Seasonal Dynamic Factor Analysis and Bootstrap Inference: Application to Electricity Market Forecasting

In this work, we propose the Seasonal Dynamic Factor Analysis (SeaDFA), an extension of Nonstationary Dynamic Factor Analysis, through which one can deal with dimensionality reduction in vectors of time series in such a way that both common and specific components are extracted. Furthermore, common factors are able to capture not only regular dynamics (stationary or not) but also seasonal ones, by means of the common factors following a multiplicative seasonal VARIMA(p, d, q) × (P, D, Q)s model. Additionally, a bootstrap procedure that does not need a backward representation of the model is proposed to be able to make inference for all the parameters in the model. A bootstrap scheme developed for forecasting includes uncertainty due to parameter estimation, allowing enhanced coverage of forecasting intervals. A challenging application is provided. The new proposed model and a bootstrap scheme are applied to an innovative subject in electricity markets: the computation of long-term point forecasts and prediction intervals of electricity prices. Several appendices with technical details, an illustrative example, and an additional table are available online as Supplementary Materials.

Factor analysis for metal grade exploration at Pallancata Vein in Peru

Se investiga la distribución espacial de contenidos metálicos analizados sobre testigos de sondeos obtenidos en las campañas de exploración de la Veta Pallancata. Se aplica el análisis factorial a dicha distribución y a los cocientes de los valores metálicos, discriminando los que están correlacionados con la mineralización argentífera y que sirven como guías de exploración para hallar zonas de potenciales reservas por sus gradientes de variación.Abstract:The metal distribution in a vein may show the paths of hydrothermal fluid flow at the time of mineralization. Such information may assist for in-fill drilling. The Pallancata Vein has been intersected by 52 drill holes, whose cores were sampled and analysed, and the results plotted to examine the mineralisation trends. The spatial distribution of the ore is observed from the logAg/logPb ratio distribution. Au is in this case closely related to Ag (electrum and uytenbogaardtite, Ag3AuS2 ). The Au grade shows the same spatial distribution as the Ag grade. The logAg/logPb ratio distribution also suggests possible ore to be expected at deeper locations. Shallow supergene Ag enrichment was also observed.

A methodology to estimate uncertainty for emission projections through sensitivity analysis

Air pollution abatement policies must be based on quantitative information on current and future emissions of pollutants. As emission projections uncertainties are inevitable and traditional statistical treatments of uncertainty are highly time/resources consuming, a simplified methodology for nonstatistical uncertainty estimation based on sensitivity analysis is presented in this work. The methodology was applied to the “with measures” scenario for Spain, concretely over the 12 highest emitting sectors regarding greenhouse gas and air pollutants emissions. Examples of methodology application for two important sectors (power plants, and agriculture and livestock) are shown and explained in depth. Uncertainty bands were obtained up to 2020 by modifying the driving factors of the 12 selected sectors and the methodology was tested against a recomputed emission trend in a low economic-growth perspective and official figures for 2010, showing a very good performance. Implications: A solid understanding and quantification of uncertainties related to atmospheric emission inventories and projections provide useful information for policy negotiations. However, as many of those uncertainties are irreducible, there is an interest on how they could be managed in order to derive robust policy conclusions. Taking this into account, a method developed to use sensitivity analysis as a source of information to derive nonstatistical uncertainty bands for emission projections is presented and applied to Spain. This method simplifies uncertainty assessment and allows other countries to take advantage of their sensitivity analyses.

A static heap analysis for shape and connectivity: Unified memory analysis: The base framework

Modeling the evolution of the state of program memory during program execution is critical to many parallehzation techniques. Current memory analysis techniques either provide very accurate information but run prohibitively slowly or produce very conservative results. An approach based on abstract interpretation is presented for analyzing programs at compile time, which can accurately determine many important program properties such as aliasing, logical data structures and shape. These properties are known to be critical for transforming a single threaded program into a versión that can be run on múltiple execution units in parallel. The analysis is shown to be of polynomial complexity in the size of the memory heap. Experimental results for benchmarks in the Jolden suite are given. These results show that in practice the analysis method is efflcient and is capable of accurately determining shape information in programs that créate and manipúlate complex data structures.

Urban built environment analysis: evidence from a mobility survey in Madrid

Applications involving travel behavior from the perspective of land use are dating from the 1990s. Usually, four important components are distinguished: density, diversity and design (3D?s of Cervero and Kockelman) and accessibility (introduced by Geurs and van Wee). But there is not a general agreement on how to measure each of those 4 components. Density is used to be measured as population and employment densities, but others authors separate population density between residential and building densities. A lot of measures have been developed to estimate diversity: among others, a dissimilarity index to indicate the degree to which different land uses lie within one another?s surrounding, an entropy index to quantify the degree of balance across various land use types or proximities to commercial-retail uses. Design has been characterized by site design, and dwelling and street characteristics. Lastly, accessibility has become a frequently used concept, but its meaning on travel behavior field always refers to the ability ?to reach activities or locations by means of a travel mode?, measured as accessibility to jobs, to leisure activities, and others. Furthermore, the previous evidence is mainly based on US data or on north European countries. Therefore, this paper adds some new evidence from a Spanish perspective to the research debate. Through a Madrid smartphone-based survey, factor analysis is used to linearly combine variables into the 3D?s and accessibility dimensions of the built environment. At a first step for future investigations, land use variables will be treated to define accurately the previous 4 components.

A dynamic factor model for mid-term forecasting of wind power generation

The main objective of this paper is the development and application of multivariate time series models for forecasting aggregated wind power production in a country or region. Nowadays, in Spain, Denmark or Germany there is an increasing penetration of this kind of renewable energy, somehow to reduce energy dependence on the exterior, but always linked with the increaseand uncertainty affecting the prices of fossil fuels. The disposal of accurate predictions of wind power generation is a crucial task both for the System Operator as well as for all the agents of the Market. However, the vast majority of works rarely onsider forecasting horizons longer than 48 hours, although they are of interest for the system planning and operation. In this paper we use Dynamic Factor Analysis, adapting and modifying it conveniently, to reach our aim: the computation of accurate forecasts for the aggregated wind power production in a country for a forecasting horizon as long as possible, particularly up to 60 days (2 months). We illustrate this methodology and the results obtained for real data in the leading country in wind power production: Denmark

Nonlinear seismic behaviour of concrete arch bridges

Arch bridge structural solution has been known for centuries, in fact the simple nature of arch that require low tension and shear strength was an advantage as the simple materials like stone and brick were the only option back in ancient centuries. By the pass of time especially after industrial revolution, the new materials were adopted in construction of arch bridges to reach longer spans. Nowadays one long span arch bridge is made of steel, concrete or combination of these two as "CFST", as the result of using these high strength materials, very long spans can be achieved. The current record for longest arch belongs to Chaotianmen bridge over Yangtze river in China with 552 meters span made of steel and the longest reinforced concrete type is Wanxian bridge which also cross the Yangtze river through a 420 meters span. Today the designer is no longer limited by span length as long as arch bridge is the most applicable solution among other approaches, i.e. cable stayed and suspended bridges are more reasonable if very long span is desired. Like any super structure, the economical and architectural aspects in construction of a bridge is extremely important, in other words, as a narrower bridge has better appearance, it also require smaller volume of material which make the design more economical. Design of such bridge, beside the high strength materials, requires precise structural analysis approaches capable of integrating the combination of material behaviour and complex geometry of structure and various types of loads which may be applied to bridge during its service life. Depend on the design strategy, analysis may only evaluates the linear elastic behaviour of structure or consider the nonlinear properties as well. Although most of structures in the past were designed to act in their elastic range, the rapid increase in computational capacity allow us to consider different sources of nonlinearities in order to achieve a more realistic evaluations where the dynamic behaviour of bridge is important especially in seismic zones where large movements may occur or structure experience P - _ effect during the earthquake. The above mentioned type of analysis is computationally expensive and very time consuming. In recent years, several methods were proposed in order to resolve this problem. Discussion of recent developments on these methods and their application on long span concrete arch bridges is the main goal of this research. Accordingly available long span concrete arch bridges have been studied to gather the critical information about their geometrical aspects and properties of their materials. Based on concluded information, several concrete arch bridges were designed for further studies. The main span of these bridges range from 100 to 400 meters. The Structural analysis methods implemented in in this study are as following: Elastic Analysis: Direct Response History Analysis (DRHA): This method solves the direct equation of motion over time history of applied acceleration or imposed load in linear elastic range. Modal Response History Analysis (MRHA): Similar to DRHA, this method is also based on time history, but the equation of motion is simplified to single degree of freedom system and calculates the response of each mode independently. Performing this analysis require less time than DRHA. Modal Response Spectrum Analysis (MRSA): As it is obvious from its name, this method calculates the peak response of structure for each mode and combine them using modal combination rules based on the introduced spectra of ground motion. This method is expected to be fastest among Elastic analysis. Inelastic Analysis: Nonlinear Response History Analysis (NL-RHA): The most accurate strategy to address significant nonlinearities in structural dynamics is undoubtedly the nonlinear response history analysis which is similar to DRHA but extended to inelastic range by updating the stiffness matrix for every iteration. This onerous task, clearly increase the computational cost especially for unsymmetrical buildings that requires to be analyzed in a full 3D model for taking the torsional effects in to consideration. Modal Pushover Analysis (MPA): The Modal Pushover Analysis is basically the MRHA but extended to inelastic stage. After all, the MRHA cannot solve the system of dynamics because the resisting force fs(u; u_ ) is unknown for inelastic stage. The solution of MPA for this obstacle is using the previously recorded fs to evaluate system of dynamics. Extended Modal Pushover Analysis (EMPA): Expanded Modal pushover is a one of very recent proposed methods which evaluates response of structure under multi-directional excitation using the modal pushover analysis strategy. In one specific mode,the original pushover neglect the contribution of the directions different than characteristic one, this is reasonable in regular symmetric building but a structure with complex shape like long span arch bridges may go through strong modal coupling. This method intend to consider modal coupling while it take same time of computation as MPA. Coupled Nonlinear Static Pushover Analysis (CNSP): The EMPA includes the contribution of non-characteristic direction to the formal MPA procedure. However the static pushovers in EMPA are performed individually for every mode, accordingly the resulted values from different modes can be combined but this is only valid in elastic phase; as soon as any element in structure starts yielding the neutral axis of that section is no longer fixed for both response during the earthquake, meaning the longitudinal deflection unavoidably affect the transverse one or vice versa. To overcome this drawback, the CNSP suggests executing pushover analysis for governing modes of each direction at the same time. This strategy is estimated to be more accurate than MPA and EMPA, moreover the calculation time is reduced because only one pushover analysis is required. Regardless of the strategy, the accuracy of structural analysis is highly dependent on modelling and numerical integration approaches used in evaluation of each method. Therefore the widely used Finite Element Method is implemented in process of all analysis performed in this research. In order to address the study, chapter 2, starts with gathered information about constructed long span arch bridges, this chapter continuous with geometrical and material definition of new models. Chapter 3 provides the detailed information about structural analysis strategies; furthermore the step by step description of procedure of all methods is available in Appendix A. The document ends with the description of results and conclusion of chapter 4.

Diseño de un modelo para la identificación y análisis de tramos de carreteras sin accidentes : una nueva visión de la seguridad vial

El planteamiento tradicional de análisis de la accidentalidad en carretera pasa por la consideración de herramientas paliativas, como son la identificación y gestión de los puntos negros o tramos de concentración de accidentes, o preventivas, como las auditorías e inspecciones de seguridad vial. En esta tesis doctoral se presenta un planteamiento complementario a estas herramientas, desde una perspectiva novedosa: la consideración de los tramos donde no se producen accidentes; son los denominados Tramos Blancos. La tesis persigue demostrar que existen determinados parámetros del diseño de las carreteras y del tráfico que, bajo características generales similares de las vías, tienen influencia en el hecho de que se produzcan o no accidentes, adicionalmente a la exposición al riesgo, como factor principal, y a otros factores. La propia definición de los Tramos Blancos, entendidos como tramos de carreteras de longitud representativa donde no se han producido accidentes con víctimas mortales o heridos graves durante un periodo largo de tiempo, garantiza que esta situación no se produzca como consecuencia de la aleatoriedad de los accidentes, sino que pudiera deberse a una confluencia específica de determinados parámetros de la geometría de la vía y del tráfico total y de vehículos pesados. Para el desarrollo de esta investigación se han considerado la red de autopistas de peaje y las carreteras convencionales de la Red del Estado de España, que supone un total de 17.000 kilómetros, y los datos de accidentes con víctimas mortales y heridos graves en el periodo 2006-2010, ambos incluidos, en estas redes (un total de 10.000 accidentes). La red viaria objeto de análisis supone el 65% de la longitud de la Red de Carreteras del Estado, por la que circula el 33% de su tráfico; en ella se produjeron en el año 2013 el 47% de los accidentes con víctimas y el 60% de las víctimas mortales de la Red de Carreteras del Estado. Durante la investigación se ha desarrollado una base de datos de 250.130 registros y más de 3.5 millones de datos en el caso de las autopistas de peaje de la Red de Carreteras del Estado y de 935.402 registros y más de 14 millones de datos en el caso de la red convencional del Estado analizada. Tanto las autopistas de peaje como las carreteras convencionales han sido clasificadas según sus características de tráfico, de manera que se valoren vías con nivel de exposición al riesgo similar. Para cada tipología de vía, se ha definido como longitud de referencia para que un tramo se considere Tramo Blanco la longitud igual al percentil 95 de las longitudes de tramos sin accidentes con heridos graves o víctimas mortales durante el periodo 2006-2010. En el caso de las autopistas de peaje, en la tipología que ha sido considerada para la definición del modelo, esta longitud de referencia se estableció en 14.5 kilómetros, mientras que en el caso de las carreteras convencionales, se estableció en 7.75 kilómetros. Para cada uno de los tipos de vía considerados se han construido una base de datos en la que se han incluido las variables de existencia o no de Tramo Blanco, así como las variables de tráfico (intensidad media diaria total, intensidad de vehículos pesados y porcentaje de vehículos pesados ), la velocidad media y las variables de geometría (número de carriles, ancho de carril, ancho de arcén derecho e izquierdo, ancho de calzada y plataforma, radio, peralte, pendiente y visibilidad directa e inversa en los casos disponibles); como variables adicionales, se han incluido el número de accidentes con víctimas, los fallecidos y heridos graves, índices de peligrosidad, índices de mortalidad y exposición al riesgo. Los trabajos desarrollados para explicar la presencia de Tramos Blancos en la red de autopistas de peaje han permitido establecer las diferencias entre los valores medios de las variables de tráfico y diseño geométrico en Tramos Blancos respecto a tramos no blancos y comprobar que estas diferencias son significativas. Así mismo, se ha podido calibrar un modelo de regresión logística que explica parcialmente la existencia de Tramos Blancos, para rangos de tráfico inferiores a 10.000 vehículos diarios y para tráficos entre 10.000 y 15.000 vehículos diarios. Para el primer grupo (menos de 10.000 vehículos al día), las variables que han demostrado tener una mayor influencia en la existencia de Tramo Blanco son la velocidad media de circulación, el ancho de carril, el ancho de arcén izquierdo y el porcentaje de vehículos pesados. Para el segundo grupo (entre 10.000 y 15.000 vehículos al día), las variables independientes más influyentes en la existencia de Tramo Blanco han sido la velocidad de circulación, el ancho de calzada y el porcentaje de vehículos pesados. En el caso de las carreteras convencionales, los diferentes análisis realizados no han permitido identificar un modelo que consiga una buena clasificación de los Tramos Blancos. Aun así, se puede afirmar que los valores medios de las variables de intensidad de tráfico, radio, visibilidad, peralte y pendiente presentan diferencias significativas en los Tramos Blancos respecto a los no blancos, que varían en función de la intensidad de tráfico. Los resultados obtenidos deben considerarse como la conclusión de un análisis preliminar, dado que existen otros parámetros, tanto de diseño de la vía como de la circulación, el entorno, el factor humano o el vehículo que podrían tener una influencia en el hecho que se analiza, y no se han considerado por no disponer de esta información. En esta misma línea, el análisis de las circunstancias que rodean al viaje que el usuario de la vía realiza, su tipología y motivación es una fuente de información de interés de la que no se tienen datos y que permitiría mejorar el análisis de accidentalidad en general, y en particular el de esta investigación. Adicionalmente, se reconocen limitaciones en el desarrollo de esta investigación, en las que sería preciso profundizar en el futuro, reconociendo así nuevas líneas de investigación de interés. The traditional approach to road accidents analysis has been based in the use of palliative tools, such as black spot (or road sections) identification and management, or preventive tools, such as road safety audits and inspections. This thesis shows a complementary approach to the existing tools, from a new perspective: the consideration of road sections where no accidents have occurred; these are the so-called White Road Sections. The aim of this thesis is to show that there are certain design parameters and traffic characteristics which, under similar circumstances for roads, have influence in the fact that accidents occur, in addition to the main factor, which is the risk exposure, and others. White Road Sections, defined as road sections of a representative length, where no fatal accidents or accidents involving serious injured have happened during a long period of time, should not be a product of randomness of accidents; on the contrary, they might be the consequence of a confluence of specific parameters of road geometry, traffic volumes and heavy vehicles traffic volumes. For this research, the toll motorway network and single-carriageway network of the Spanish National Road Network have been considered, which is a total of 17.000 kilometers; fatal accidents and those involving serious injured from the period 2006-2010 have been considered (a total number of 10.000 accidents). The road network covered means 65% of the total length of the National Road Network, which allocates 33% of traffic volume; 47% of accidents with victims and 60% of fatalities happened in these road networks during 2013. During the research, a database of 250.130 registers and more than 3.5 million data for toll motorways and 935.042 registers and more than 14 million data for single carriageways of the National Road Network was developed. Both toll motorways and single-carriageways have been classified according to their traffic characteristics, so that the analysis is performed over roads with similar risk exposure. For each road type, a reference length for White Road Section has been defined, as the 95 percentile of all road sections lengths without accidents (with fatalities or serious injured) for 2006-2010. For toll motorways, this reference length concluded to be 14.5 kilometers, while for single-carriageways, it was defined as 7.75 kilometers. A detailed database was developed for each type of road, including the variable “existence of White Road Section”, as well as variables of traffic (average daily traffic volume, heavy vehicles average daily traffic and percentage of heavy vehicles from the total traffic volume), average speed and geometry variables (number of lanes, width of lane, width of shoulders, carriageway width, platform width, radius, superelevation, slope and visibility); additional variables, such as number of accidents with victims, number of fatalities or serious injured, risk and fatality rates and risk exposure, have also been included. Research conducted for the explanation of the presence of White Road Sections in the toll motorway network have shown statistically significant differences in the average values of variables of traffic and geometric design in White Road Sections compared with other road sections. In addition, a binary logistic model for the partial explanation of the presence of White Road Sections was developed, for traffic volumes lower than 10.000 daily vehicles and for those running from 10.000 to 15.000 daily vehicles. For the first group, the most influent variables for the presence of White Road Sections were the average speed, width of lane, width of left shoulder and percentage of heavy vehicles. For the second group, the most influent variables were found to be average speed, carriageway width and percentage of heavy vehicles. For single-carriageways, the different analysis developed did not reach a proper model for the explanation of White Road Sections. However, it can be assumed that the average values of the variables of traffic volume, radius, visibility, superelevation and slope show significant differences in White Road Sections if compared with others, which also vary with traffic volumes. Results obtained should be considered as a conclusion of a preliminary analysis, as there are other parameters, not only design-related, but also regarding traffic, environment, human factor and vehicle which could have an influence in the fact under research, but this information has not been considered in the analysis, as it was not available. In parallel, the analysis of the circumstances around the trip, including its typology and motivation is an interesting source of information, from which data are not available; the availability of this information would be useful for the improvement of accident analysis, in general, and for this research work, in particular. In addition, there are some limitations in the development of the research work; it would be necessary to develop an in-depth analysis in the future, thus assuming new research lines of interest.