Spelling suggestions: "subject:"algoritmo nen?tica"" "subject:"algoritmo been?tica""
1 |
Otimiza??o global para resolver problemas inversos em eletrorresistividade com flexibilidade na escolha dos v?nculosBarboza, Francisco M?rcio 28 November 2017 (has links)
Submitted by Automa??o e Estat?stica (sst@bczm.ufrn.br) on 2018-01-23T14:44:54Z
No. of bitstreams: 1
FranciscoMarcioBarboza_TESE.pdf: 8557102 bytes, checksum: 12d460591e55e242af090bdb4f4cfbb9 (MD5) / Approved for entry into archive by Arlan Eloi Leite Silva (eloihistoriador@yahoo.com.br) on 2018-01-25T13:26:13Z (GMT) No. of bitstreams: 1
FranciscoMarcioBarboza_TESE.pdf: 8557102 bytes, checksum: 12d460591e55e242af090bdb4f4cfbb9 (MD5) / Made available in DSpace on 2018-01-25T13:26:13Z (GMT). No. of bitstreams: 1
FranciscoMarcioBarboza_TESE.pdf: 8557102 bytes, checksum: 12d460591e55e242af090bdb4f4cfbb9 (MD5)
Previous issue date: 2017-11-28 / Conselho Nacional de Desenvolvimento Cient?fico e Tecnol?gico (CNPq) / Invers?o em eletrorresistividade ? um problema inverso mal posto, porque diferentes
realiza??es de um mesmo modelo podem satisfazer aproximadamente o mesmo crit?rio
de ajuste. Se faz necess?rio portanto o uso de v?nculos para obter solu??es ?nicas e/ou
est?veis ? pequenas perturba??es nas medidas. Contudo, em geral, a introdu??o de v?nculos
tem ficado restrita aos casos de v?nculos diferenci?veis e que podem ser tratados com
algoritmos de otimiza??o local. A modelagem direta 1D e 2D em resistividade DC ? computacionalmente
barata, permitindo o uso de m?todos de otimiza??o global (GOMs) para
resolver problemas inversos 1.5D e 2D com flexibilidade na incorpora??o de v?nculos.
As modifica??es da fun??o de custo, seja na mudan?a de v?nculos ou no crit?rio de ajuste
de dados, podem ser realizadas com facilidade, j? que cada termo da fun??o de custo
? devidamente normalizado para permitir a invari?ncia aproximada dos multiplicadores
Lagrange. Os GOMs t?m potencial para suportar um ambiente computacional adequado
para interpreta??o quantitativa em que a compara??o de solu??es que incorporam diferentes
restri??es ? uma maneira de inferir caracter?sticas da distribui??o real da resistividade
subterr?nea. Neste trabalho foram desenvolvidas: (i) Compara??o das performances dos
m?todos Simulated Annealing (SA), Algoritmo Gen?tico (GA) e Particle Swarm Otmization
(PSO) para resolver o problema inverso 1.5D na resistividade DC usando dados
sint?ticos e de campo; (ii) Apresenta??o de uma abordagem de invers?o baseada no Particle
Swarm Optimization (PSO) para os dados 2D de resistividade de corrente cont?nua
(DC); (iii) Explora??o de v?rios v?nculos na varia??o de log da resistividade: continuidade
espacial tanto nas normas L1 quanto L2, incluindo o caso de restri??o de varia??o
apenas na dire??o horizontal, varia??o total e v?nculos de esparsidade usando transformada
discreta do cosseno e bases de Daubechies. Al?m disso, exploramos o v?nculo de
m?nimo momento de in?rcia, incluindo o caso de usar a superf?cie da Terra como eixo
alvo, para impor a concentra??o de materiais resistivos ou condutores ao longo dos eixos
alvo. Os principais resultados da compara??o para o case 1.5D s?o: a) todos os m?todos
reproduzem bastante a distribui??o de resistividade de modelos sint?ticos, b) PSO e GA
s?o muito robustos para mudan?as na fun??o de custo e SA ? comparativamente muito
mais sens?vel, c) primeiro PSO e GA segundo apresentam o melhores desempenhos computacionais,
exigindo um menor n?mero de modelos de encaminhamento do que SA, e d)
GA mostra o melhor desempenho em rela??o ao valor final alcan?ado da fun??o de custo
e seu desvio padr?o, enquanto a SA tem o pior desempenho neste aspecto. Igualmente
importante para ambos os casos 1.5D e 2D, a partir dos crit?rios de parada do algoritmo
PSO resulta n?o apenas a melhor solu??o, mas tamb?m um conjunto de quase-solu??es
sub-?timas a partir dos quais as an?lises de incerteza podem ser realizadas. Como resultado,
o int?rprete tem liberdade para realizar um processo de interpreta??o quantitativa com base em uma abordagem de invers?o de julgamento e erro, de forma semelhante, ele
tem ao usar um software de modelagem avan?ado amig?vel, sendo capaz de conduzir a
solu??o para incorporar suas concep??es sobre o ambiente geol?gico, al?m de avaliar o
ajuste de dados e a estabilidade das solu??es obtidas. Apresentamos exemplos de dados
sint?ticos e de campo para ambos os casos de invers?o. / Inversion in DC-resistivity is an ill-posed inverse problem because different realizations
of the same model might satisfy approximately the same data fitting criterium. It is
therefore necessary to use constraints to obtain unique and / or stable solutions to small
perturbations in the measurements. However, in general, the introduction of constraints
has been restricted to cases of differentiable constraints, which can be treated with local
optimization algorithms. 1D and 2D modeling in DC-resistivity is computationally inexpensive,
allowing the use of global optimization methods (GOMs) to solve 1.5D and 2D
inverse problems with flexibility in constraint incorporation. Changes in the cost function,
either in the constraints or data fitting criteria, can be easily performed, since each term
of the cost function is properly normalized to allow the approximate invariance of the
Lagrange multipliers. GOMs have the potential to support a computational environment
suitable for quantitative interpretation in which the comparison of solutions incorporating
different constraints is one way of inferring characteristics of the actual distribution of the
underground resistivity. In this work, we developed: (i) comparison of the performances
of the Simulated Annealing (SA), Genetic Algorithm (GA) and Particle Swarm Optimization
(PSO) methods to solve the 1.5D inverse problem in DC resistivity using synthetic
and field data; (ii) an inversion approach based on particle swarm optimization (PSO) to
solve the 2D DC-resistivity inverse problem; (iii) exploration of several constraints in the
variation of log-resistivity, including spatial continuity in both L1 andL2 norms, total variation
and sparsity constraints using discrete cosine and Daubechies bases. In addition,
we explore the minimum inertia constraint, including the case of using the Earth?s surface
as the target axis, to impose the concentration of resistive or conductive materials along
target axes. The main results of the comparison for the 1.5D case are: a) all methods
reproduce quite well the resistivity distribution of synthetic models, b) PSO and GA are
very robust to changes in the cost function and SA is comparatively much more sensitive,
c) PSO first and GA second present the best computational performances, requiring smaller
number of forwarding modeling than SA, and d) GA shows the best performance with
respect to the final attained value of the cost function and its standard deviation, whilst
SA has the worst performance in this aspect. Equally important for both 1.5 and 2D
cases, from the stopping criteria of the PSO algorithm results not only the best solution
but also a cluster of suboptimal quasi-solutions from which uncertainty analyses can be
performed. As a result, the interpreter has freedom to perform a quantitative interpretation
process based on a feedback trial-and-error inversion approach, in a similar manner
he/she has when using a friendly forward modeling software, being capable of driving
the solution to incorporate his/her conceptions about the geologic environment, besides
appraising data fitting and stability of the obtained solutions. We present both synthetic and field data examples for all inversion cases.
|
2 |
Proposta de implementa??o paralela de algoritmo gen?tico em FPGATorquato, Matheus Fernandes 01 December 2017 (has links)
Submitted by Automa??o e Estat?stica (sst@bczm.ufrn.br) on 2018-02-15T11:21:03Z
No. of bitstreams: 1
MatheusFernandesTorquato_DISSERT.pdf: 2354339 bytes, checksum: 32d0b6ee431808fccf375481ef94b8a9 (MD5) / Approved for entry into archive by Arlan Eloi Leite Silva (eloihistoriador@yahoo.com.br) on 2018-02-16T11:39:52Z (GMT) No. of bitstreams: 1
MatheusFernandesTorquato_DISSERT.pdf: 2354339 bytes, checksum: 32d0b6ee431808fccf375481ef94b8a9 (MD5) / Made available in DSpace on 2018-02-16T11:39:52Z (GMT). No. of bitstreams: 1
MatheusFernandesTorquato_DISSERT.pdf: 2354339 bytes, checksum: 32d0b6ee431808fccf375481ef94b8a9 (MD5)
Previous issue date: 2017-12-01 / Coordena??o de Aperfei?oamento de Pessoal de N?vel Superior (CAPES) / Os Algoritmos Gen?ticos (AGs) s?o utilizados para resolver problemas de busca e
otimiza??o no qual, uma solu??o ?tima pode ser encontrada utilizando um processo iterativo
e transi??es probabil?sticas. Todavia, dependendo do tipo de problema, o tempo
para encontrar a solu??o pode ser elevado em m?quinas sequenciais devido ? complexidade
computacional do algoritmo gen?tico. Assim, esse trabalho possui como objetivo
o desenvolvimento de um prot?tipo associado a uma implementa??o paralela de um algoritmo
gen?tico em FPGA (Field-programmable gate array). O principal objetivo do
desenvolvimento dessa arquitetura ? a otimiza??o do tempo de processamento do sistema.
Resultados associados com o tempo de processamento e a ?rea ocupada para v?rios
tamanhos de popula??o foram analisados. Estudos relativos ? precis?o da resposta do
algoritmo gen?tico para o problema de otimiza??o de fun??es com uma e duas vari?veis
tamb?m foram analisados para a implementa??o em hardware. Todo projeto foi desenvolvido
utilizando a plataforma de desenvolvimento System Generator da Xilinx tendo como
FPGA alvo um Virtex-7 xc7vx550t-1ffg1158 FPGA. / Genetic Algorithms (GAs) are used to solve search and optimization problems in
which an optimal solution can be found using an iterative process and using probabilistic
transitions. However, depending on the type of problem, the time required to find a solution
can be high in sequential machines due to the computational complexity of genetic
algorithm. This work proposes a parallel implementation of a genetic algorithm on fieldprogrammable
gate array (FPGA). Optimization of the system?s processing time is the
main goal of this project. Results associated with the processing time and area occupancy
(in FPGA) for various population size are analyzed. Studies concerning the accuracy of
the GA response for the optimization of functions with one and two variables were also
analyzed for the hardware implementation. The project was developed using the System
Generator software (Xilinx development platform) and the Virtex-7 xc7vx550t-1ffg1158
FPGA.
|
3 |
Estudo de clusters met?licos de alum?nio-s?dio, alum?niopot?ssio, alum?nio-l?tio e s?dio-l?tio pelas abordagens de algoritmos gen?ticos, c?lculos qu?nticos e an?lise topol?gicaSantos, Acassio Rocha 21 February 2017 (has links)
Submitted by Automa??o e Estat?stica (sst@bczm.ufrn.br) on 2017-07-03T12:42:00Z
No. of bitstreams: 1
AcassioRochaSantos_DISSERT.pdf: 9390259 bytes, checksum: d25d8e92963faba78cb36774741179e4 (MD5) / Approved for entry into archive by Arlan Eloi Leite Silva (eloihistoriador@yahoo.com.br) on 2017-07-06T11:41:22Z (GMT) No. of bitstreams: 1
AcassioRochaSantos_DISSERT.pdf: 9390259 bytes, checksum: d25d8e92963faba78cb36774741179e4 (MD5) / Made available in DSpace on 2017-07-06T11:41:22Z (GMT). No. of bitstreams: 1
AcassioRochaSantos_DISSERT.pdf: 9390259 bytes, checksum: d25d8e92963faba78cb36774741179e4 (MD5)
Previous issue date: 2017-02-21 / O estudo te?rico de clusters met?licos tem despertado um interesse consider?vel,
devido ? possibilidade de criar novas ligas de materiais em nanoescala, as chamadas
"nanoligas". Pesquisas sobre nanoligas desempenham papel significativo na Ci?ncia de
Materiais, pois, entre seus objetivos mais importantes, est?o o de prever a estabilidade das
estruturas, seus modos de crescimento, bem como o de auxiliar a interpreta??o de medidas
espectrosc?picas e outras medi??es experimentais. Nesse contexto, um grande n?mero de
m?todos foi relatado nos ?ltimos anos para a otimiza??o do m?nimo global de grupos
at?micos e moleculares, sendo um dos mais utilizados atualmente o do Algoritmo Gen?tico
(doravante, GA), o qual baseia-se em princ?pios relacionados a processos evolutivos, em
operadores inspirados na Teoria da Evolu??o e na Gen?tica, isto ?, na recombina??o, muta??o
e sele??o natural. Particularmente, o GA com a implementa??o do potencial Gupta tem se
mostrado eficiente na busca de solu??es ??timas? em problemas de otimiza??o de clusters
met?licos. Esta disserta??o ? composta por cap?tulos de introdu??o, de metodologia, de
abordagem te?rica (Cap. 1, 2 e 3); e tamb?m por cap?tulos que cont?m artigos sobre o tema
proposto (Cap. 4, 5 e 6). No primeiro artigo (Cap. 4), analisaram-se clusters bimet?licos
AlxNay (x+y?55) por meio da aplica??o do GA com a implementa??o do potencial Gupta.
Com base tamb?m na aplica??o do GA, no segundo cap?tulo (Cap. 5) foram estudados
clusters de AlxLiy e AlxKy (x+y ? 55). Em ambos os trabalhos, para elevar a efici?ncia do GA,
introduziu-se mais dois operadores: o Aniquilador e o Hist?ria. Ao serem comparadas as
estruturas obtidas por meio do GA com potencial Gupta para clusters de alum?nio puro, l?tio
puro e alum?nio-l?tio com resultados recentes da literatura, verificou-se que para os sistemas
Al2, Al3, Al6, Al8, Al9, Li5, Li6, Li7, Al1Li5, Al1Li7 e Al1Li8 as geometrias obtidas foram muito
semelhantes ?quelas resultantes de c?lculos de funcional de densidade e ab initio[como
CCSD(T)]. No terceiro artigo (Cap. 6), analisou-se um novo algoritmo gen?tico qu?ntico (Q-GA)
para pequenos sistemas de clusters NaxLiy com (x+y ? 10). Constatou-se que o Q-GA
apresenta maior efici?ncia na busca do m?nimo global em rela??o ao GA com o potencial
Gupta. Isso porque o primeiro utiliza m?todo qu?ntico, enquanto o segundo usa um m?todo
cl?ssico. Por ser mais preciso, o Q-GA possui uma abrang?ncia menor. Neste artigo, al?m de
c?lculos ab inito, tamb?m foram realizados c?lculos topol?gicos a partir da Teoria Qu?ntica
de ?tomos em Mol?culas (QTAIM) para as estruturas Na1Li5, Na2Li4, Na3Li3, Na4Li2 e
Na5Li1, obtidas pelo Q-GA. Nessas estruturas, chama a aten??o o fato de n?o haver caminho
de liga??o envolvendo diretamente os metais, sendo unidos por pseudo?tomos, com exce??o
do Na5Li1. Algumas intera??es at?micas n?o foram indicadas pelo caminho de liga??o e sua
an?lise foi feita pelo ?ndice de deslocaliza??o (DI). No sistema Na1Li5, os pares at?micos
Na1-Li2 e Na1-Li6 t?m as intera??es mais fortes (e equivalentes ? do sistema NaLi) de todos
os pares Na-Li de todos clusters NaxLiy(x+y=6); ao mesmo tempo, os outros pares Na-Li t?m
intera??es dez vezes mais fracas do que aquelas do sistema NaLi. As intera??es Na-Na dos
clusters Na4Li2 e Na5Li1 s?o as mais fortes quando comparadas com sistemas puros. Por fim,
verificou-se que a f?rmula do grau de degeneresc?ncia do ?ndice de aromaticidade D3BIA e a
carga at?mica indicaram que os ?tomos de l?tio mais pr?ximo ao ?tomo de s?dio transferem
carga para esse ?ltimo. / The theoretical study of metal clusters has drawn considerable interest due to the
possibility of creating new alloys from materials in nanoscale, the so-called "nanoalloys".
Research on nanoalloys has had an important role in materials science, since, among some of
its most relevant objectives, we may find the prediction of stability in structures, their
manners of growth and further assistance in the interpretation of spectroscopic and other
experimental measures. In this context, several methods have been reported in the last few
years towards the global minimum optimization of atomic and molecular groups, where the
Genetic Algorithm (henceforth GA) is currently considered one of the most used methods,
whilst based on principles related to evolutionary processes as well as operators inspired by
the Theory of Evolution and Genetics, i. e., by recombination, mutation and natural selection.
The GA method in particular, and altogether with the implementation of the Gupta potential,
has become efficient in the search for ?optimal? solutions for optimization problems in
metallic clusters. The present dissertation is composed of chapters consisting of introduction,
methodology and theoretical considerations (Chap. 1, 2 and 3), as well as of chapters
containing articles on the proposed subject (Chap. 4, 5 and 6). In the first article (Chap. 4), we
may find the analysis of AlxNay (x + y ? 55) bimetallic clusters through the Genetic
Algorithm method with the implementation of the Gupta potential. Also based on the GA
application, in the following chapter (Chap. 5) we may find a study regarding AlxLiy e AlxKy
(x+y ? 55) clusters. In both works, in order to improve GA efficiency, two additional
operators have been introduced: Annihilator and History. By being compared to structures
obtained by means of GA with Gupta potential for pure aluminum, pure lithium and
aluminum-lithium clusters in recent results from literature, it has been verified that, regarding
systems Al2, Al3, Al6, Al8, Al9, Li5, Li6, Li7, Al1Li5, Al1Li7 e Al1Li8, the obtained geometries
were very similar to those resulting from density functional and ab initio calculations [such as
CCSD(T)]. In the third chapter (Chap. 6), we analyzed a new quantum genetic algorithm (QGA)
for small cluster systems NaxLiy with (x+y ? 10). It has been observed that Q-GA
presents an improved efficiency towards a global minimum regarding the GA with the Gupta
potential. That has been the case since the former uses the quantum method, while the latter
uses a classic method. More specifically, the Q-GA has a narrower scope. In this article,
besides ab initio calculations, topological calculations were performed as well, grounded on
the Quantum Theory of Atoms in Molecules (QTAIM) for the structures Na1Li5, Na2Li4,
Na3Li3, Na4Li2 e Na5Li1 obtained by the Q-GA. In these structures, it is evident that there is no
bonding path between the metals, since they are bonded by pseudo atoms, with the exception
of the Na5Li1. Some of the atomic interactions have not been suggested by the bonding path,
being their analysis performed according to the delocalization index (DI). In the Na5Li1
system, the atomic pairs Na1-Li2 and Na1-Li6 have the strongest interactions (equivalent to
the NaLi system) of all Na-Li pairs in all of the NaxLiy (x+y=6) clusters; concurrently, other
Na-Li pairs bear interactions ten times weaker than those from the NaLi system. The Na-Na
interactions from the clusters Na4Li2 e Na5Li1 are stronger when compared to pure systems.
Finally, it has been verified that the degree of degeneracy formula of the aromaticity index
D3BIA and the atomic charge suggest that the lithium atoms that are closer to the sodium
atom transfer charge to the latter.
|
4 |
Algoritmos gen?ticos: uso de l?gica nebulosa e an?lise de converg?ncia por cadeia de MarkovCarlos, Luiz Amorim 05 November 2013 (has links)
Made available in DSpace on 2014-12-17T14:55:16Z (GMT). No. of bitstreams: 1
LuizAC_TESE.pdf: 1792266 bytes, checksum: ec666aacc6a6bc6fc6e042bf7f0ce339 (MD5)
Previous issue date: 2013-11-05 / In this work, the Markov chain will be the tool used in the modeling and analysis of
convergence of the genetic algorithm, both the standard version as for the other versions
that allows the genetic algorithm. In addition, we intend to compare the performance of
the standard version with the fuzzy version, believing that this version gives the genetic algorithm
a great ability to find a global optimum, own the global optimization algorithms.
The choice of this algorithm is due to the fact that it has become, over the past thirty yares,
one of the more importan tool used to find a solution of de optimization problem. This
choice is due to its effectiveness in finding a good quality solution to the problem, considering
that the knowledge of a good quality solution becomes acceptable given that there
may not be another algorithm able to get the optimal solution for many of these problems.
However, this algorithm can be set, taking into account, that it is not only dependent on
how the problem is represented as but also some of the operators are defined, to the standard
version of this, when the parameters are kept fixed, to their versions with variables
parameters. Therefore to achieve good performance with the aforementioned algorithm
is necessary that it has an adequate criterion in the choice of its parameters, especially the
rate of mutation and crossover rate or even the size of the population. It is important to
remember that those implementations in which parameters are kept fixed throughout the
execution, the modeling algorithm by Markov chain results in a homogeneous chain and
when it allows the variation of parameters during the execution, the Markov chain that
models becomes be non - homogeneous. Therefore, in an attempt to improve the algorithm
performance, few studies have tried to make the setting of the parameters through
strategies that capture the intrinsic characteristics of the problem. These characteristics
are extracted from the present state of execution, in order to identify and preserve a pattern
related to a solution of good quality and at the same time that standard discarding of
low quality. Strategies for feature extraction can either use precise techniques as fuzzy
techniques, in the latter case being made through a fuzzy controller. A Markov chain is
used for modeling and convergence analysis of the algorithm, both in its standard version
as for the other. In order to evaluate the performance of a non-homogeneous algorithm
tests will be applied to compare the standard fuzzy algorithm with the genetic algorithm,
and the rate of change adjusted by a fuzzy controller. To do so, pick up optimization
problems whose number of solutions varies exponentially with the number of variables / Neste trabalho, a cadeia de Markov ser? a ferramenta usada na modelagem e na an?lise
de converg?ncia do algoritmo gen?tico, tanto para sua vers?o padr?o quanto para as
demais vers?es que o algoritmo gen?tico permite. Al?m disso, pretende-se comparar o
desempenho da vers?o padr?o com a vers?o nebulosa, por acreditar que esta vers?o d?
ao algoritmo gen?tico uma grande capacidade para encontrar um ?timo global, pr?prio
dos algoritmos de otimiza??o global. A escolha deste algoritmo deve-se tamb?m ao fato
do mesmo ter se tornado, nos ?ltimos anos, uma das ferramentas mais usadas para achar
uma solu??o do problema de otimiza??o. Esta escolha deve-se ? sua comprovada efic?cia
na busca de uma solu??o de boa qualidade para o problema, considerando que o
conhecimento de uma solu??o de boa qualidade torna-se aceit?vel tendo em vista que
pode n?o existir um outro algorimo capaz de obter a solu??o ?tima, para muitos desses
problemas. Entretanto, esse algoritmo pode ser definido, levando em conta que o mesmo
? dependente n?o apenas da forma como o problema ? representado, mas tamb?m como
s?o definidos alguns dos operadores, desde sua vers?o padr?o, quando os par?metros s?o
mantidos fixos, at? suas vers?es com par?metros vari?veis. Por isso, para se alcan?ar
um bom desempenho com o aludido algoritmo ? necess?rio que o mesmo tenha um adequado
crit?rio na escolha de seus par?metros, principalmente da taxa de muta??o e da
taxa de cruzamento ou, at? mesmo, do tamanho da popula??o. ? importante lembrar que
as implementa??es em que par?metros s?o mantidos fixos durante toda a execu??o, a modelagem
do algoritmo por cadeia de Markov resulta numa cadeia homog?nea, e quando
permite a varia??o de par?metros ao longo da execu??o, a cadeia de Markov que o modela
passa a ser do tipo n?o-homog?nea. Portanto, na tentativa de melhorar o desempenho
do algoritmo, alguns trabalhos t?m procurado realizar o ajuste dos par?metros atrav?s de
estrat?gias que captem caracter?sticas intr?nsecas ao problema. Essas caracter?sticas s?o
extra?das do estado presente de execu??o, com o fim de identificar e preservar algum padr?o
relacionado a uma solu??o de boa qualidade e, ao mesmo tempo, descartando aquele
padr?o de baixa qualidade. As estrat?gias de extra??o das caracter?sticas tanto podem usar
t?cnicas precisas quanto t?cnicas nebulosas, sendo neste ?ltimo caso feita atrav?s de um
controlador nebuloso. Com o fim de avaliar empiriccamente o desempenho de um algoritmo
n?o-homog?neo, apresenta-se testes onde se compara o algoritmo gen?tico padr?o
com o algoritmo gen?tico nebuloso, sendo a taxa de muta??o ajustada por um controlador
nebuloso. Para isso, escolhe-se problemas de otimiza??o cujo n?mero de solu??es varia
exponencialmente com o n?mero de vari?veis
|
5 |
Dilema da diversidade-acur?cia: um estudo emp?rico no contexto de multiclassificadoresOliveira, Diogo Fagundes de 01 September 2008 (has links)
Made available in DSpace on 2014-12-17T15:47:49Z (GMT). No. of bitstreams: 1
DiogoFO.pdf: 866073 bytes, checksum: bf59c2597aef9b7382b7e14bd4914265 (MD5)
Previous issue date: 2008-09-01 / Conselho Nacional de Desenvolvimento Cient?fico e Tecnol?gico / Multi-classifier systems, also known as ensembles, have been widely used to solve several problems, because they, often, present better performance than the individual classifiers that form these systems. But, in order to do so, it s necessary that the base classifiers to be as accurate as diverse among themselves this is also known as diversity/accuracy dilemma. Given its importance, some works have investigate the ensembles behavior in
context of this dilemma. However, the majority of them address homogenous ensemble, i.e., ensembles composed only of the same type of classifiers. Thus, motivated by this limitation, this thesis, using genetic algorithms, performs a detailed study on the dilemma diversity/accuracy for heterogeneous ensembles / Sistemas Multiclassificadores, tamb?m conhecidos como comit?s de classificadores, t?m sido amplamente utilizados para resolver os mais variados problemas, pois em geral t?m
melhores desempenhos que os classificadores base que formam esses sistemas. Para que isso ocorra, por?m, ? necess?rio que os classificadores base sejam t?o acurados quanto diversos entre si isso ? conhecido como dilema da diversidade-acur?cia. Dado a sua import?ncia, alguns trabalhos sobre o estudo do omportamento dos comit?s no contexto desse dilema foram propostos. Entretanto, a maioria dos trabalhos estudou tal problema para comit?s homog?neos, ou seja, comit?s formados apenas por classificadores do mesmo tipo. Sendo assim, motivado por esta limita??o, esta disserta??o, usando algoritmos gen?ticos, efetua um estudo mais detalhado sobre o dilema da diversidade-acur?cia em comit?s heterog?neos
|
6 |
Especifica??o e implementa??o de um algoritmo gen?tico para otimiza??o de projetos de ilumina??o p?blica / Specification and implementation of a genetic algorithm for optimization of public illumination projectsOliveira, R?mulo Alves de 27 January 2015 (has links)
Submitted by Automa??o e Estat?stica (sst@bczm.ufrn.br) on 2016-05-30T23:14:12Z
No. of bitstreams: 1
RomuloAlvesDeOliveira_TESE.pdf: 1869244 bytes, checksum: af0453c8083aee85607b79c3a17c1d4c (MD5) / Approved for entry into archive by Arlan Eloi Leite Silva (eloihistoriador@yahoo.com.br) on 2016-06-01T22:30:48Z (GMT) No. of bitstreams: 1
RomuloAlvesDeOliveira_TESE.pdf: 1869244 bytes, checksum: af0453c8083aee85607b79c3a17c1d4c (MD5) / Made available in DSpace on 2016-06-01T22:30:48Z (GMT). No. of bitstreams: 1
RomuloAlvesDeOliveira_TESE.pdf: 1869244 bytes, checksum: af0453c8083aee85607b79c3a17c1d4c (MD5)
Previous issue date: 2015-01-27 / Atualmente os projetos de Ilumina??o P?blica (IP), ou seja, ruas, avenidas, pra?as, estacionamentos e similares s?o realizados com a utiliza??o de softwares comerciais ou livres, em geral, fornecidos por fabricantes ou grupos de fabricantes de produtos de ilumina??o, aplicando o M?todo Ponto a Ponto para o c?lculo dos n?veis de ilumina??o. Outros pontos em comum s?o: a falta de preocupa??o na redu??o dos custos dos projetos e a dificuldade em modificar as estruturas utilizadas, tais como: localiza??o e altura dos postes e lumin?rias, ?ngulo de inclina??o das lumin?rias, quantidade de lumin?rias por poste, entre outros. Qualquer altera??o nas estruturas ter? que ser feita manualmente, geralmente em um ambiente CAD, para depois obter os novos resultados e comparar com os anteriores. Para auxiliar nessa tarefa, ? proposta aqui a utiliza??o da Metaheur?stica Col?nia de Formigas, onde os par?metros e localiza??o das estruturas passam a ser definidos automaticamente, de forma a atender os n?veis de ilumina??o estabelecidos nas normas t?cnicas, al?m de otimizar o custo de material por unidade de ?rea. / The development of public lighting projects in Brazil must meet the standards established in Brazilian standards. Many of these projects is developed through the use of knowledge about "practical rules" practiced by the designers of this area. In some cases are also used computational tools offered, generally, by leading manufacturers of lamps/luminaires. These tools have served only as calculation tools, with some limitations, such as: are not able to verify compliance or not the parameters established by Brazilian standards, most of the luminaires offered in your database are not sold in Brazil, not have no concern about the analysis of the implementation costs of elaborate designs and, finally, present an enormous difficulty in performing tests on a large volume of possible projects. It is the goal of this thesis to develop a methodology and a computational tool for the development of public lighting projects based on genetic algorithm techniques that not only perform the calculations of these projects, but can also test several possible projects using in your database the luminaires marketed in Brazil, providing the user, as a solution, a set of projects that meet the Brazilian standards and classified according the implementation costs of each project. To adjust the proposed algorithm the following performance parameters were modified: number of individuals in the initial population; probability of achievement of the cross-over; probability of achievement of the mutation. A comparison of this method with the projects developed with the use of "practical rules" is performed for various types of existing roads. The results obtained using the proposed methodology and the developed computational tool show that the methodology, including the adjustments in performance parameters, is able to meet the objectives of the work.
|
7 |
M?todo h?brido para projeto de superf?cies seletivas em frequ?nciaAra?jo, Gilmara Linhares de 10 December 2015 (has links)
Submitted by Automa??o e Estat?stica (sst@bczm.ufrn.br) on 2016-08-09T22:51:26Z
No. of bitstreams: 1
GilmaraLinharesDeAraujo_TESE.pdf: 11647948 bytes, checksum: 94a5e966f18f8dd6bfc6646d2519542b (MD5) / Approved for entry into archive by Arlan Eloi Leite Silva (eloihistoriador@yahoo.com.br) on 2016-08-10T23:44:37Z (GMT) No. of bitstreams: 1
GilmaraLinharesDeAraujo_TESE.pdf: 11647948 bytes, checksum: 94a5e966f18f8dd6bfc6646d2519542b (MD5) / Made available in DSpace on 2016-08-10T23:44:37Z (GMT). No. of bitstreams: 1
GilmaraLinharesDeAraujo_TESE.pdf: 11647948 bytes, checksum: 94a5e966f18f8dd6bfc6646d2519542b (MD5)
Previous issue date: 2015-12-10 / Coordena??o de Aperfei?oamento de Pessoal de N?vel Superior (CAPES) / Esse trabalho apresenta uma t?cnica h?brida de projeto de superf?cies seletivas em frequ?ncia, sobre uma camada diel?trica isotr?pica, considerando diversas geometrias para os elementos das c?lulas unit?rias. Especificamente, a t?cnica h?brida usa o m?todo do circuito equivalente em conjunto com algoritmos gen?ticos, visando a s?ntese de estruturas com resposta singleband e dual-band. O m?todo do circuito equivalente permite modelar a estrutura por meio de um circuito equivalente e tamb?m a obten??o de circuitos para diferentes geometrias. A partir da obten??o dos par?metros desses circuitos, podem-se obter as caracter?sticas de transmiss?o e de reflex?o das estruturas modeladas. Para obter a otimiza??o das estruturas modeladas, de acordo com a resposta em frequ?ncia desejada, a ferramenta de otimiza??o do Matlab optimtool mostrou-se de f?cil utiliza??o, permitindo explorar resultados importantes na an?lise de otimiza??o. No trabalho, s?o apresentados resultados num?ricos e experimentais para as caracter?sticas de transmiss?o de diferentes geometrias analisadas. Foram efetuadas compara??es com resultados apresentados na literatura, tendo-se observado uma boaconcord?ncia nos casos analisados para estruturas com substratos isotr?picos. / This thesis presents a hybrid technique of frequency selective surfaces project
(FSS) on a isotropic dielectric layer, considering various geometries for the elements of
the unit cell. Specifically, the hybrid technique uses the equivalent circuit method in
conjunction with genetic algorithm, aiming at the synthesis of structures with response
single-band and dual-band. The equivalent circuit method allows you to model the
structure by using an equivalent circuit and also obtaining circuits for different
geometries. From the obtaining of the parameters of these circuits, you can get the
transmission and reflection characteristics of patterned structures. For the optimization
of patterned structures, according to the desired frequency response, Matlab?
optimization tool named optimtool proved to be easy to use, allowing you to explore
important results on the optimization analysis. In this thesis, numeric and experimental
results are presented for the different characteristics of the analyzed geometries. For
this, it was determined a technique to obtain the parameter N, which is based on genetic
algorithms and differential geometry, to obtain the algebraic rational models that
determine values of N more accurate, facilitating new projects of FSS with these
geometries. The optimal results of N are grouped according to the occupancy factor of
the cell and the thickness of the dielectric, for modeling of the structures by means of
rational algebraic equations. Furthermore, for the proposed hybrid model was developed
a fitness function for the purpose of calculating the error occurred in the definitions of
FSS bandwidths with transmission features single band and dual band. This thesis deals
with the construction of prototypes of FSS with frequency settings and band widths
obtained with the use of this function. The FSS were initially reviewed through
simulations performed with the commercial software Ansoft Designer ?, followed by
simulation with the equivalent circuit method for obtaining a value of N in order to
converge the resonance frequency and the bandwidth of the FSS analyzed, then the
results obtained were compared. The methodology applied is validated with the
construction and measurement of prototypes with different geometries of the cells of the
arrays of FSS.
|
8 |
An?lise comparativa do desempenho de um Controlador Fuzzy acoplado a um PID Neural sintonizado por um Algoritmo Gen?tico com Controladores Inteligentes ConvencionaisVale, Marcelo Roberto Bastos Guerra 05 December 2007 (has links)
Made available in DSpace on 2014-12-17T14:55:04Z (GMT). No. of bitstreams: 1
MarceloRBGV.pdf: 514846 bytes, checksum: f54178f2d4a2cb1b74a993d664aad1b3 (MD5)
Previous issue date: 2007-12-05 / On this paper, it is made a comparative analysis among a controller fuzzy coupled to a PID neural adjusted by an AGwith several traditional control techniques, all of them
applied in a system of tanks (I model of 2nd order non lineal). With the objective of making possible the techniques involved in the comparative analysis and to validate the control to be compared, simulations were accomplished of some control techniques (conventional PID adjusted by GA, Neural PID (PIDN) adjusted by GA, Fuzzy PI, two Fuzzy attached to a PID Neural adjusted by GA and Fuzzy MISO (3 inputs) attached to a PIDN adjusted by GA) to have some comparative effects with the considered controller. After doing, all the tests, some control structures were elected from all the tested techniques on the simulating stage (conventional PID adjusted by GA, Fuzzy PI, two Fuzzy attached to a PIDN adjusted by GA and Fuzzy MISO (3 inputs) attached to a PIDN adjusted by GA), to be implemented at the real system of tanks. These two kinds of operation, both the simulated and the real, were very important to achieve a solid basement in order to
establish the comparisons and the possible validations show by the results / Neste trabalho ?realizado uma an?lise comparativa entre um controlador fuzzy acoplado aum PID neural ajustado por um AGcom diversas t?cnicas de controle tradicionais,
todas elas aplicadas em um sistema de tanques (modelo de 2a ordem n?o linear). Com o objetivo de viabilizar as t?cnicas envolvidas nas an?lises comparativas e validar o controle a ser comparado, foram realizadas simula??es de algumas t?cnicas de controle (PID convencional ajustado por AG, PID Neural (PIDN) ajustado por AG, Fuzzy PI, Fuzzy cascata acoplado a um PIDN ajustado por AG e Fuzzy MISO (3 entradas) acoplado a um PIDN ajustado por AG) para efeitos comparativos com o controlador proposto. Depois de realizar todos os testes simulados, foram eleitas, dentre as t?cnicas testadas na fase de simula??o, algumas estruturas de controle (PID convencional ajustado por AG, Fuzzy PI, Fuzzy cascata acoplado aum PIDN ajustado por AGeFuzzy MISO (3entradas) acoplado a um PIDN ajustado por AG) para serem implementadas no sistema real de tanques. Esses dois modos de opera??o, tanto o simulado como o real, se fizeram importantes para um embasamento s?lido para fazer as compara??es e valida??es poss?veis mostradas nos
resultados
|
9 |
Estudo de Superf?cies Seletivas de Frequ?ncia com o Uso de Intelig?ncia ComputacionalBarreto, Edwin Luize Ferreira 20 July 2012 (has links)
Made available in DSpace on 2014-12-17T14:56:11Z (GMT). No. of bitstreams: 1
EwinLFB_DISSERT.pdf: 1133428 bytes, checksum: 7db5058d9183e9d7e1e87f596676c111 (MD5)
Previous issue date: 2012-07-20 / Coordena??o de Aperfei?oamento de Pessoal de N?vel Superior / The main objective of this work is to optimize the performance of frequency
selective surfaces (FSS) composed of crossed dipole conducting patches. The
optimization process is performed by determining proper values for the width of the
crossed dipoles and for the FSS array periodicity, while the length of the crossed dipoles
is kept constant. Particularly, the objective is to determine values that provide wide
bandwidth using a search algorithm with representation in bioinspired real numbers.
Typically FSS structures composed of patch elements are used for band rejection
filtering applications. The FSS structures primarily act like filters depending on the type
of element chosen. The region of the electromagnetic spectrum chosen for this study is
the one that goes from 7 GHz to 12 GHz, which includes mostly the X-band. This
frequency band was chosen to allow the use of two X-band horn antennas, in the FSS
measurement setup. The design of the FSS using the developed genetic algorithm
allowed increasing the structure bandwidth / Este trabalho tem como objetivo principal efetuar a otimiza??o do desempenho de
estruturas de superf?cies seletivas de frequ?ncia FSS (Frequency Selective Surface),
com patches condutores na forma de dipolos em cruz. A otimiza??o foi realizada
atrav?s da identifica??o de valores ?timos para a largura do dipolo e a periodicidade do
arranjo, considerando o valor do comprimento do dipolo fixo. Especificamente,
objetiva-se determinar valores que permitam aumentar a largura de banda, utilizando
um algoritmo de busca bioinspirado com representa??o em n?meros reais. As aplica??es
t?picas de estruturas de FSS com patches condutores utilizam frequ?ncias selecionadas
atrav?s das faixas de rejei??o. As estruturas de FSS funcionam basicamente como filtros
dependendo do tipo de elemento escolhido. A regi?o do espectro eletromagn?tico
escolhida para este estudo foi a faixa de 7 GHz a 12 GHz, que inclui basicamente a
banda X. Essa regi?o do espectro eletromagn?tico foi escolhida para possibilitar a
medi??o do dispositivo com a utiliza??o de antenas de abertura do tipo corneta, que
operam na banda X. O projeto da FSS com a utiliza??o do algoritmo gen?tico GA
(Genetic Algorithm) permitiu aumentar a largura de banda da estrutura
|
10 |
An?lise multin?vel wavelet como fitness na sintonia de controladores utilizando meta-heur?sticasPires, Andr? Henrique Matias 06 December 2017 (has links)
Submitted by Automa??o e Estat?stica (sst@bczm.ufrn.br) on 2018-03-12T13:20:55Z
No. of bitstreams: 1
AndreHenriqueMatiasPires_DISSERT.pdf: 5565361 bytes, checksum: f5665a81cdc1abafd016753cfb9016e6 (MD5) / Approved for entry into archive by Arlan Eloi Leite Silva (eloihistoriador@yahoo.com.br) on 2018-03-13T20:49:03Z (GMT) No. of bitstreams: 1
AndreHenriqueMatiasPires_DISSERT.pdf: 5565361 bytes, checksum: f5665a81cdc1abafd016753cfb9016e6 (MD5) / Made available in DSpace on 2018-03-13T20:49:03Z (GMT). No. of bitstreams: 1
AndreHenriqueMatiasPires_DISSERT.pdf: 5565361 bytes, checksum: f5665a81cdc1abafd016753cfb9016e6 (MD5)
Previous issue date: 2017-12-06 / Coordena??o de Aperfei?oamento de Pessoal de N?vel Superior (CAPES) / O controle de sistemas din?micos apresenta-se como um desafio. Os m?todos tradicionalmente
utilizados na sintonia apresentam a dificuldade em expressar as especifica??es
pretendidas e conseguir encontrar controladores que atendam a esses requerimentos, sobretudo
quando o caso exige controladores mais complexos, como no caso de problemas
MIMO (Multiple Input Multiple Output). Devido ? crescente competitividade na ind?stria,
torna-se imprescind?vel utilizar t?cnicas de sintonia mais eficientes e que de fato consigam
encontrar controladores com desempenho pretendido. Pode-se, para isso, utilizar
meta-heur?sticas, como Particle Swarm Optimization (PSO), Algoritmo Gen?tico (AG) e
Algoritmo do Vagalume(AV) para a obten??o dos par?metros do controlador de acordo
com uma fun??o de avalia??o, a qual deve conseguir, de fato, codificar o qu?o bom ? um
dado controlador, expressando de forma adequada as especifica??es desejadas, de modo
que a meta-heur?stica empregada consiga encontrar o controlador que melhor satisfa?a tal
fun??o. Em vista disso, prop?e-se a utiliza??o da an?lise wavelet multin?veis, j? muito
presente na literatura, voltada para outras aplica??es, sobretudo na an?lise de sinais, sons
e imagens, para a confec??o de um ?ndice a ser utilizado como fun??o de avalia??o na
otimiza??o de controladores. A an?lise wavelet permite a apreens?o de informa??es do
comportamento e forma do sinal, informando frequ?ncia de um sinal ao longo do tempo,
caracter?stica que pode ser desej?vel, na avalia??o e projeto de controladores sendo, assim,
poss?vel avaliar separadamente o desempenho do transit?rio e do regime permanente. Foi
feito um estudo de caso, encontrando o controle otimizado de um sistema MIMO de quatro
tanques acoplados. Foi feito um estudo comparativo com outras fun??es de avalia??o
apresentadas na literatura e com o m?todo do LGR (Lugar Geom?trico das Raizes). Os
controladores implementados apresentaram o desempenho esperado, e aquele encontrado
utilizando o ?ndice proposto presentou melhor desempenho. / The control of dynamic systems is a challenge, the methods traditionally used in tuning
present the difficulty in expressing the desired specifications and being able to find
controllers that produce these requirements, especially when the case requires more complex
controllers, as in the case of Multiple Input Multiple Output (MIMO) problems. Due
to the increasing competitiveness in the industry, it becomes imperative to use more efficient
tuning techniques and that in fact can find controllers with the desired performance.
For this, one can use metaheuristics, such as Particle Swarm Optimization (PSO), Genetic
Algorithm (AG) and Vagalume Algorithm (AV) to obtain the parameters of the controller
according to a fitness function, which should in fact code how good a given controller is,
adequately expressing the desired specifications, so that the metaheuristic employed can
find the optimal controller, which best satisfies the chosen fitness function. Therefore, it is
proposed to use the multilevel wavelet analysis, already present in the literature, focused
on other applications, especially in the analysis of signals, sounds and images, for the creation
of an index to be used as a fitness function in control optimization. Wavelet analysis
allows to capture information on the behavior and shape of the signal by informing the
frequency of a signal over time, a characteristic that may be desirable, in the evaluation
and design of controllers and, thus, it is possible to separately evaluate the transient and
steady-state performances. A case study will be done, finding control of a MIMO system
of four coupled tanks. A comparative study was made with other fitness functions
presented in the literature and with the LGR (Geometric Place of Roots) method. The
implemented controllers presented the expected performance, and the one found using
the proposed index presented better performance.
|
Page generated in 0.319 seconds