A novel ranking procedure for forecasting approaches using Data Envelopment Analysis

To compare the accuracy of different forecasting approaches an error measure is required. Many error measures have been proposed in the literature, however in practice there are some situations where different measures yield different decisions on forecasting approach selection and there is no agreement on which approach should be used. Generally forecasting measures represent ratios or percentages providing an overall image of how well fitted the forecasting technique is to the observations. This paper proposes a multiplicative Data Envelopment Analysis (DEA) model in order to rank several forecasting techniques. We demonstrate the proposed model by applying it to the set of yearly time series of the M3 competition. The usefulness of the proposed approach has been tested using the M3-competition where five error measures have been applied in and aggregated to a single DEA score.

Ontology forecasting in scientific literature:semantic concepts prediction based on innovation-adoption priors

The ontology engineering research community has focused for many years on supporting the creation, development and evolution of ontologies. Ontology forecasting, which aims at predicting semantic changes in an ontology, represents instead a new challenge. In this paper, we want to give a contribution to this novel endeavour by focusing on the task of forecasting semantic concepts in the research domain. Indeed, ontologies representing scientific disciplines contain only research topics that are already popular enough to be selected by human experts or automatic algorithms. They are thus unfit to support tasks which require the ability of describing and exploring the forefront of research, such as trend detection and horizon scanning. We address this issue by introducing the Semantic Innovation Forecast (SIF) model, which predicts new concepts of an ontology at time t + 1, using only data available at time t. Our approach relies on lexical innovation and adoption information extracted from historical data. We evaluated the SIF model on a very large dataset consisting of over one million scientific papers belonging to the Computer Science domain: the outcomes show that the proposed approach offers a competitive boost in mean average precision-at-ten compared to the baselines when forecasting over 5 years.

GA-ANN Short-Term Electricity Load Forecasting

This paper presents a methodology for short-term load forecasting based on genetic algorithm feature selection and artificial neural network modeling. A feed forward artificial neural network is used to model the 24-h ahead load based on past consumption, weather and stock index data. A genetic algorithm is used in order to find the best subset of variables for modeling. Three data sets of different geographical locations, encompassing areas of different dimensions with distinct load profiles are used in order to evaluate the methodology. The developed approach was found to generate models achieving a minimum mean average percentage error under 2 %. The feature selection algorithm was able to significantly reduce the number of used features and increase the accuracy of the models.

Coastal modelling studies for forecasting and remediation solutions

The coastal area along the Emilia-Romagna (ER), in the Italian side of the northern Adriatic Sea, is considered to implement an unstructured numerical ocean model with the aim to develop innovative tools for the coastal management and a forecasting system for the storm surge risk reduction. The Adriatic Sea has been the focus of several studies because of its peculiar dynamics driven by many forcings acting at basin and local scales. The ER coast is particularly exposed to storm surge events. In particular conditions, winds, tides and seicehs may combine and contribute to the flooding of the coastal area. The global sea level rise expected in the next decades will increase even more the hazard along the ER and Adriatic coast. Reliable Adriatic and Mediterranean scale numerical ocean models are now available to allow the dynamical downscaling of very high-resolution models in limited coastal areas. In this work the numerical ocean model SHYFEM is implemented in the Goro lagoon (named GOLFEM) and along the ER coast (ShyfER) to test innovative solutions against sea related coastal hazards. GOLFEM was succesfully applied to analyze the Goro lagoon dynamics and to assess the dynamical effects of human interventions through the analysis of what-if scenarios. The assessment of storm surge hazard in the Goro lagoon was carried out through the development of an ensemble storm surge forecasting system with GOLFEM using forcing from different operational meteorological and ocean models showing the fundamental importance of the boundary conditions. The ShyfER domain is used to investigate innovative solutions against storm surge related hazard along the ER coast. The seagrass is assessed as a nature-based solution (NBS) for coastal protection under present and future climate conditions. The results show negligible effects on sea level but sensible effects in reducing bottom current velocity.

Urban air quality: new high-resolution modeling approaches and forecasting tools for citizens

Air pollution is one of the greatest health risks in the world. At the same time, the strong correlation with climate change, as well as with Urban Heat Island and Heat Waves, make more intense the effects of all these phenomena. A good air quality and high levels of thermal comfort are the big goals to be reached in urban areas in coming years. Air quality forecast help decision makers to improve air quality and public health strategies, mitigating the occurrence of acute air pollution episodes. Air quality forecasting approaches combine an ensemble of models to provide forecasts from global to regional air pollution and downscaling for selected countries and regions. The development of models dedicated to urban air quality issues requires a good set of data regarding the urban morphology and building material characteristics. Only few examples of air quality forecast system at urban scale exist in the literature and often they are limited to selected cities. This thesis develops by setting up a methodology for the development of a forecasting tool. The forecasting tool can be adapted to all cities and uses a new parametrization for vegetated areas. The parametrization method, based on aerodynamic parameters, produce the urban spatially varying roughness. At the core of the forecasting tool there is a dispersion model (urban scale) used in forecasting mode, and the meteorological and background concentration forecasts provided by two regional numerical weather forecasting models. The tool produces the 1-day spatial forecast of NO2, PM10, O3 concentration, the air temperature, the air humidity and BLQ-Air index values. The tool is automatized to run every day, the maps produced are displayed on the e-Globus platform, updated every day. The results obtained indicate that the forecasting output were in good agreement with the observed measurements.

Power flow forecasting and smart charging for electric vehicle charging systems with photovoltaic integration

An essential role in the global energy transition is attributed to Electric Vehicles (EVs) the energy for EV traction can be generated by renewable energy sources (RES), also at a local level through distributed power plants, such as photovoltaic (PV) systems. However, EV integration with electrical systems might not be straightforward. The intermittent RES, combined with the high and uncontrolled aggregate EV charging, require an evolution toward new planning and paradigms of energy systems. In this context, this work aims to provide a practical solution for EV charging integration in electrical systems with RES. A method for predicting the power required by an EV fleet at the charging hub (CH) is developed in this thesis. The proposed forecasting method considers the main parameters on which charging demand depends. The results of the EV charging forecasting method are deeply analyzed under different scenarios. To reduce the EV load intermittency, methods for managing the charging power of EVs are proposed. The main target was to provide Charging Management Systems (CMS) that modulate EV charging to optimize specific performance indicators such as system self-consumption, peak load reduction, and PV exploitation. Controlling the EV charging power to achieve specific optimization goals is also known as Smart Charging (SC). The proposed techniques are applied to real-world scenarios demonstrating performance improvements in using SC strategies. A viable alternative to maximize integration with intermittent RES generation is the integration of energy storage. Battery Energy Storage Systems (BESS) may be a buffer between peak load and RES production. A sizing algorithm for PV+BESS integration in EV charging hubs is provided. The sizing optimization aims to optimize the system's energy and economic performance. The results provide an overview of the optimal size that the PV+BESS plant should have to improve whole system performance in different scenarios.

A coupled ocean-atmosphere system for short term forecasting in the southern european seas

This coupled model combines two state-of-the-art numerical models, NEMO for the oceanic component and WRF for the atmospheric component and implements them at an appropriate resolution. The oceanic model has been implemented starting from the Mediterranean Forecasting System with a resolution of 1/24° and the domain was extended to exactly match the grid of a newly implemented atmospheric model for the same area. The uncoupled ocean model has been validated against SST observed data, both in the simulation of an extreme event and in the short-term forecast of two seasonal periods. A new setup of the model was successfully tested in which the downward radiative fluxes were prescribed from atmospheric forecasts. Various physical schemes, domain, boundary, and initial conditions were tested with the atmospheric model to obtain the best representation of medicane Ianos. The heat fluxes calculated by the uncoupled models were compared to determine which setup gave the best energy balance between the components of the coupled model. The coupling strategy used is the traditional one, where the ocean is driven by the surface stress, heat fluxes, and radiative fluxes computed in the atmospheric component, which in turn receives the SST and surface currents. As expected, the overall skills of the coupled model are slightly degraded compared to the uncoupled models, even though the positioning and timing of the cyclone at the time of the landfall is enhanced. The mean heat fluxes do not change compared to the uncoupled model, whereas the pattern of the shortwave radiation and latent heat is changed. Moreover, the two energy fluxes are larger in absolute values than those calculated with the MFS formulas. The fact that they have opposite signs give raise to a compensation error that limits the overall degradation of the coupled simulation.

Sales Forecasting e Demand Planning all'interno del processo di Sales & Operations Planning: il caso Cefla s.c.

Il presente elaborato di tesi è stato realizzato coerentemente con quanto osservato in Cefla s.c., azienda italiana composta attualmente da 4 Business Unit che operano a livello internazionale in settori distinti. I temi trattati riguardano nel dettaglio la Business Unit Medical Equipment, la quale realizza prodotti a supporto del professionista sanitario in tutte le fasi della sua attività, comprendendo riuniti odontoiatrici, apparecchiature per l’imaging e radiologia digitale e sistemi di sterilizzazione. L’obiettivo di questo elaborato è quello di descrivere l’attuale processo di Sales & Operations Planning all’interno di questa divisione dell’azienda e contribuire alla progettazione del piano per la sua strutturazione, reso necessario dalla situazione di forte criticità che Cefla s.c. è stata costretta ad affrontare. Vengono quindi descritte le problematiche che caratterizzano i processi interni all’azienda allo stato attuale, la cui valutazione è stata supportata da consulenti esterni, al fine di evidenziare gli aspetti più critici ed elaborare proposte di miglioramento. Queste ultime sono distinte in funzione delle diverse figure coinvolte che hanno contribuito alla loro realizzazione e ai sottoprocessi interessati e che costituiscono il Sales & Operations Planning: Sales Forecasting, Demand Planning e Supply Planning. In particolare, vengono approfonditi i processi che riguardano la previsione della domanda, in quanto per essi è stato possibile collaborare nell’elaborazione di proposte di miglioramento mirate. Visti i tempi medio lunghi che caratterizzano le soluzioni proposte all’azienda si è cercato di contribuire con la progettazione di proposte quick-win in ambito di Sales Forecasting e Demand Planning. Infine, si è tentato di quantificare i costi sostenuti da Cefla s.c. per far fronte alla situazione di criticità affrontata tramite valutazioni economiche e KPI, potendo così stimare l’impatto dato dall’implementazione di proposte di miglioramento.

Estimativas de insegurança alimentar grave nos municípios Brasileiros

Prevalence of severe food insecurity was estimated for Brazilian municipalities based on the 2004 National Household Sample Survey (PNAD). First, a logistic regression model was developed and tested with this database. The model was then applied to the 2000 census data, generating severe food insecurity estimates for the Brazilian municipalities, which were subsequently analyzed according to the proportion of families exposed to severe food insecurity. Severe food insecurity was mainly concentrated in the North and Northeast regions, where 46.1% and 65.3% of municipalities showed high prevalence of severe food insecurity, respectively. Most municipalities in the Central West region showed intermediate prevalence of severe food insecurity. There was wide intra-regional variation in severe food insecurity, while the South of Brazil showed the most uniform distribution. In conclusion, Brazil displays wide inter and intra-regional variations in the occurrence of severe food insecurity. Such variations should be identified and analyzed in order to plan appropriate public policies.

Estimativas de insegurança alimentar grave nos municípios brasileiros

Estimou-se a prevalência de insegurança alimentar grave para os municípios brasileiros, com base na Pesquisa Nacional por Amostra de Domicílios (PNAD) 2004. Inicialmente, foi gerado e testado um modelo por regressão logística multivariada com base nesse banco de dados. O modelo foi aplicado à amostra do Censo Demográfico de 2000, gerando estimativas de prevalências de insegurança alimentar grave para os municípios brasileiros, que foram analisadas de acordo com o percentual de famílias em condição de insegurança alimentar grave. Essa insegurança alimentar está mais concentrada nas regiões Norte e Nordeste, onde 46,1 por cento e 65,3 por cento dos municípios, respectivamente, apresentam altas prevalências de insegurança alimentar grave. Predominam nas regiões Sudeste e Sul municípios com baixa exposição à insegurança alimentar grave. No Centro-oeste a maior parte dos municípios mostra estimativas de insegurança alimentar grave classificadas como médias. Verificou-se grande variação intra-regional na ocorrência da insegurança alimentar, sendo a Região Sul a mais uniforme. Conclui-se que o Brasil apresenta grandes variações inter e intra-regionais na ocorrência da insegurança alimentar, sendo importante realçá-las e analisá-las, no intuito de subsidiar políticas públicas

System Identification and Prediction of Dengue Fever Incidence in Rio de Janeiro

Identification, prediction, and control of a system are engineering subjects, regardless of the nature of the system. Here, the temporal evolution of the number of individuals with dengue fever weekly recorded in the city of Rio de Janeiro, Brazil, during 2007, is used to identify SIS (susceptible-infective-susceptible) and SIR (susceptible-infective-removed) models formulated in terms of cellular automaton (CA). In the identification process, a genetic algorithm (GA) is utilized to find the probabilities of the state transition S -> I able of reproducing in the CA lattice the historical series of 2007. These probabilities depend on the number of infective neighbors. Time-varying and non-time-varying probabilities, three different sizes of lattices, and two kinds of coupling topology among the cells are taken into consideration. Then, these epidemiological models built by combining CA and GA are employed for predicting the cases of sick persons in 2008. Such models can be useful for forecasting and controlling the spreading of this infectious disease.

Methodology to forecast the harvest date of banana bunches

Introduction. This method is used to forecast the harvest date of banana bunches from as early as the plant shooting stage. It facilitates the harvest of bunches with the same physiological age. The principle, key advantages, time required and expected results are presented. Materials and methods. Details of the four steps of the method ( installation of the temperature sensor, tagging bunches at the flowering stage, temperature sum calculation and estimation of bunch harvest date) are described. Possible problems are discussed. Results. The application of the method allows drawing a curve of the temperature sum accumulated by the bunches which have to be harvested at exactly 900 degree-days physiological age.

Integrating technology roadmapping and portfolio management at the front-end of new product development

Many authors point out that the front-end of new product development (NPD) is a critical success factor in the NPD process and that numerous companies face difficulties in carrying it out appropriately. Therefore, it is important to develop new theories and proposals that support the effective implementation of this earliest phase of NPD. This paper presents a new method to support the development of front-end activities based on integrating technology roadmapping (TRM) and project portfolio management (PPM). This new method, called the ITP Method, was implemented at a small Brazilian high-tech company in the nanotechnology industry to explore the integration proposal. The case study demonstrated that the ITP Method provides a systematic procedure for the fuzzy front-end and integrates innovation perspectives into a single roadmap, which allows for a better alignment of business efforts and communication of product innovation goals. Furthermore, the results indicated that the method may also improve quality, functional integration and strategy alignment. (C) 2010 Elsevier Inc. All rights reserved.

Model to predict the chloride penetration in concrete with profile forming peak: verification of its precision

Chloride attack in marine environments or in structures where deicing salts are used will not always show profiles with concentrations that decrease from the external surface to the interior of the concrete. Some profiles show an increase in chloride concentrations from when a peak is formed. This type of profile must be analyzed in a different way from the traditional model of Fick`s second law to generate more precise service life models. A model for forecasting the penetration of chloride ions as a function of time for profiles having formed a peak. To confirm the efficiency of this model, it is necessary to observe the behavior of a chloride profile with peak in a specific structure over a period of time. To achieve this, two chloride profiles with different ages (22 and 27 years) were extracted from the same structure. The profile obtained from the 22-year sample was used to estimate the chloride profile at 27 years using three models: a) the traditional model using Fick`s second law and extrapolating the value of C(S)-external surface chloride concentration; b) the traditional model using Fick`s second law and shifting the x-axis to the peak depth; c) the previously proposed model. The results from these models were compared with the actual profile measured in the 27-year sample and the results were analyzed. The model was presented with good precision for this study of case, requiring to be tested with other structures in use.

Development of Ramulosis Disease of Cotton Under Controlled Environment and Field Conditions

Colletotrichum gossypii var. cephalosporioides, the fungus that causes ramulosis disease of cotton, is widespread in Brazil and can cause severe yield loss. Because weather conditions greatly affect disease development, the objective of this work was to develop weather-based models to assess disease favorability. Latent period, incidence, and severity of ramulosis symptoms were evaluated in controlled environment experiments using factorial combinations of temperature (15, 20, 25, 30, and 35 degrees C) and leaf wetness duration (0, 4, 8, 16, 32, and 64 h after inoculation). Severity was modeled as an exponential function of leaf wetness duration and temperature. At the optimum temperature of disease development, 27 degrees C, average latent period was 10 days. Maximum ramulosis severity occurred from 20 to 30 degrees C, with sharp decreases at lower and higher temperatures. Ramulosis severity increased as wetness periods were increased from 4 to 32 h. In field experiments at Piracicaba, Sao Paulo State, Brazil, cotton plots were inoculated (10(5) conidia ml(-1)) and ramulosis severity was evaluated weekly. The model obtained from the controlled environment study was used to generate a disease favorability index for comparison with disease progress rate in the field. Hourly measurements of solar radiation, temperature, relative humidity, leaf wetness duration, rainfall, and wind speed were also evaluated as possible explanatory variables. Both the disease favorability model and a model based on rainfall explained ramulosis growth rate well, with R(2) of 0.89 and 0.91, respectively. They are proposed as models of ramulosis development rate on cotton in Brazil, and weather-disease relationships revealed by this work can form the basis of a warning system for ramulosis development.