吹大,表面不断膨胀,表面积是不断增大的,上面的点之间的距离也自然不断增大。而整个气球表面是没有明确的边界的,这就是有限无界。宇宙空间就是这样的,小时候经常想太阳的外面、银河的外面,外面的外面又是什么,用有限无界就可以解释了。mind状态所处的相空间,也可以看做是有限无界的,所以整个空间还是有限的,人的想象力虽然可能有无限的状态,但所能想象之“物”,触角所能延伸的范围仍然可能是有限的。所以,思想没多远。 另外,还有个有趣的东西,分形(Fractal),也是有限中蕴含无限。最开始是测量英国海岸线长度发现的,用不同尺子去量,得到的长度是不一样的!尺子无限小,长度就会无限大!而英国是在地球上的,范围是有限的。另外,分形描述了自然界的一些复杂物体,可能只遵循简单的规律,就可以演化出来。典型的就是Mandelbrot set,在一定的面积下,可以有无限复杂的形态。你可以无限放大,里面总是有特异性的结构! Mandelbrot set,是我用Qt的例子程序生成的。 -------------------- 今天的Nature出了一篇文章,和我的看法不谋而合!认为人的认知能力是有限的,那些我们不知道不知道的东西,才是最急迫需要开拓的。我们应该focus到寻找问题上。下面这段话太经典了: “There are known knowns; there are things we know we know. We also know there are known unknowns; that is to say, we know there are some things we do not know. But there are also unknown unknowns, the ones we don't know we don't know,” Shermer, Michael. “Philosophy: What We Don’t Know.” Nature 484, no. 7395 (April 25, 2012): 446–447. ------------------------ 参考资料: 1. Stephen Hawking,The Grand Design,Bantam, 2011 2. Barry Masters, Physics and Biology: Fractals and the Human Retinal Blood Vessels, oral presentation, 2012 3.罗杰 彭罗斯 (许明贤 吴忠超 ),皇帝新脑,湖南科学技术出版社,2010(第二版) 4. BBC系列:神秘的混沌理论 这个视频给了很多例子,试图揭示分形的深层次含义,暗示了,光怪陆离、形态各异的花姿世界,包括人等形形色色的生命,可能都只遵循简单的规则,就可以演化出来了。
研究了一类树状规则分形上带有单个陷阱点的随机游走问题,其中陷阱固定在中心节点上。得到了这类分形上陷阱问题对应的随机主方程的全部特征值及其重数,其中特征值通过一个显示的递推关系式给出。此外,给出了最小特征值的近似解,并指出它的倒数与平均陷阱时间近似相等。所提出的计算网络特征值及其重数的方法还适用于其它树状规则分形。 相关结果已经被《 EPL (Europhysics Letters) 》正式录用,拟于近期发表。 国际专家的评论: I read with much interest the submitted manuscript, whose authors succeeded in determining exactly the spectra of an important family of tree-like fractals. This is a very significant achievement........ The work is written in a very clear and concise manner and should be readily understood by specialists and non-specialists alike. I hence recommend publication of the manuscript inEPL. 文章发表的 PDF 版本: Complete spectrum of stochastic master equation for random walks on treelike fractals.pdf
汉诺塔问题是一个古老的“游戏”,在每本计算机程序设计教课书里,几乎都把求解汉诺塔问题作为递归算法的范例。经典的汉诺塔问题可以描述如下:有三根柱子与 n 个大小不一的盘子,初始时,这 n 个盘子从大到小叠放在第一根柱子上,并且小盘子位于大盘子上面。问题是如何把这 n 个盘子从第一个柱子全部移动到第三个柱子上,移动时满足这样的规则:每一次只能移动一个盘子,并且满足小盘子只能在大盘子上面。就这一问题本身而言,无论是最佳的移动方法还是最少的移动步数,都已成功解决。 从经典的汉诺塔问题可以拓展出许多其它的版本。例如,我们最近提出了一个扩展的汉诺塔问题,即在上述移动规则下,并不按照最优的方法移动盘子,而是对其进行随机移动。我们的问题是:从初始状态(所有盘子都在第一个柱子上)出发,按照移动规则,随机地移动盘子,请问:所有盘子恰好首次均在第三个柱子上时,期望移动盘子的次数是多少?这一问题当初由我本人提出来,让来自台湾参加大陆 ACM 总决赛的同学思考,问题的答案最终由我的硕士生伍顺琪在我与陈关荣老师的共同指导下得到了圆满解决。我们将所提出的问题归结为求解对偶 Sierpinski 分形上随机游走的平均首达时间,相关结果已经被《 The European Physical Journal B 》正式录用,以下是论文的中文摘要。 摘要: 本文研究了 d 维对偶 Sierpinski 分形( Dual Sierpinski gaskets, DSGs )上的随机游走问题。根据电阻距离与随机游走平均首达时间的关系,首先计算了 d 维 DSGs 上两个特殊点之间的平均首达时间,然后利用 DSGs 拉普拉斯矩阵的谱,计算了 DSGs 中所有节点对之间的平均首达时间。通过递归的方法,得到了上述两个问题的精确解,并给出了它们与网络规模大小的变化关系。最后,给出了 d=2 时所得 DSGs 上随机游走的研究结果与扩展的汉诺塔问题的对应关系。 发表的论文PDF版本: Random walks on dual Sierpinski gaskets.pdf
我做了一个梦,梦见春秋时代一个国家 yan 国,我要就 yan 国写一篇文章,可这 yan 字不会写,我看书上 yan 是这么写的:撬。看起来是撬棍的撬。可再仔细看,发现撬字中的每个“毛”仍然由三个小的“毛”组成,用放大镜看,这个小“毛”由更小的“毛”组成,以此类推,以至无穷。旁边还有一个画外音说道:这是一个典型的分形结构,这个字本质上无法书写。 三个小毛如何组成一个大毛,着实令人费解。在梦境里,这一切都是合理的。梦幻世界必有一套自己的逻辑。那么,存在一个没有悖论的世界吗?要想消除悖论,得用一种什么样的逻辑呢?
最近在研究复杂网络的分形和自相似问题时遇到了很多困惑,在这里把思路整理一下。 基本理论 分形理论首先是一门数学,但由于可作为描述系统科学中很多问题的强有力工具因而被视为一种重要的系统理论。传统的几何学只研究规则齐整的形状,即整形。但是现实世界中存在大量不规则、不整齐的琐碎形状,因而简单性科学是无法描述他们的,这样的复杂几何现象引起了人们的注意并由此诞生了分形几何学,分形理论逐步发展成熟。 大自然中存在着大量的分形现象,我们称之为自然分形。一个典型问题即为 Mandelbrot 提出的英国海岸线有多长?由于海岸线是由大大小小的曲折嵌套而成的,所以不同的测度单位会带来计算结果的巨大差异。同样,山川河流也都具有分形特征,主脉分出支脉,大支脉嵌套小支脉(或支流)。山的表面既不是平面也非光滑曲面,同样,水的表面也不是绝对平面。 数学家用数学的方法造出的分形则称为数学分形,比如对某个规则整形按照一定的规则进行变换,以产生更多更深层次的细节,使得图形越来越纷繁、琐碎、复杂。典型的例子有康托尘埃、科赫曲线、谢尔宾斯基垫子、谢尔宾斯基海绵等等。这样的生成规则也不一定是完全确定的,可以加入一定的随机因子,按照概率使用某些规则,可以生成更复杂同时更接近自然分形的图案来。 分形与自相似 分形至今没有一个严格的定义,常用通俗的描述来解释分形。一般认为,分形具有不规整性、层次嵌套性和自相似性。 按照 Mandelbrot 的定义, fractal is a rough or fragmented geometric shape that can be split into parts, each of which is (at least approximately) a reduced-size copy of the whole (Mandelbrot, B.B. 1982) 。这样的定义就默认了分形特征是一种 a property called self-similarity ,也就是说分形中包括了自相似。 所谓自相似,是一种尺度变换下的不变性 (scale-invariance) ,即在不同尺度下观察分形可以看到近似相同的形象,若把整个对象的局部放大,再把局部的局部放大,都可以看到相似的结构特征。但是这种自相似并不像整形的相似那么严格,允许相似中的不相似,不需要也不可能完全相同。比如,科赫曲线,整体是闭合的,但任一部分都不是封闭曲线。分形自相似意味着部分与整体有一样的复杂性:一样曲折、琐碎、纷乱、不规整、不光滑。并且,分形的部分与部分之间也是相似的。山重水复疑无路就是从审美的角度对山水分形中的自相似的描述,以至于让外人只看到相同之处而难以了解细微的差别,便生出迷路的疑惑。 整数维与分数维 我们都知道传统几何中点、线、面、体分别是 0 、 1 、 2 、 3 维的,这里的维数都是非负整数,故称为整数维或者拓扑维。但是分形几何对象的独特属性是不能用整数维来描述的,特别是其不规则性和复杂性,如科赫曲线在性质上不同于一维曲线但也远非二维的面。因此, Mandelbrot 引入分数维来刻画分形对象的不规则程度和复杂性程度。设 为b几何对象, a为单位线段,令 D 为分数维,定义 ,则 D=logb/loga 。 分形时间序列 查到的关于分形时间序列的文献大多是金融时间序列,这是已被公认为分形布朗运动的一种时间序列。分形布朗运动是统计自相似的,具有长期记忆性的,也就是说有一种记忆效应使得未来的变化趋势与现在相同。这种长期相关性可由相关指数 Hurst 指数表征,当H=0.5 时,序列是完全随机的;当H0.5 时,序列具有长期相关性,未来的发展趋势倾向于和过去相同;而当H0.5 时,序列是反相关的,未来的发展趋势倾向于与过去相反。分形维数可由 Hurst 指数求出,定义: 时间尺度分形维:Df=2-H. 概率空间分形维:Dp=1/H.
分形之父Mandelbrot所得部分奖励和荣誉 A partial list of awards received by Mandelbrot 2004 Best Business Book of the Year Award AMS Einstein Lectureship Barnard Medal Caltech Service Casimir Funk Natural Sciences Award Charles Proteus Steinmetz Medal Franklin Medal Harvey Prize Honda Prize Humboldt Preis Fellow, American Geophysical Union IBM Fellowship Japan Prize John Scott Award Lewis Fry Richardson Medal Medaglia della Presidenza della Repubblica Italiana Mdaille de Vermeil de la Ville de Paris Nevada Prize Science for Art Sven Berggren-Priset W?adys?aw Orlicz Prize Wolf Foundation Prize for Physics
据美国新闻媒体报道,美国东部时间10月14日法裔美国数学家Benoit Mandelbrot 85岁(20 November 1924 14 October 2010)在马塞诸塞州剑桥市临终医院因胰腺癌辞世。Mandelbrot是分形之父, 1982年他出版了著名的自然的分形几何著作,标志着分形几何的诞生。他在分形方面的工作成为混沌理论的基础,也是计算机数据压缩和医学图像纹理以及模拟 湍流对飞机机翼造型设计的关键。Benoit Mandelbrot 出生于波兰(父母是犹太人),孩童时代移居法国,他大部分时间在美国生活和工作,他具有法国和美国双重国籍。 1958-1987年 Mandelbrot 一直在IBM工作,1987-2005在耶鲁大学工作,2005年退休后一直生活在麻州的剑桥市。 Benoit Mandelbrot, a mathematics pioneer and the father of the principle of fractal geometry, has died in the US at the age of 85. The fractal principle uses mathematical fromulas to attempt to understand complexity of natural world In his seminal 1982 work The Fractal Geometry of Nature, Mandelbrot argued that seemingly random patterns could in fact be the same infinitely repeated shape. He once used a cauliflower to describe the mathematical principle, pointing out that the shape of the vegetable was repeated over and over The mathematical principle has been used to measure shapes previously thought unmeasurable, including coastlines and mountains. Mandelbrot also applied the concept to economics, but he was critical of the global financial system, believing it to be too complex to properly function. Fractal geometry can be depicted in intricate and colourful computer designs which have become popular as artworks in their own right. One fractal variation was even named after Mandelbrot. The Mandelbrot Set has had a huge influence on mathematics and culture - examples have even been known to appear as crop formations. Mandelbrot的早年生活 Early years Mandelbrot was born in Warsaw into a Jewish family from Lithuania .He was born into a family with a strong academic traditionhis mother was a medical doctor and he was introduced to mathematics by two uncles, one of whom, Szolem Mandelbrojt , was a Parisian mathematician. However, his father made his living trading clothing. Anticipating the threat posed by Nazi Germany , the family fled from Poland to France in 1936 when he was 11. Mandelbrot attended the Lyce Rolin in Paris until the start of World War II , when his family moved to Tulle . He was helped by Rabbi David Feuerwerker , the Rabbi of Brive-la-Gaillarde , to continue his studies. In 1944 he returned to Paris. He studied at the Lyce du Parc in Lyon and in 1945-47 attended the cole Polytechnique , where he studied under Gaston Julia and Paul Lvy . From 1947 to 1949 he studied at California Institute of Technology , where he earned a master's degree in aeronautics .Returning to France, he obtained a PhD in Mathematical Sciences at the University of Paris in 1952. From 1949 to 1958 Mandelbrot was a staff member at the Centre National de la Recherche Scientifique . During this time he spent a year at the Institute for Advanced Study in Princeton, New Jersey , where he was sponsored by John von Neumann . In 1955 he married Aliette Kagan and moved to Geneva, Switzerland , and later to the Universit Lille Nord de France . In 1958 the couple moved to the United States where Mandelbrot joined the research staff at the IBM Thomas J. Watson Research Center in Yorktown Heights, New York . He remained at IBM for thirty-two years, becoming an IBM Fellow , and later Fellow Emeritus . 学术生涯 Academic career From 1951 onward, Mandelbrot worked on problems and published papers not only in mathematics but in applied fields such as information theory , economics , and fluid dynamics . He became convinced that two key themes, fat tails and self-similar structure, ran through a multitude of problems encountered in those fields. Mandelbrot found that price changes in financial markets did not follow a Gaussian distribution , but rather Lvy stable distributions having theoretically infinite variance . He found, for example, that cotton prices followed a Lvy stable distribution with parameter equal to 1.7 rather than 2 as in a Gaussian distribution. Stable distributions have the property that the sum of many instances of a random variable follows the same distribution but with a larger scale parameter . Mandelbrot also put his ideas to work in cosmology . He offered in 1974 a new explanation of Olbers' paradox (the dark night sky riddle), demonstrating the consequences of fractal theory as a sufficient, but not necessary, resolution of the paradox. He postulated that if the stars in the universe were fractally distributed (for example, like Cantor dust ), it would not be necessary to rely on the Big Bang theory to explain the paradox. His model would not rule out a Big Bang, but would allow for a dark sky even if the Big Bang had not occurred. In 1975, Mandelbrot coined the term fractal to describe these structures, and published his ideas in Les objets fractals, forme, hasard et dimension (1975; an English translation Fractals: Form, Chance and Dimension was published in 1977). Mandelbrot developed here ideas from the article Deux types fondamentaux de distribution statistique (1938; an English translation Two Basic Types of Statistical Distribution ) of Czech geographer , demographer and statistician Jaromr Kor?k . While on secondment as Visiting Professor of Mathematics at Harvard University in 1979, Mandelbrot began to study fractals called Julia sets that were invariant under certain transformations of the complex plane . Building on previous work by Gaston Julia and Pierre Fatou , Mandelbrot used a computer to plot images of the Julia sets of the formula z ² . While investigating how the topology of these Julia sets depended on the complex parameter he studied the Mandelbrot set fractal that is now named after him. (Note that the Mandelbrot set is now usually defined in terms of the formula z ² + c , so Mandelbrot's early plots in terms of the earlier parameter are leftright mirror images of more recent plots in terms of the parameter c .) In 1982, Mandelbrot expanded and updated his ideas in The Fractal Geometry of Nature . This influential work brought fractals into the mainstream of professional and popular mathematics, as well as silencing critics, who had dismissed fractals as program artifacts . Mandelbrot left IBM in 1987, after 35 years and 12 days, when IBM decided to end pure research in his division. He joined the Department of Mathematics at Yale , and obtained his first tenured post in 1999, at the age of 75. At the time of his retirement in 2005, he was Sterling Professor of Mathematical Sciences. His awards include the Wolf Prize for Physics in 1993, the Lewis Fry Richardson Prize of the European Geophysical Society in 2000, the Japan Prize in 2003, and the Einstein Lectureship of the American Mathematical Society in 2006. The small asteroid 27500 Mandelbrot was named in his honor. In November 1990, he was made a Knight in the French Legion of Honour . In December 2005, Mandelbrot was appointed to the position of Battelle Fellow at the Pacific Northwest National Laboratory . Mandelbrot was promoted to Officer of the Legion of Honour in January 2006. An honorary degree from Johns Hopkins University was bestowed on Mandelbrot in the May 2010 commencement exercises. 分形Fractals and regular roughness Although Mandelbrot coined the term fractal , some of the mathematical objects he presented in The Fractal Geometry of Nature had been described by other mathematicians. Before Mandelbrot, they had been regarded as isolated curiosities with unnatural and non-intuitive properties. Mandelbrot brought these objects together for the first time and turned them into essential tools for the long-stalled effort to extend the scope of science to non-smooth objects in the real world. He highlighted their common properties, such as self-similarity (linear, non-linear, or statistical), scale invariance , and a (usually) non-integer Hausdorff dimension . He also emphasized the use of fractals as realistic and useful models of many rough phenomena in the real world. Natural fractals include the shapes of mountains , coastlines and river basins ; the structures of plants, blood vessels and lungs ; the clustering of galaxies ; and Brownian motion . Fractals are found in human pursuits, such as music , painting , architecture , and stock market prices. Mandelbrot believed that fractals, far from being unnatural, were in many ways more intuitive and natural than the artificially smooth objects of traditional Euclidean geometry : Clouds are not spheres, mountains are not cones, coastlines are not circles, and bark is not smooth, nor does lightning travel in a straight line. Mandelbrot, in his introduction to The Fractal Geometry of Nature Mandelbrot has been called a visionary and a maverick. His informal and passionate style of writing and his emphasis on visual and geometric intuition (supported by the inclusion of numerous illustrations) made The Fractal Geometry of Nature accessible to non-specialists. The book sparked widespread popular interest in fractals and contributed to chaos theory and other fields of science and mathematics. When visiting the Museu de la Cincia de Barcelona in 1988, he told its director that the painting The Face of War had given him the intuition about the transcendence of the fractal geometry when making intelligible the omnipresent similitude in the forms of nature. He also said that, fractally, Gaud was superior to Van der Rohe . Death Mandelbrot died in a hospice in Cambridge, Massachusetts , on 14 October 2010 from pancreatic cancer , at the age of 85. Reacting to news of his death, mathematician Heinz-Otto Peitgen said if we talk about impact inside mathematics, and applications in the sciences, he is one of the most important figures of the last 50 years. Chris Anderson described Mandelbrot as an icon who changed how we see the world. French President Nicolas Sarkozy said Mandelbrot had a powerful, original mind that never shied away from innovating and shattering preconceived notions. Sarkozy also added, His work, developed entirely outside mainstream research, led to modern information theory.
刚收到 Physical Review E ( PRE ) 编辑部发来的一个好消息:由我们组林苑、吴斌两位本科同学和我一起合作的文章《 Determining mean first-passage time on a class of treelike regular fractals 》已在 PRE 上 正式发表,这是我们组以本科生为第一作者发表在 PRE 上的第一篇论文。 这个工作针对一类树状网络,研究了将陷阱置于某一特殊节点的随机游走时间与全局随机游走时间问题。针对这两个问题,我们分别提出了新的计算方法,该方法计算简单、便捷。为了说明所提出方法的计算过程,我们提出了一类确定性 T 形树,并针对这类树状网络,给出我们所提出方法的计算细节,得到了精确的结果。文章还通过多个例子,说明了所给出的新方法具有普适性。 为了获得计算平均首达时间的简捷方法,林苑和吴斌两位同学付出了大量的时间与精力,他们对这两个问题进行了反复讨论、计算与验算,最终攻克了这些难题。这篇文章的工作量非常大,共有 12 个 PRE 版面(全文见附件)。如今文章在 PRE 上正式发表,这是对两位同学辛勤付出的肯定与鼓励。 本科生以第一作者身份在国际著名期刊上发表文章,说明了在适当的引导与培养之下,本科生是可以做出很出色的工作的。希望我们课题组的本科生同学再接再厉,争取今后做出更加优秀的研究成果。 文章发表的PDF版本