Transport and dynamical properties for a bouncing ball model with regular and stochastic perturbations

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

Circular, elliptic and oval billiards in a gravitational field

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

Parâmetros ruminais e desaparecimento da FDN da forragem em bovinos suplementados em pastagem na estação das águas

O objetivo do presente trabalho foi determinar o efeito de suplementos concentrados com diferentes degradabilidades da proteína (alta-75%, média-55% e baixa-35%) e o efeito da quantidade dos mesmos (0,5, 1,0 e 1,5 kg de MS/dia) sobre os parâmetros ruminais (pH e N-NH3) e o desaparecimento da FDN da forragem em bovinos pastejando Brachiaria brizantha cv. Marandu no período das águas. Foram utilizados dez bovinos canulados no rúmen com peso médio de 463 kg num esquema fatorial 3 x 3 com três repetições (blocos). A técnica do saco de náilon e a simulação do pastejo foram usadas para alcançar a cinética de degradação da FDN da forragem. Foram determinadas as concentrações N-NH3 e o pH ruminais durante dois dias, em vários horários. Não houve efeito dos suplementos, tampouco da quantidade destes sobre o desaparecimento da FDN da forragem, mas houve para os diferentes tempos de incubação, alcançando o potencial máximo de degradação (67,58%) às 96 horas de incubação. O pH ruminal variou de 5,7 a 7,4 com média de 6,48, não havendo efeito de tratamento e nem interação tratamento x tempo. Já para os horários de determinação houve diferenças significativas. Para o N-NH3 ruminal houve diferença significativa entre os tratamentos, sendo as maiores concentrações observadas para aqueles que apresentavam maior degradabilidade da proteína. Os primeiros horários após a suplementação apresentavam um maior pico de N-NH3 e os picos do período da manhã foram maiores que os da tarde. As diferentes concentrações de N-NH3 ruminal não alteraram a degradação da fibra.

Eficácia do glyphosate e 2,4-d no controle de espécies de trapoerabas (commelina spp.)

The aim of this study was to evaluate the efficacy of glyphosate and 2,4-D for the Commelina benghalensis, Commelina diffusa and Commelina erecta. Three experiments were conducted in a greenhouse in a completely randomized in factorial scheme 3x7+1 (three herbicides x seven periods - hour after application) for two methods for measuring absorption of herbicides (simulated rain and cut the leaves applied), and a control without application, with four replications. The herbicides used were: glyphosate (1,080 g ha(-1)), 2,4-D amine (720 g ha(-1)) and, mixture glyphosate + 2,4-D (720 + 720 g ha(-1)). It was evaluated seven time intervals for washing (simulating rainfall after application) and cut (simulation abortion as a strategy of defense) the leaves applied: 0, 2, 4, 6, 8, 12 and 24 hours after herbicide application (HHA). The minimum period for absorption of glyphosate alone and in mixture with 2,4 -D presented a satisfactory control (> 90.0%) was around 12 HHA to C. benghalensis, C. diffusa and C. erecta, independently of washing or cut the leaves applied. This observed behavior can influence the time required without rain after herbicide application, besides that, the plants aborted part of the stem with injuries to avoid the herbicide translocation, especially when was used the 2,4-D alone. Concluded that glyphosate alone and in mixture with 2,4-D were efficient and that the differences in the control of species may be due to variation of time to absorb the herbicide.

Numerical solution of the eXtended Pom-Pom model for viscoelastic free surface flows

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

A novel approach based on recurrent neural networks applied to nonlinear systems optimization

This paper presents an efficient approach based on recurrent neural network for solving nonlinear optimization. More specifically, a modified Hopfield network is developed and its internal parameters are computed using the valid subspace technique. These parameters guarantee the convergence of the network to the equilibrium points that represent an optimal feasible solution. The main advantage of the developed network is that it treats optimization and constraint terms in different stages with no interference with each other. Moreover, the proposed approach does not require specification of penalty and weighting parameters for its initialization. A study of the modified Hopfield model is also developed to analyze its stability and convergence. Simulation results are provided to demonstrate the performance of the proposed neural network. (c) 2005 Elsevier B.V. All rights reserved.

Single-phase high power-factor boost ZCS pre-regulator operating in critical conduction mode

This paper presents a new single-phase interleaved high power factor boost pre-regulator operating in critical conduction mode, where the switches and boost diode performing zero-current commutations during its turn-off, eliminating the disadvantages related to the reverse recovery losses and electromagnetic interference problems of the boost diode, when operating in the continuous conduction mode. The interleaving technique is applied in the power cell, providing a significant input current ripple reduction in comparison to discontinuous mode of operation, due to its input current continuous conduction operation. This paper presents a complete modeling for the converter operating in critical conduction mode, resulting in an improved design procedure for interleaved techniques with high input power factor, a complete design procedure, and main simulation results from a design example with two interleaved cells rated at 1kW, 400V output voltage and 220V rms input voltage.

Hardware implementation of an analog neural nonderivative optimizer

Analog neural systems that can automatically find the minimum value of the outputs of unknown analog systems, described by convex functions, are studied. When information about derivative or gradient are not used, these systems are called analog nonderivative optimizers. An electronic circuit for the analog neural nonderivative optimizer proposed by Teixeira and Zak, and its simulation with software PSPICE, is presented. With the simulation results and hardware implementation of the system, the validity of the proposed optimizer can be verified. These results are original, from the best of the authors knowledge.

Generating VHDL-AMS models of digital-to-analogue converters from MATLAB (R)/SIMULINK (R)

Today, the trend within the electronics industry is for the use of rapid and advanced simulation methodologies in association with synthesis toolsets. This paper presents an approach developed to support mixed-signal circuit design and analysis. The methodology proposed shows a novel approach to the problem of developing behvioural model descriptions of mixed-signal circuit topologies, by construction of a set of subsystems, that supports the automated mapping of MATLAB (R)/SINIULINK (R) models to structural VHDL-AMS descriptions. The tool developed, named (MSSV)-S-2, reads a SIMULINK (R) model file and translates it to a structural VHDL-AMS code. It also creates the file structure required to simulate the translated model in the SystemVision (TM). To validate the methodology and the developed program, the DAC08, AD7524 and AD5450 data converters were studied and initially modelled in MATLAB (R)/SIMULINK (R). The VHDL-AMS code generated automatically by (MSSV)-S-2, (MATLAB (R)/SIMULINK (R) to SystemVision (TM)), was then simulated in the SystemVision (TM). The simulation results show that the proposed approach, which is based on VHDL-AMS descriptions of the original model library elements, allows for the behavioural level simulation of complex mixed-signal circuits.

Electrical load forecasting formulation by a fast neural network

The objective of this work is to develop a methodology for electric load forecasting based on a neural network. Here, backpropagation algorithm is used with an adaptive process that based on fuzzy logic and using a decaying exponential function to avoid instability in the convergence process. This methodology results in fast training, when compared to the conventional formulation of backpropagation algorithm. The results are presented using data from a Brazilian Electric Company, and shows a very good performance for the proposal objective.

Control aspects in nonlinear Hill's equation

The Hill's equations-even in the linear original version are a describer of phenomenon having chaotic flavor, giving sometimes very unusual situations. The theory of the so called intervals of instability in the equation provides the precise description for most of these phenomena. Considerations on nonlinearities into the Hill's equation is a quite recent task. The linearized version for almost of these systems it reduces to the Hill's classical linear one. In this paper, some indicative facts are pointed out on the possibility of having the linear system stabilizable and/or exactly controllable. As consequence of such an approach we get results having strong classical aspects, like the one talking about location of parameters in intervals of stability. A result for nonlinear proper periodic controls, is considered too. (C) 2010 Elsevier B.V. All rights reserved.