科学研究費補助金 研究種目:一般研究(C) 課題番号:026600124 研究代表者:巽 二郎 研究期間:1990-1991年度

Fractal was first used in measuring the length of the coastline, with the fractal research and development, not only to break the traditional Archimedean geometry, but also to explain many scientific to ignore the complexity and nature of nonlinear phenomena structure .Fractal has been widely applied to such as physics, astronomy, geography and sociology and other fields, as a wave of interdisciplinary research in recent years. Coastal areas has always been cultural, economic and activities areas since ancient times. Coastal zone was land and sea for the interaction region by a variety of factors (ex: waves, tides, currents and wind, etc.) continue to function, derived from different coastal terrain. Therefore changes in the coast of the deep impact of humanity. Under the principle of the conservation and development, Coastal areas should be use of modern technology to prediction, analysis, assessment, planning, and management, so that a sustainable preservation of coastal resources. In this study, static and dynamic predict and simulation the coast shape base on fractal. The static part is observation of 29 beaches in South China coast. And collect and calculate the parameters and fractal dimensions of the coast. Through the shape of image processing and analysis of information, to find two generators of the coast. Through the data mining technology to identify the criteria for classification, and to simulation the coastline by generate iterations method. The dynamic part is based on hydraulic model¡¦s results, the use of traditional multiple linear regression and neural network to compare the dynamic prediction of the coastline. The results show that the use of neural networks to predict than the use of multiple linear regression, and effect of use difference angle (£c) to predict sub-coastlines than the effect of not use difference angle (£c) to predict, and add fractal dimension can effectively reduce the predict error and increase the degree of interpretation.

Cette thèse adresse la problématique de la construction des intergiciels hautement adaptables. Ces intergiciels se caractérisent par une grande diversité des fonctionnalités fournies. Dans le domaine du transactionnel, cette diversité concerne non seulement les modèles de transactions, les protocoles de contrôle de concurrence et de reprise après défaillance, mais aussi les normes et les standards d'intégration. Notre proposition consiste à définir un canevas intergiciel capitalisant la diversité du domaine transactionnel, et permettant de construire des services de transactions hautement adaptables. Ce type de services requiert la mise en place d'une démarche de construction à granularité extrêmement fine afin de pouvoir adapter les nombreuses caractéristiques de l'intergiciel.<br /><br />Nous proposons donc de compléter l'approche issue des exogiciels avec quatre nouveaux éléments. Ainsi, nous définissons le modèle de programmation Fraclet à base d'annotations pour favoriser la programmation des abstractions fonctionnelles de l'intergiciel. Nous proposons ensuite un langage de description et de vérification de motifs d'architecture pour fiabiliser la modélisation des abstractions architecturales. Ces deux premiers éléments servent à la conception d'un canevas intergiciel à base de composants utilisant les motifs de conception comme structure architecturale extensible. Enfin, nous décrivons les configurations possibles en utilisant différents modèles de haut niveau dédiés aux caractéristiques de l'intergiciel. Nous illustrons ces concepts en présentant GoTM, un canevas intergiciel à composants pour la construction de services de transactions hautement adaptables.<br /><br />Notre approche est validée au travers de trois expériences originales. Tout d'abord, nous proposons de faciliter l'intégration des services de transactions dans les plates-formes intergicielles par la définition de politiques de démarcation transactionnelle indépendantes de la plate-forme et du type de service intégré. Ensuite, nous définissons un service de transactions composant plusieurs personnalités simultanément pour faciliter l'interopérabilité transactionnelle d'applications hétérogènes. Enfin, nous sommes en mesure de sélectionner différents protocoles de validation à deux phases pour optimiser le temps d'exécution des transactions face aux changements des conditions d'exécution de l'application.

Open and connected fractures, where present, control fluid flow and dominate solute transport. Flow through fractures has major implications for water resource management, underground waste repositories, contaminant remediation, and hydrocarbon exploitation. Complex fracture morphology makes it difficult to quantify and predict flow and transport accurately. The difficulty in usefully describing the complex morphology of a real fracture from a small 3-D volume or 2-D profile sample remains unresolved. Furthermore, even when complex fracture morphology is measured across three-dimensions, accurate prediction of discharge remains difficult. High resolution x-ray computed tomography (HXRCT) data collected for over 20 rock surfaces and fractures provide a useful dataset to study fracture morphology across scales of several orders of magnitude. Samples include fractured rock of varying lithology, including sandstone, volcanic tuffs and crystalline igneous and metamorphic rocks. Results suggest that the influence of grain size on surface roughness is not readily apparent due to other competing variables such as mechanics, skins and coatings, and weathering and erosion. Flow tests of HXRCT-scanned fractures provide real discharge data allowing the hydraulic aperture to be directly measured. Scale-invariant descriptions of surface roughness can produce constrained estimates of aperture variability and possibly yield better predictions of fluid flow through fractures. Often, a distinction is not made between the apparent and true fracture apertures for rough fractures measured on a 2-D topographic grid. I compare a variety of local aperture measurements, including the apparent aperture, two-dimensional circular tangential aperture, and three-dimensional spherical tangential aperture. The mechanical aperture, the arithmetic mean of the apparent local aperture, is always the largest aperture. The other aperture metrics vary in their ranking, but remain similar. Results suggest that it may not be necessary to differentiate between the apparent and true apertures. Rock fracture aperture is the predominant control on permeability, and surface roughness controls fracture aperture. A variety of surface roughness characterizations using statistical and fractal methods are compared. A combination of the root-mean-square roughness and the surface-to-footprint ratio are found to be the most useful descriptors of rock fracture roughness. Mated fracture surfaces are observed to have nearly identical characterizations of fracture surface roughness, suggesting that rock fractures can be sampled by using only one surface, resulting in a significantly easier sampling requirement. For mated fractures that have at least one point in contact, a maximum potential aperture can be constrained by reflecting and translating a single surface. The maximized aperture has a nearly perfect correlation with the RMS roughness of the surface. These results may allow better predictions of fracture permeability thereby providing a better understanding of subsurface fracture flow for applications to contaminant remediation and water and hydrocarbon management. Further research must address upscaling fracture morphology from hand samples to outcrops and characterizing entire fracture networks from samples of single fractures. / text

Este trabajo de Tesis tiene por objetivo general dar soluciones a diferentes problemas relacionados con la detección automática de señales sísmicas. Dichos problemas incluyen el picado de los tiempos de los primeros arribos en exploración sísmica, la detección y determinación de los tiempos de llegada de fases de terremotos en registros de estaciones sismológicas, y la declaración y picado de los arribos de eventos en registros de microsismicidad. Para ello, se estudian diversos atributos sísmicos convencionales y no convencionales diseñados para realzar la llegada de las señales de interés por sobre el ruido de fondo de los datos. A partir de estos atributos se proponen nuevos algoritmos, estrategias y métodos para detectar las señales buscadas de forma eficiente, automática y precisa. En particular, se analiza la dimensión fractal de la traza sísmica, se aclaran algunas suposiciones teóricas e hipótesis que deben realizarse al considerar una traza sísmica como una curva fractal, y se evalúan distintas estrategias para estimar su dimensión. A partir de ejemplos con datos sintéticos y reales, se prueba que la dimensión fractal obtenida a través del método del variograma es un atributo que evidencia claramente la transición entre el ruido y la señal con ruido. Se desarrollan también tres nuevos métodos para detectar los tiempos de los primeros arribos en registros sísmicos complejos de forma automática. Los atributos en los que se basan estos nuevos algoritmos están dados por un cociente de energías, por la entropía de la traza sísmica, y por su dimensión fractal, aprovechando en este último caso los resultados del análisis previo. Los cambios detectados en los atributos son realzados mediante la aplicación de un filtro de suavizado que preserva los saltos bruscos. Estos algoritmos determinan los primeros arribos analizando las trazas sísmicas individualmente. Se introduce luego un novedoso proceso de corrección de picados erróneos que considera simultáneamente todo el conjunto de las trazas pertenecientes a un mismo registro de disparo común teniendo en cuenta la alineación aproximada de los primeros arribos dada por el modelo de refracción. Este procedimiento permite además descartar las trazas malas de forma automática. Los métodos propuestos son luego mejorados y aplicados a un gran volumen de datos reales provistos por la industria que presentan distintos desafíos y dificultades. Para estudiar la detección de señales en sismología, se realiza un análisis comparativo entre ocho métodos V de detección automática de fases de terremotos. El conjunto de ocho métodos analizado está formado por la adaptación de los tres algoritmos desarrollados para picar primeros arribos, por tres métodos tradicionales, y por dos métodos obtenidos a partir de modificaciones a dos de los métodos tradicionales. El estudio es llevado a cabo utilizando registros obtenidos en la Estación Sismológica de La Plata (LPA). Como resultado, se realiza una valoración relativa de los ocho métodos de detección en la que algunos de ellos se destacan por sobre otros, y se obtienen así herramientas computacionales para realizar la detección automática de fases de terremotos en estaciones sismológicas. Por último, se seleccionan tres de estos ocho métodos y a partir de ellos se genera una nueva metodología para detectar microsismos de forma automática que consta de dos pasos. En el primer paso se detectan y determinan todos los tiempos de aquellos eventos que representan un potencial arribo de un microsismo, traza por traza en todo el registro. En el segundo paso, se utiliza un criterio que analiza todo el registro en su conjunto para declarar la presencia de microsismos. Como resultado se obtienen el número de microsismos declarados, un indicador de confianza asociado a cada uno de ellos, y los tiempos de llegada para aquellas trazas en las que el evento es detectado.

El presente trabajo de Tesis aborda el problema de la caracterización del flujo de agua en medios porosos heterogéneos mediante el uso de métodos numéricos y soluciones analíticas. En primer lugar, se propone un experimento numérico que simula las técnicas de laboratorio usualmente empleadas para la obtención experimental de las relaciones constitutivas. Con este fin se resuelve la ecuación de flujo no saturado en estado estacionario en dominios bidimensionales y tridimensionales considerando campos de conductividad hidráulica altamente heterogéneos. La linealización del problema se realiza mediante un esquema de Picard y la aproximación numérica mediante métodos híbridos mixtos de elementos finitos. Este procedimiento numérico es utilizado para obtener parámetros hidráulicos efectivos y relaciones constitutivas de rocas fracturadas y rocas heterogéneas con distinto grado de correlación espacial. Se presenta además un análisis comparativo de las diferencias que surgen al representar las heterogeneidades en dos y tres dimensiones. La caracterización hidráulica de rocas fracturadas también es abordada mediante modelos analíticos. Utilizando conceptos de geometría fractal para representar la red de fracturas se derivan dos modelos constitutivos con expresiones analíticas cerradas para una y dos fases (agua y aire). Las expresiones analíticas son analizadas en términos de los parámetros involucrados, y validadas mediante su ajuste con datos experimentales y valores obtenidos con el experimento numérico descrito anteriormente. Por último, se analiza el efecto de la heterogeneidad de la conductividad hidráulica a escala de campo en acuíferos costeros afectados por la marea. Para ello se derivan soluciones analíticas exactas y aproximadas que permiten estimar las fluctuaciones inducidas en pozos para el caso de heterogeneidades lineales de la conductividad hidráulica. Las nuevas soluciones analíticas son comparadas con la correspondiente a un acuífero homogéneo y utilizadas para estimar parámetros hidráulicos a partir de datos experimentales disponibles en la literatura.

Light scattering is a powerful characterization tool for determining shape, size, and size distribution of fine particles, as well as complex, irregular structures of their aggregates. Small angle static light scattering and elliptically polarized light scattering techniques produce accurate results and provide real time, non-intrusive, and in-situ observations on prevailing process conditions in three-dimensional systems. As such, they complement conventional characterization tools such as SEM and TEM which have their known disadvantages and limitations. In this study, we provide a thorough light scattering analysis of colloidal tungsten trioxide (WO3) nanoparticles in the shape of irregular nanospheres and cylindrical nanowires, and of the resulting aggregate morphologies. Aggregation characteristics as a function of primary particle geometry, aspect ratio of nanowires, and the change in dispersion stability in various polar solvents without the use of dispersants are monitored over different time scales and are described using the concepts of fractal theory. Using forward scattered intensities, sedimentation rates as a result of electrolyte addition and particle concentration at low solution pH are quantified, in contrast to widely reported visual observations, and are related to the aggregate structure in the dispersed phase. For nanowires of high aspect ratios, when aggregate structures cannot directly be inferred from measurements, an analytical and a quasiexperimental method are used.

Diagnosis of faults associated with hydraulic actuators is essential to avoid accidents or loss of system functionality. This thesis focuses on internal leakage fault diagnosis in valve controlled hydraulic actuation systems (VCA) as well as electrohydrostatic actuation systems (EHA). For the VCA, the hydraulic actuator is driven in a closed loop mode to track a pseudorandom input signal whereas for the EHA, an actuator is driven in an open loop mode to track a sinusoidal input. Motivated by developing a method that does not rely on the model of the system or type of fault, signal processing techniques based on the ratio of metric lengths of pressure signals, autocorrelation of pressure signal, cross correlation between chamber pressure signals, and cross correlation between control signal and piston displacement is employed for internal leakage diagnosis. For the VCA, autocorrelation of pressure signals performed well at lower lags (less than 4) and at a window size of 200 data points; both cross correlation between pressure signals and cross correlation between control signal and piston displacement performed well at higher lags (higher than 8) and at a window size of 100 data points; ratio of metric lengths of pressure signals was found to be more effective at higher lag ratios (more than 16:3). All methods were sensitive to the lowest simulated leakage of 0.047 L/min, though with different level of success; ratio of metric lengths produced 84% sensitivity, autocorrelation 19% sensitivity, cross correlation between pressure signals 25% sensitivity and cross correlation between piston displacement and control signal 20% sensitivity. For the EHA, all methods were capable of identifying small leakage of 0.98 L/min. The ratio of metric lengths produced 6.7% sensitivity, autocorrelation 2.59% sensitivity, cross correlation between pressure signals 9.4% sensitivity and cross correlation between piston displacement and control signal 31.9% sensitivity. The low leakage detection achieved without requiring a model of the actuator or leakage type make these methods very attractive for industrial implementation

Den fraktala geometrin slog igenom i västvärlden under 60-70-talet och inspirerade vetenskapen till nya framsteg inom allt från datorteknologi till antennkonstruktion samt influerade konstens metodik och filosofi. Fraktaler som matematiska modeller hade studerats i väst redan på 1800 talet och ur afrikansk konst samt ur fontida egyptiska skulpturer kan vi skåda en betydligt äldre förståelse för de metoder som genererar dessa komplexa mönster. Inom musiken har fraktaler och självliknelse varit en del av kompositionsteorin sedan medeltiden, Josquin Des Prez och Johannes Ockeghem är två tidiga exempel som komponerade skalbara melodier som överlagrades i olika stämmor. I vår tid är Per Nörgård det mest framstående exemplet på tonsättare som inkorporerat fraktala tekniker i sina kompositioner. Oändlighetsserien, en heltalssekvens som upptäcktes av Nörgård i slutet av 50-talet inspirerade flera verk där de fraktala kvaliteterna återigen fick stå i förgrunden. I detta examensarbete diskuteras vad en fraktal är samt dess historiska bakgrund. Med utgång ifrån musikaliska exempel så härleds ett antal kompositionstekniker för att komponera musikaliska kanon samt fraktala sekvenser. Dessa kan göras för hand eller med hjälp av dator, kod finns i bilagorna. Slutligen med utgång från min egna musik för symfoniorkester samt stråkorkester tas konkreta exempel upp där dessa metoder använts. / <p>Patrik Ohlsson, Subtle Subsets för stråkorkester (2014), framfördes av Musica Vitae under ledning av dirigent Michael Bartosch.</p>

Humans are experts at understanding what they see. Similarity and analogy play a significant role in making sense of the visual world by forming analogies to similar images encountered previously. Yet, while these acts of visual reasoning may be commonplace, the processes of visual analogy are not yet well understood. In this dissertation, I investigate the utility of representing visual information in a fractal manner for computing visual similarity and analogy. In particular, I develop a computational technique of fractal reasoning for addressing problems of visual similarity and novelty. I illustrate the effectiveness of fractal reasoning on problems of visual similarity and analogy on the Raven’s Progressive Matrices and Miller’s Analogies tests of intelligence, problems of visual novelty and oddity on the Odd One Out test of intelligence, and problems of visual similarity and oddity on the Dehaene test of core geometric reasoning. I show that the performance of my computational model on these various tests is comparable to human performance. Fractal reasoning provides a new method for computing answers to such problems. Specifically, I show that the choice of the level of abstraction of problem representation determines the degree to which an answer may be regarded as confident, and that that choice of abstraction may be controlled automatically by the algorithm as a means of seeking that confident answer. This emergence of ambiguity and its remedy via problem re-representation is afforded by the fractal representation. I also show how reasoning over sparse data (at coarse levels of abstraction) or homogeneous data (at finest levels of abstraction) could both drive the automatic exclusion of certain levels of abstraction, as well as provide a signal to shift the analogical reasoning from consideration of simple analogies (such as analogies between pairs of objects) to more complex analogies (such as analogies among triplets, or larger groups of objects). My dissertation also explores fractal reasoning in perception, including both biologically-inspired imprinting and bistable perception. In particular, it provides a computational explanation of bistable perception in the famous Necker cube problem that is directly tied to the process of determining a confident interpretation via re-representation. Thus, my research makes two primary contributions to AI theories of visual similarity and analogy. The first contribution is the Extended Analogy By Recall (ABR*) algorithm, the computational technique for visual reasoning that automatically adjusts fractal representations to an appropriate level of abstraction. The second contribution is the fractal representation itself, a knowledge representation that add the notion of self-similarity and re-representation to analogy making.