Detectando não-linearidades nos retornos dos fundos multimercados

Estudos recentes apontam que diversas estratégias implementadas em hedge funds geram retornos com características não lineares. Seguindo as sugestões encontradas no paper de Agarwal e Naik (2004), este trabalho mostra que uma série de hedge funds dentro da indústria de fundos de investimentos no Brasil apresenta retornos que se assemelham ao de uma estratégia em opções de compra e venda no índice de mercado Bovespa. Partindo de um modelo de fatores, introduzimos um índice referenciado no retorno sobre opções de modo que tal fator possa explicar melhor que os tradicionais fatores de risco a característica não linear dos retornos dos fundos de investimento.

Modelagem do produto interno bruto brasileiro utilizando modelos não lineares

O trabalho tem como objetivo aplicar uma modelagem não linear ao Produto Interno Bruto brasileiro. Para tanto foi testada a existência de não linearidade do processo gerador dos dados com a metodologia sugerida por Castle e Henry (2010). O teste consiste em verificar a persistência dos regressores não lineares no modelo linear irrestrito. A seguir a série é modelada a partir do modelo autoregressivo com limiar utilizando a abordagem geral para específico na seleção do modelo. O algoritmo Autometrics é utilizado para escolha do modelo não linear. Os resultados encontrados indicam que o Produto Interno Bruto do Brasil é melhor explicado por um modelo não linear com três mudanças de regime, que ocorrem no inicio dos anos 90, que, de fato, foi um período bastante volátil. Através da modelagem não linear existe o potencial para datação de ciclos, no entanto os resultados encontrados não foram suficientes para tal análise.

Identificação em tempo real de modelo dinâmico de robô móvel com acionamento diferencial e zona morta

Several mobile robots show non-linear behavior, mainly due friction phenomena between the mechanical parts of the robot or between the robot and the ground. Linear models are efficient in some cases, but it is necessary take the robot non-linearity in consideration when precise displacement and positioning are desired. In this work a parametric model identification procedure for a mobile robot with differential drive that considers the dead-zone in the robot actuators is proposed. The method consists in dividing the system into Hammerstein systems and then uses the key-term separation principle to present the input-output relations which shows the parameters from both linear and non-linear blocks. The parameters are then simultaneously estimated through a recursive least squares algorithm. The results shows that is possible to identify the dead-zone thresholds together with the linear parameters

Controle por modos deslizantes com compensação difusa aplicado a sistemas com descontinuidade

The development of non-linear controllers gained space in the theoretical ambit and of practical applications on the moment that the arising of digital computers enabled the implementation of these methodologies. In comparison with the linear controllers more utilized, the non -linear controllers present the advantage of not requiring the linearity of the system to determine the parameters of control, which permits a more efficient control especially when the system presents a high level of non-linearity. Another additional advantage is the reduction of costs, since to obtain the efficient control through linear controllers it is necessary the utilization of sensors and more refined actuators than when it is utilized a non-linear controller. Among the non-linear theories of control, the method of control by gliding ways is detached for being a method that presents more robustness, before uncertainties. It is already confirmed that the adoption of compensation on the region of residual error permits to improve better the performance of these controllers. So, in this work it is described the development of a non-linear controller that looks for an association of strategy of control by gliding ways, with the fuzzy compensation technique. Through the implementation of some strategies of fuzzy compensation, it was searched the one which provided the biggest efficiency before a system with high level of nonlinearities and uncertainties. The electrohydraulic actuator was utilized as an example of research, and the results appoint to two configurations of compensation that permit a bigger reduction of the residual error

An efficient Hopfield network to solve economic dispatch problems with transmission system representation

Economic dispatch (ED) problems have recently been solved by artificial neural network approaches. Systems based on artificial neural networks have high computational rates due to the use of a massive number of simple processing elements and the high degree of connectivity between these elements. The ability of neural networks to realize some complex non-linear function makes them attractive for system optimization. All ED models solved by neural approaches described in the literature fail to represent the transmission system. Therefore, such procedures may calculate dispatch policies, which do not take into account important active power constraints. Another drawback pointed out in the literature is that some of the neural approaches fail to converge efficiently toward feasible equilibrium points. A modified Hopfield approach designed to solve ED problems with transmission system representation is presented in this paper. The transmission system is represented through linear load flow equations and constraints on active power flows. The internal parameters of such modified Hopfield networks are computed using the valid-subspace technique. These parameters guarantee the network convergence to feasible equilibrium points, which represent the solution for the ED problem. Simulation results and a sensitivity analysis involving IEEE 14-bus test system are presented to illustrate efficiency of the proposed approach. (C) 2004 Elsevier Ltd. All rights reserved.

Genetic Algorithms (Binary and Real Codes) for the Optimisation of a Fermentation Process for Butanol Production

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

Nonlinear effects generation in suspended core chalcogenide fibre

In this work we report our achievements in the elaboration and optical characterizations of low-losses suspended core optical fibers elaborated from As2S3 glass. For preforms elaboration, alternatively to other processes like the stack and draw or extrusion, we use a process based on mechanical drilling. The drawing of these drilled performs into fibers allows reaching a suspended core geometry, in which a 2 μm diameter core is linked to the fiber clad region by three supporting struts. The different fibers that have been drawn show losses close to 0.9 dB/m at 1.55 μm. The suspended core waveguide geometry has also an efficient influence on the chromatic dispersion and allows its management. Indeed, the zero dispersion wavelength, which is around 5 μm in the bulk glass, is calculated to be shifted towards around 2μm in our suspended core fibers. In order to qualify their nonlinearity we have pumped them at 1.995 μm with the help of a fibered ns source. We have observed a strong non linear response with evidence of spontaneous Raman scattering and strong spectral broadening. © 2011 SPIE.

Fenômenos não-lineares, incluindo-se os não-ideias, em captura de energia utilizando-se dispositivos piezoelétricos

Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES)

Utilização de redes neurais recorrentes na caracterização de cargas não lineares em sistemas elétricos

Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES)

Caos e termalização na teoria de Yang-Mills com quebra espontânea de simetria

Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES)

Systemic risk in dynamical networks with stochastic failure criterion

Complex non-linear interactions between banks and assets we model by two time-dependent Erdos-Renyi network models where each node, representing a bank, can invest either to a single asset (model I) or multiple assets (model II). We use a dynamical network approach to evaluate the collective financial failure -systemic risk- quantified by the fraction of active nodes. The systemic risk can be calculated over any future time period, divided into sub-periods, where within each sub-period banks may contiguously fail due to links to either i) assets or ii) other banks, controlled by two parameters, probability of internal failure p and threshold T-h ("solvency" parameter). The systemic risk decreases with the average network degree faster when all assets are equally distributed across banks than if assets are randomly distributed. The more inactive banks each bank can sustain (smaller T-h), the smaller the systemic risk -for some Th values in I we report a discontinuity in systemic risk. When contiguous spreading becomes stochastic ii) controlled by probability p(2) -a condition for the bank to be solvent (active) is stochasticthe- systemic risk decreases with decreasing p(2). We analyse the asset allocation for the U.S. banks. Copyright (C) EPLA, 2014

Crises de fronteira e estruturas complexas no espaço de parâmetros em um mapeamento dissipativo bidimensional

Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq)

Influence of structural nonlinearities in stall-induced aeroelastic response of pitching airfoils

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

Artificial neural networks applied to bandsawing process control

In the search for productivity increase, industry has invested on the development of intelligent, flexible and self-adjusting method, capable of controlling processes through the assistance of autonomous systems, independently whether they are hardware or software. Notwithstanding, simulating conventional computational techniques is rather challenging, regarding the complexity and non-linearity of the production systems. Compared to traditional models, the approach with Artificial Neural Networks (ANN) performs well as noise suppression and treatment of non-linear data. Therefore, the challenges in the wood industry justify the use of ANN as a tool for process improvement and, consequently, add value to the final product. Furthermore, Artificial Intelligence techniques such as Neuro-Fuzzy Networks (NFNs) have proven effective, since NFNs combine the ability to learn from previous examples and generalize the acquired information from the ANNs with the capacity of Fuzzy Logic to transform linguistic variables in rules.

Contribution to assessing the stiffness reduction of structural elements in the global stability analysis of precast concrete multi-storey buildings

This study deals with the reduction of the stiffness in precast concrete structural elements of multi-storey buildings to analyze global stability. Having reviewed the technical literature, this paper present indications of stiffness reduction in different codes, standards, and recommendations and compare these to the values found in the present study. The structural model analyzed in this study was constructed with finite elements using ANSYS® software. Physical Non-Linearity (PNL) was considered in relation to the diagrams M x N x 1/r, and Geometric Non-Linearity (GNL) was calculated following the Newton-Raphson method. Using a typical precast concrete structure with multiple floors and a semi-rigid beam-to-column connection, expressions for a stiffness reduction coefficient are presented. The main conclusions of the study are as follows: the reduction coefficients obtained from the diagram M x N x 1/r differ from standards that use a simplified consideration of PNL; the stiffness reduction coefficient for columns in the arrangements analyzed were approximately 0.5 to 0.6; and the variation of values found for stiffness reduction coefficient in concrete beams, which were subjected to the effects of creep with linear coefficients from 0 to 3, ranged from 0.45 to 0.2 for positive bending moments and 0.3 to 0.2 for negative bending moments.