Algoritmo Q-learning como estratégia de exploração e/ou explotação para metaheurísticas GRASP e algoritmo genético

Techniques of optimization known as metaheuristics have achieved success in the resolution of many problems classified as NP-Hard. These methods use non deterministic approaches that reach very good solutions which, however, don t guarantee the determination of the global optimum. Beyond the inherent difficulties related to the complexity that characterizes the optimization problems, the metaheuristics still face the dilemma of xploration/exploitation, which consists of choosing between a greedy search and a wider exploration of the solution space. A way to guide such algorithms during the searching of better solutions is supplying them with more knowledge of the problem through the use of a intelligent agent, able to recognize promising regions and also identify when they should diversify the direction of the search. This way, this work proposes the use of Reinforcement Learning technique - Q-learning Algorithm - as exploration/exploitation strategy for the metaheuristics GRASP (Greedy Randomized Adaptive Search Procedure) and Genetic Algorithm. The GRASP metaheuristic uses Q-learning instead of the traditional greedy-random algorithm in the construction phase. This replacement has the purpose of improving the quality of the initial solutions that are used in the local search phase of the GRASP, and also provides for the metaheuristic an adaptive memory mechanism that allows the reuse of good previous decisions and also avoids the repetition of bad decisions. In the Genetic Algorithm, the Q-learning algorithm was used to generate an initial population of high fitness, and after a determined number of generations, where the rate of diversity of the population is less than a certain limit L, it also was applied to supply one of the parents to be used in the genetic crossover operator. Another significant change in the hybrid genetic algorithm is the proposal of a mutually interactive cooperation process between the genetic operators and the Q-learning algorithm. In this interactive/cooperative process, the Q-learning algorithm receives an additional update in the matrix of Q-values based on the current best solution of the Genetic Algorithm. The computational experiments presented in this thesis compares the results obtained with the implementation of traditional versions of GRASP metaheuristic and Genetic Algorithm, with those obtained using the proposed hybrid methods. Both algorithms had been applied successfully to the symmetrical Traveling Salesman Problem, which was modeled as a Markov decision process

Modelo bioeconômico para cálculo de custos e receitas em sistemas de produção de gado de corte visando à obtenção de valores econômicos de características produtivas e reprodutivas

Foi desenvolvido um modelo bioeconômico para o cálculo do desempenho, dos custos e das receitas para obtenção de valores econômicos de características de interesse econômico em sistemas de produção de gado de corte no Brasil. As informações de desempenho e de parâmetros biológicos e econômicos foram obtidas em uma propriedade de gado Nelore que realiza ciclo completo com venda de reprodutores. O modelo é determinístico e estático e foram usadas planilhas Excel para a realização dos cálculos. Com base nas informações originais, foram simulados dois sistemas de produção, um fazendo o ciclo completo com venda de reprodutores (CcoR) e o outro, de cria (Cc). Foram calculados os custos e as receitas para esses dois sistemas e, a partir disto, foram obtidos seus lucros anuais. Para o cálculo dos valores econômicos foram escolhidas as características peso à desmama (PD), peso adulto da vaca (PAV), taxa de prenhez (TP) e taxa de desmama (TD), que são de interesse nos dois sistemas de produção. Para avaliar o impacto das mudanças no desempenho das características sobre o lucro anual do sistema de produção, os valores iniciais das características foram aumentados em 1%. Esse aumento resultou em mudanças positivas no lucro, observando-se que a TD foi a característica que apresentou maior impacto nos dois sistemas. Para o CcoR, os valores econômicos para PD, PAV, TP e TD foram, respectivamente, R$ 0,40/kg, R$ 0,09/kg, R$ 3,20/1% e R$ 10,15/1%. Para Cc, estes valores foram, respectivamente, de R$ 1,31/kg, R$ 0,09/kg, R$ 2,41/1% e R$ 3,36/1%. O modelo foi capaz de reproduzir satisfatoriamente o sistema de produção de gado de corte estudado e pode ser adaptado para outras circunstâncias de produção.

Uma implementação paralela híbrida para o problema do caixeiro viajante usando algoritmos genéticos, GRASP e aprendizagem por reforço

The metaheuristics techiniques are known to solve optimization problems classified as NP-complete and are successful in obtaining good quality solutions. They use non-deterministic approaches to generate solutions that are close to the optimal, without the guarantee of finding the global optimum. Motivated by the difficulties in the resolution of these problems, this work proposes the development of parallel hybrid methods using the reinforcement learning, the metaheuristics GRASP and Genetic Algorithms. With the use of these techniques, we aim to contribute to improved efficiency in obtaining efficient solutions. In this case, instead of using the Q-learning algorithm by reinforcement learning, just as a technique for generating the initial solutions of metaheuristics, we use it in a cooperative and competitive approach with the Genetic Algorithm and GRASP, in an parallel implementation. In this context, was possible to verify that the implementations in this study showed satisfactory results, in both strategies, that is, in cooperation and competition between them and the cooperation and competition between groups. In some instances were found the global optimum, in others theses implementations reach close to it. In this sense was an analyze of the performance for this proposed approach was done and it shows a good performance on the requeriments that prove the efficiency and speedup (gain in speed with the parallel processing) of the implementations performed

Digital monitoring of mycelium growth kinetics and vigor of shiitake (Lentinula edodes (Berk.) Pegler) on agar medium

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

Modelagem e simulação de um separador bifásico com inversão de fases induzida

The present work has the main goal to study the modeling and simulation of a biphasic separator with induced phase inversion, the MDIF, with the utilization of the finite differences method for the resolution of the partial differencial equations which describe the transport of contaminant s mass fraction inside the equipment s settling chamber. With this aim, was developed the deterministic differential model AMADDA, wich was admensionalizated and then semidiscretizated with the method of lines. The integration of the resultant system of ordinary differential equations was realized by means of a modified algorithm of the Adam-Bashfort- Moulton method, and the sthocastic optimization routine of Basin-Hopping was used in the model s parameter estimation procedure . With the aim to establish a comparative referential for the results obtained with the model AMADDA, were used experimental data presented in previous works of the MDIF s research group. The experimental data and those obtained with the model was assessed regarding its normality by means of the Shapiro-Wilk s test, and validated against the experimental results with the Student s t test and the Kruskal-Wallis s test, depending on the result. The results showed satisfactory performance of the model AMADDA in the evaluation of the MDIF s separation efficiency, being possible to determinate that at 1% significance level the calculated results are equivalent to those determinated experimentally in the reference works

The succession dynamics of a macroalgal community after a flood disturbance in a tropical stream from São Paulo State, southeastern Brazil

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

Ondas de spin em quasi-cristais magnônicos

In this paper we investigate the spectra of band structures and transmittance in magnonic quasicrystals that exhibit the so-called deterministic disorders, specifically, magnetic multilayer systems, which are built obeying to the generalized Fibonacci (only golden mean (GM), silver mean (SM), bronze mean (BM), copper mean (CM) and nickel mean (NM) cases) and k-component Fibonacci substitutional sequences. The theoretical model is based on the Heisenberg Hamiltonian in the exchange regime, together with the powerful transfer matrix method, and taking into account the RPA approximation. The magnetic materials considered are simple cubic ferromagnets. Our main interest in this study is to investigate the effects of quasiperiodicity on the physical properties of the systems mentioned by analyzing the behavior of spin wave propagation through the dispersion and transmission spectra of these structures. Among of these results we detach: (i) the fragmentation of the bulk bands, which in the limit of high generations, become a Cantor set, and the presence of the mig-gap frequency in the spin waves transmission, for generalized Fibonacci sequence, and (ii) the strong dependence of the magnonic band gap with respect to the parameters k, which determines the amount of different magnetic materials are present in quasicrystal, and n, which is the generation number of the sequence k-component Fibonacci. In this last case, we have verified that the system presents a magnonic band gap, whose width and frequency region can be controlled by varying k and n. In the exchange regime, the spin waves propagate with frequency of the order of a few tens of terahertz (THz). Therefore, from a experimental and technological point of view, the magnonic quasicrystals can be used as carriers or processors of informations, and the magnon (the quantum spin wave) is responsible for this transport and processing

ALGORITHM FOR PROBABILISTIC ANALYSIS of DISTRIBUTION SYSTEMS WITH DISTRIBUTED GENERATION

In this paper an efficient algorithm for probabilistic analysis of unbalanced three-phase weakly-meshed distribution systems is presented. This algorithm uses the technique of Two-Point Estimate Method for calculating the probabilistic behavior of the system random variables. Additionally, the deterministic analysis of the state variables is performed by means of a Compensation-Based Radial Load Flow (CBRLF). Such load flow efficiently exploits the topological characteristics of the network. To deal with distributed generation, a strategy to incorporate a simplified model of a generator in the CBRLF is proposed. Thus, depending on the type of control and generator operation conditions, the node with distributed generation can be modeled either as a PV or PQ node. To validate the efficiency of the proposed algorithm, the IEEE 37 bus test system is used. The probabilistic results are compared with those obtained using the Monte Carlo method.

Utilizando mapas de conectividade fuzzy no desenvolvimento de algoritmos reparadores de imagens binárias 3D

A 3D binary image is considered well-composed if, and only if, the union of the faces shared by the foreground and background voxels of the image is a surface in R3. Wellcomposed images have some desirable topological properties, which allow us to simplify and optimize algorithms that are widely used in computer graphics, computer vision and image processing. These advantages have fostered the development of algorithms to repair bi-dimensional (2D) and three-dimensional (3D) images that are not well-composed. These algorithms are known as repairing algorithms. In this dissertation, we propose two repairing algorithms, one randomized and one deterministic. Both algorithms are capable of making topological repairs in 3D binary images, producing well-composed images similar to the original images. The key idea behind both algorithms is to iteratively change the assigned color of some points in the input image from 0 (background)to 1 (foreground) until the image becomes well-composed. The points whose colors are changed by the algorithms are chosen according to their values in the fuzzy connectivity map resulting from the image segmentation process. The use of the fuzzy connectivity map ensures that a subset of points chosen by the algorithm at any given iteration is the one with the least affinity with the background among all possible choices

Excitações elementares em super-redes quasiperiódicas com simetria espelho

In this thesis, we investigated the magnonic and photonic structures that exhibit the so-called deterministic disorder. Speci cally, we studied the effects of the quasiperiodicity, associated with an internal structural symmetry, called mirror symmetry, on the spectra of photonics and magnonics multilayer. The quasiperiodicity is introduced when stacked layers following the so-called substitutional sequences. The three sequences used here were the Fibonacci sequence, Thue-Morse and double-period, all with mirror symmetry. Aiming to study the propagation of light waves in multilayer photonic, and spin waves propagation in multilayer magnonic, we use a theoretical model based on transfer matrix treatment. For the propagation of light waves, we present numerical results that show that the quasiperiodicity associated with a mirror symmetry greatly increases the intensity of transmission and the transmission spectra exhibit a pro le self-similar. The return map plotted for this system show that the presence of internal symmetry does not alter the pattern of Fibonacci maps when compared with the case without symmetry. But when comparing the maps of Thue-Morse and double-time sequences with their case without the symmetry mirror, is evident the change in the pro le of the maps. For magnetic multilayers, we work with two di erent systems, multilayer composed of a metamagnetic material and a non-magnetic material, and multilayers composed of two cubic Heisenberg ferromagnets. In the rst case, our calculations are carried out in the magnetostatic regime and calculate the dispersion relation of spin waves for the metamgnetic material considered FeBr2. We show the e ect of mirror symmetry in the spectra of spin waves, and made the analysis of the location of bulk bands and the scaling laws between the full width of the bands allowed and the number of layers of unit cell. Finally, we calculate the transmission spectra of spin waves in quasiperiodic multilayers consisting of Heisenberg ferromagnets. The transmission spectra exhibit self-similar patterns, with regions of scaling well-de ned in frequency and the return maps indicates only dependence of the particular sequence used in the construction of the multilayer

Estudo integrado de análogo a reservatórios petrolíferos fluviais: caracterização, parametrização e modelagem tridimensional de depósitos recentes do Rio Assu (Rio Grande do Norte/Brasil)

This work shows a integrated study of modern analog to fluvial reservoirs of Açu Formation (Unit 3). The modern analog studied has been Assu River located in the same named city, Rio Grande do Norte State, Northeast of Brazil. It has been developed a new methodology to parameterizating the fluvial geological bodies by GPR profile (by central frequency antennas of 50, 100 and 200 MHz). The main parameters obtained were width and thickness. Still in the parameterization, orthophotomaps have been used to calculate the canal sinuosity and braided parameters of Assu River. These information are integrated in a database to supply input data in 3D geological models of fluvial reservoirs. It was made an architectural characterization of the deposit by trench description, GPR profile interpretation and natural expositions study to recognize and describe the facies and its associations, external and internal geometries, boundary surfaces and archtetural elements. Finally, a three-dimensional modeling has been built using all the acquired data already in association with real well data of a reservoir which Rio Assu is considered as analogous. Facies simulations have been used simple kriging (deterministic algorithm), SIS and Boolean (object-based, both stochastics). And, for modeling porosities have used the stochastic algorithm SGS

Sistemas de dunas do parque das Dunas e Barreiras do Inferno/Natal (RN: levantamento geológico /geofísico, elaboração do modelo deterministico e avaliação da vulnerabilidade/suscetibilidade

The current work was developed on the dune systems of the Parque das Dunas and Barreira do Inferno. These places are located in the cities of Natal and Parnamirim (RN, Brazil), respectively. This project has the purpose of developing the deterministic model on a specific blowout at Parque das Dunas, based in the geophysical interpretations of the lines gotten with the Ground Penetration Radar and the planialtimetric acquisitions of the topographical surface of the land. Also analyses of the vulnerability/susceptibility of these dune systems had been done in relation to the human pressures. To develop its deterministic model, it is necessary to acquire inner and outer geometries of the cited blowout. In order to depict inner geometries underneath the surface are used the GPR observing the altimetric control for topographical correction of the GPR lines. As for the outer geometries, the geodesic GPS gives us the planialtimetric points (x, y and z points) with milimetric precision, resulting in high-resolution surfaces. Using interpolation methods of the planialtimetric points was possible create Digital Elevations Models (DEM´s) of these surfaces. As a result, 1,161.4 meters of GPR lines were acquired on the blowout at the Parque das Dunas and 3,735.27 meters on the blowout at the Barreira do Inferno. These lines had been acquired with a 200 MHz antenna, except the 7 and 8 lines, for which we had been used a 100 MHz antenna. The gotten data had been processed and interpreted, being possible to identify boundary surfaces of first, second and third order. The first order boundary surface is related with the contact of the rocks of the Barreiras Group with the aeolian deposits. These deposits had been divided in two groups (Group 1 and Group 2) which are related with the geometry of stratum and the dip of its stratifications. Group 1 presented stratum of sigmoidal and irregular geometries and involved bodies where the reflectors had presented dips that had varied of 20 to the 28 degrees for the Parque das Dunas blowout and of 22 to the 29 degrees for the Barreira do Inferno blowout. Usually, it was limited in the base for the first order surface and in the top for the second order surface. Group 2 presented stratum of trough, wedge or lens geometries, limited in the base for the second order vi surface, where the corresponding deposits had more shown smoothed reflectors or with dips of low angle. The Deterministic and Digital Elevation Models had been developed from the integration and interpretation of the 2D data with the GOCAD® program. In Digital Elevations Models it was possible to see, for the localities, corridor or trough-shaped blowouts. In Deterministic Model it was possible to see first and second order boundary surfaces. For the vulnerability/susceptibility of the dune systems it was applied the methodology proposal by Boderè al (1991); however the same one did not show adequate because it evaluates actual coastal dunes. Actual coastal dunes are dunes that are presented in balance with the current environmental conditions. Therefore, a new methodology was proposal which characterizes the supplying and activity sedimentary, as well as the human pressures. For the methodology developed in this work, both the localities had presented a good management. The Parque das Dunas was characterized as a relic dune system and the Barreira do Inferno was characterized as a palimpsestic dune system. Also two Thematic Maps had been elaborated for the environmental characterization of the studied dune systems, with software ArcGis 8.3, and its respective data bases

Análise de barreiras e o modelo de ressonância funcional de acidentes de Erik Hollnagel

Partindo de apresentação da proposta de análise de acidentes baseada no modelo sistêmico e não linear de acidentes, dito de ressonância funcional, desenvolvido por Erik Hollnagel, este texto discute críticas às concepções deterministas de acidente e aos modelos lineares de representação desses eventos. Aspectos do modelo de ressonância funcional são usados como eixo definidor de proposta de agenda de discussões a ser desenvolvida pelos interessados no tema da análise de acidentes, destacando-se: a) a defesa do abandono da idéia de causas de acidentes e sua substituição pela de explicação desses eventos; b) apresentação dos conceitos de variabilidade de desempenhos e adaptações locais e a proposta de sua utilização como bases para a identificação de perfis ou aspectos típicos de acidentes. Por fim, discute-se a necessidade de incorporação dos conceitos apresentados no debate sobre novos caminhos para as práticas de análises e de prevenção de acidentes em nosso país.

Indeterminismo e liberdade em Hume

Embora Hume eventualmente pareça supor que os processos naturais ocorrem de modo determinístico, sua concepção da causalidade proíbe que os conheçamos enquanto tais, pois, para ele, toda idéia de conexão necessária entre eventos é de natureza psicológica. em nossa análise dessa questão, argumentamos que: a) nas suas principais passagens sobre a causalidade, Hume não assume um compromisso forte com o determinismo; b) certas partes de seus textos sugerem que o indeterminismo é ineliminável, no conhecimento da natureza; e c) sua concepção causal da liberdade supõe o indeterminismo dos processos naturais, e não apenas o indeterminismo em nível de seu conhecimento.