Short-term energy forecasting using time series techniques: Chile, a case study

Includes bibliography

Previsão de séries temporais financeiras por meio de redes neurais dinâmicas e processos de transformação de dados: uma abordagem empírico-comparativa

Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq)

Performance Analysis of MODIS 500-m Spatial Resolution Products for Estimating Chlorophyll-a Concentrations in Oligo-to Meso-Trophic Waters Case Study: Itumbiara Reservoir, Brazil

Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)

The bias in reversing the Box-Cox transformation in time series forecasting: An empirical study based on neural networks

The Box-Cox transformation is a technique mostly utilized to turn the probabilistic distribution of a time series data into approximately normal. And this helps statistical and neural models to perform more accurate forecastings. However, it introduces a bias when the reversion of the transformation is conducted with the predicted data. The statistical methods to perform a bias-free reversion require, necessarily, the assumption of Gaussianity of the transformed data distribution, which is a rare event in real-world time series. So, the aim of this study was to provide an effective method of removing the bias when the reversion of the Box-Cox transformation is executed. Thus, the developed method is based on a focused time lagged feedforward neural network, which does not require any assumption about the transformed data distribution. Therefore, to evaluate the performance of the proposed method, numerical simulations were conducted and the Mean Absolute Percentage Error, the Theil Inequality Index and the Signal-to-Noise ratio of 20-step-ahead forecasts of 40 time series were compared, and the results obtained indicate that the proposed reversion method is valid and justifies new studies. (C) 2014 Elsevier B.V. All rights reserved.

INFLUENCE OF THE VENEER-FRAMEWORK INTERFACE ON THE MECHANICAL BEHAVIOR OF CERAMIC VENEERS: A NONLINEAR FINITE ELEMENT ANALYSIS

Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)

Calibration of a Species-Specific Spectral Vegetation Index for Leaf Area Index (LAI) Monitoring: Example with MODIS Reflectance Time-Series on Eucalyptus Plantations

The leaf area index (LAI) is a key characteristic of forest ecosystems. Estimations of LAI from satellite images generally rely on spectral vegetation indices (SVIs) or radiative transfer model (RTM) inversions. We have developed a new and precise method suitable for practical application, consisting of building a species-specific SVI that is best-suited to both sensor and vegetation characteristics. Such an SVI requires calibration on a large number of representative vegetation conditions. We developed a two-step approach: (1) estimation of LAI on a subset of satellite data through RTM inversion; and (2) the calibration of a vegetation index on these estimated LAI. We applied this methodology to Eucalyptus plantations which have highly variable LAI in time and space. Previous results showed that an RTM inversion of Moderate Resolution Imaging Spectroradiometer (MODIS) near-infrared and red reflectance allowed good retrieval performance (R-2 = 0.80, RMSE = 0.41), but was computationally difficult. Here, the RTM results were used to calibrate a dedicated vegetation index (called "EucVI") which gave similar LAI retrieval results but in a simpler way. The R-2 of the regression between measured and EucVI-simulated LAI values on a validation dataset was 0.68, and the RMSE was 0.49. The additional use of stand age and day of year in the SVI equation slightly increased the performance of the index (R-2 = 0.77 and RMSE = 0.41). This simple index opens the way to an easily applicable retrieval of Eucalyptus LAI from MODIS data, which could be used in an operational way.

Stochastic simulation of time-series models combined with geostatistics to predict water-table scenarios in a Guarani Aquifer System outcrop area, Brazil

Stochastic methods based on time-series modeling combined with geostatistics can be useful tools to describe the variability of water-table levels in time and space and to account for uncertainty. Monitoring water-level networks can give information about the dynamic of the aquifer domain in both dimensions. Time-series modeling is an elegant way to treat monitoring data without the complexity of physical mechanistic models. Time-series model predictions can be interpolated spatially, with the spatial differences in water-table dynamics determined by the spatial variation in the system properties and the temporal variation driven by the dynamics of the inputs into the system. An integration of stochastic methods is presented, based on time-series modeling and geostatistics as a framework to predict water levels for decision making in groundwater management and land-use planning. The methodology is applied in a case study in a Guarani Aquifer System (GAS) outcrop area located in the southeastern part of Brazil. Communication of results in a clear and understandable form, via simulated scenarios, is discussed as an alternative, when translating scientific knowledge into applications of stochastic hydrogeology in large aquifers with limited monitoring network coverage like the GAS.

Neural coding tools, based on Information Theory, applied to discrete time series: from electrophysiology to neuroethology

This work is supported by Brazilian agencies Fapesp, CAPES and CNPq

Comparison between the complete Bayesian method and empirical Bayesian method for ARCH models using Brazilian financial time series

In this work we compared the estimates of the parameters of ARCH models using a complete Bayesian method and an empirical Bayesian method in which we adopted a non-informative prior distribution and informative prior distribution, respectively. We also considered a reparameterization of those models in order to map the space of the parameters into real space. This procedure permits choosing prior normal distributions for the transformed parameters. The posterior summaries were obtained using Monte Carlo Markov chain methods (MCMC). The methodology was evaluated by considering the Telebras series from the Brazilian financial market. The results show that the two methods are able to adjust ARCH models with different numbers of parameters. The empirical Bayesian method provided a more parsimonious model to the data and better adjustment than the complete Bayesian method.

CID: an efficient complexity-invariant distance for time series

The ubiquity of time series data across almost all human endeavors has produced a great interest in time series data mining in the last decade. While dozens of classification algorithms have been applied to time series, recent empirical evidence strongly suggests that simple nearest neighbor classification is exceptionally difficult to beat. The choice of distance measure used by the nearest neighbor algorithm is important, and depends on the invariances required by the domain. For example, motion capture data typically requires invariance to warping, and cardiology data requires invariance to the baseline (the mean value). Similarly, recent work suggests that for time series clustering, the choice of clustering algorithm is much less important than the choice of distance measure used.In this work we make a somewhat surprising claim. There is an invariance that the community seems to have missed, complexity invariance. Intuitively, the problem is that in many domains the different classes may have different complexities, and pairs of complex objects, even those which subjectively may seem very similar to the human eye, tend to be further apart under current distance measures than pairs of simple objects. This fact introduces errors in nearest neighbor classification, where some complex objects may be incorrectly assigned to a simpler class. Similarly, for clustering this effect can introduce errors by “suggesting” to the clustering algorithm that subjectively similar, but complex objects belong in a sparser and larger diameter cluster than is truly warranted.We introduce the first complexity-invariant distance measure for time series, and show that it generally produces significant improvements in classification and clustering accuracy. We further show that this improvement does not compromise efficiency, since we can lower bound the measure and use a modification of triangular inequality, thus making use of most existing indexing and data mining algorithms. We evaluate our ideas with the largest and most comprehensive set of time series mining experiments ever attempted in a single work, and show that complexity-invariant distance measures can produce improvements in classification and clustering in the vast majority of cases.

Pushover analysis of an existing reinforced concrete bridge:Jamboree Road Overcrossing in Irvine, California

The work for the present thesis started in California, during my semester as an exchange student overseas. California is known worldwide for its seismicity and its effort in the earthquake engineering research field. For this reason, I immediately found interesting the Structural Dynamics Professor, Maria Q. Feng's proposal, to work on a pushover analysis of the existing Jamboree Road Overcrossing bridge. Concrete is a popular building material in California, and for the most part, it serves its functions well. However, concrete is inherently brittle and performs poorly during earthquakes if not reinforced properly. The San Fernando Earthquake of 1971 dramatically demonstrated this characteristic. Shortly thereafter, code writers revised the design provisions for new concrete buildings so to provide adequate ductility to resist strong ground shaking. There remain, nonetheless, millions of square feet of non-ductile concrete buildings in California. The purpose of this work is to perform a Pushover Analysis and compare the results with those of a Nonlinear Time-History Analysis of an existing bridge, located in Southern California. The analyses have been executed through the software OpenSees, the Open System for Earthquake Engineering Simulation. The bridge Jamboree Road Overcrossing is classified as a Standard Ordinary Bridge. In fact, the JRO is a typical three-span continuous cast-in-place prestressed post-tension box-girder. The total length of the bridge is 366 ft., and the height of the two bents are respectively 26,41 ft. and 28,41 ft.. Both the Pushover Analysis and the Nonlinear Time-History Analysis require the use of a model that takes into account for the nonlinearities of the system. In fact, in order to execute nonlinear analyses of highway bridges it is essential to incorporate an accurate model of the material behavior. It has been observed that, after the occurrence of destructive earthquakes, one of the most damaged elements on highway bridges is a column. To evaluate the performance of bridge columns during seismic events an adequate model of the column must be incorporated. Part of the work of the present thesis is, in fact, dedicated to the modeling of bents. Different types of nonlinear element have been studied and modeled, with emphasis on the plasticity zone length determination and location. Furthermore, different models for concrete and steel materials have been considered, and the selection of the parameters that define the constitutive laws of the different materials have been accurate. The work is structured into four chapters, to follow a brief overview of the content. The first chapter introduces the concepts related to capacity design, as the actual philosophy of seismic design. Furthermore, nonlinear analyses both static, pushover, and dynamic, time-history, are presented. The final paragraph concludes with a short description on how to determine the seismic demand at a specific site, according to the latest design criteria in California. The second chapter deals with the formulation of force-based finite elements and the issues regarding the objectivity of the response in nonlinear field. Both concentrated and distributed plasticity elements are discussed into detail. The third chapter presents the existing structure, the software used OpenSees, and the modeling assumptions and issues. The creation of the nonlinear model represents a central part in this work. Nonlinear material constitutive laws, for concrete and reinforcing steel, are discussed into detail; as well as the different scenarios employed in the columns modeling. Finally, the results of the pushover analysis are presented in chapter four. Capacity curves are examined for the different model scenarios used, and failure modes of concrete and steel are discussed. Capacity curve is converted into capacity spectrum and intersected with the design spectrum. In the last paragraph, the results of nonlinear time-history analyses are compared to those of pushover analysis.

Die Rolle der Globalstrahlung im Klimasystem Südwestdeutschlands

Die Untersuchungen umfassen die Periode 1981 – 2000 und basieren hauptsächlich auf Daten des Deutschen Wetterdienstes (DWD). Relativwerte der Globalstrahlung beziehen sich auf die Rayleigh-Atmosphäre. Das Regressionsmodell nach Angström ermöglicht die Erweiterung des Meßnetzes. In linearer und nichtlinearer Regression und Korrelation ist die Globalstrahlung entweder abhängige (Sonnenscheindauer, Bewölkung) oder unabhängige Variable (Lufttemperatur, Bodentemperatur). Ihre Intensität in Abhängigkeit von Großwetterlagen, Großwettertypen und Luftmassen wird diskutiert. Diesbezüglich werden mit der Linearen Diskriminanzanalyse ähnliche Großwetterlagen und Stationen in signifikant unterschiedenen Gruppen zusammengefaßt, getrennt nach Sommer- und Winterhalbjahr. Abhängig von der Zeit betrachtet, enthalten Globalstrahlung, direkte und diffuse Sonnenstrahlung, Lufttemperatur, Bewölkung und Niederschlag signifikante zyklische Variationen, die gegebenenfalls klimatologisch relevant sind. Weiteren Aufschluß ergeben deshalb die Zeitreihenanalysen. Autokorrelation-Spektralanalysen (ASA) der genannten Variablen werden in integrierten Spektren dargestellt. Hinweise auf die zeitliche Konstanz signifikanter Varianzmaxima enthalten die Spektren der dynamischen (gleitenden) ASA.

Early Warning For Large Earthquakes: Observations, Models and Real-Time Data Analysis

This thesis is a collection of works focused on the topic of Earthquake Early Warning, with a special attention to large magnitude events. The topic is addressed from different points of view and the structure of the thesis reflects the variety of the aspects which have been analyzed. The first part is dedicated to the giant, 2011 Tohoku-Oki earthquake. The main features of the rupture process are first discussed. The earthquake is then used as a case study to test the feasibility Early Warning methodologies for very large events. Limitations of the standard approaches for large events arise in this chapter. The difficulties are related to the real-time magnitude estimate from the first few seconds of recorded signal. An evolutionary strategy for the real-time magnitude estimate is proposed and applied to the single Tohoku-Oki earthquake. In the second part of the thesis a larger number of earthquakes is analyzed, including small, moderate and large events. Starting from the measurement of two Early Warning parameters, the behavior of small and large earthquakes in the initial portion of recorded signals is investigated. The aim is to understand whether small and large earthquakes can be distinguished from the initial stage of their rupture process. A physical model and a plausible interpretation to justify the observations are proposed. The third part of the thesis is focused on practical, real-time approaches for the rapid identification of the potentially damaged zone during a seismic event. Two different approaches for the rapid prediction of the damage area are proposed and tested. The first one is a threshold-based method which uses traditional seismic data. Then an innovative approach using continuous, GPS data is explored. Both strategies improve the prediction of large scale effects of strong earthquakes.