Browsing Doctoral Dissertations by Title
-
Racković Babić, Kristina (Beograd , 2022)[more][less]
Abstract: Interplanetary dust grains contain important information about the Solar System. Analyzing these particles is an important aspect of the heliosphere study. Dust impacts have been observed using radio and wave instruments onboard the spacecraft since the 1980s. The interac- tion between the impact-generated plasma cloud and antenna – space- craft system elements generates the characteristic signal waveform. The present work focuses on the detection and interpretation of the dust generated signals from radio instruments onboard various space- craft orbiting at 1 AU. In the first part of the thesis, we aim to develop a model which links the observed electric signals to the dust impact properties. We propose a new model which takes into account the effect of impact - ionization - charge collection and electrostatic-influence. Our model provides an analytical expression for the pulse. It allows us to measure the amount of total ion charge, the fraction of escaping charge, the rise timescale, and the relaxation timescale. The proposed model is simple and conve- nient for large data fitting. To validate the model, we use the Time Do- main Sampler (TDS) subsystem of the STEREO/WAVES instrument, which generates high-cadence time series of voltage pulses for each monopole. Since the beginning of the STEREO mission in 2007, we have collected all the dust events detected by S/WAVES/TDS simul- taneously on all three monopoles at 1 AU. Our study confirms that the rise time vastly exceeds the spacecraft’s short timescale of elec- tron collection by the spacecraft. Aside from electron dynamics, we also obtained interesting results regarding the cloud’s electron tem- perature. The presented model provides an effective tool for analyzing vii dust waveforms, and is applicable for different space missions which investigate the distribution of dust particles, e.g., Solar Orbiter and Parker Solar Probe. In the second part of the thesis, we focus on the interstellar dust (ISD). Interplanetary and interstellar dust are the two main dust pop- ulations at 1 AU. Our objective is to search for interstellar dust by analyzing the data sets collected by STEREO and Wind, starting from the beginning of the missions. Between 2007 and 2012, while being at the solar minimum with a solar dipole pointing southward, all three spacecraft recorded ISD flux at 1 AU. However, before and after that period, the disappearance of the interstellar component was noticeable. The observed change of the impact rate suggests that the flux of inter- stellar dust at 1 AU varies with the solar cycle. Each time the magnetic dipole field changes its polarity during the solar cycle, small interstel- lar grains experience focusing or defocusing. Consequently, the dust grains are systematically deflected either towards, or away from the solar magnetic equator plane by the solar wind magnetic field which thus affects the dust dynamics and the total interstellar dust flux in the inner heliosphere. Our study provides the first quantitative de- scription of the time variation of ISD flux at 1 AU. URI: http://hdl.handle.net/123456789/5547 Files in this item: 1
Teza_KRB.pdf ( 10.95Mb ) -
Božović, Nataša (Belgrade)[more][less]
-
Surla, Katarina (Novi Sad , 1980)[more][less]
-
Jevtić, Miroljub (Belgrade)[more][less]
-
Vučković, Bojan (Beograd , 2017)[more][less]
Abstract: We present original results from the following fields of discrete mathematics: chromatic graph theory, extremal set theory and Boolean matrix theory. From the chromatic graph theory we investigate edge and total colorings satisfying the condition that neighboring vertices of a graph possess different values of multi-set, set or sum, induced by the giving coloring. Multi-set neighbor-distinguishing edge coloring of a graph is an assignment of colors to edges such that, for every edge uv of a graph, multi-set of the edges incident with the vertex u differs from the multi-set of the edges incident with the vertex v. The previous best result concerning the minimum number of colors required for such a coloring of an arbitrary graph states that four colors are sufficient. The author’s contribution is a proof that such a coloring is always possible with only three colors, which is in general case the optimal number of colors. We construct a graph for which we subsequently prove that a different number of colors is required to obtain a multi-set neighbor-distinguishing coloring and neighbor-distinguishing coloring by sum. As far as we know, this is the first example of such a graph. A few results concerning the neighbor expended sum distinguishing coloring are given. The main contribution is a proof that for an arbitrary graph there exists a total coloring from the set f1; 2; 3g, such that every two adjacent vertices have different sums of its adjacent vertices and incident edges. Also, for certain classes of graphs is proved that there exists such a coloring using only the colors from the set f1; 2g. Neighbor-distinguishing edge coloring of a graph G requires that every two adjacent edges receive different colors, while the sets of the edges incident with the vertices u and v differ for every edge uv of G. The author presents a procedure of edge coloring for an arbitrary graph without isolated edges, where we a smaller number of colors is used compared to all known results. For the adjacent vertex distinguishing total coloring of a graph G the condition is that every two adjacent and incident elements of V (G) [ E(G) receive different colors, while for every edge uv of G the set composed from the colors assigned to the edges incident with u together with the color of u, differs from such a set for v. The author improves the upper bound of the minimum number of colors needed for such a coloring, relative to the maximal degree of a graph. Frankl’s conjecture from the extremal set theory states that for every family closed under union there exists an element contained in at least half of the sets of the family. We give a proof that Frankl’s conjecture holds for every family contained from 12 elements, while it is known that this is true for families contained from 11 or less elements. Our proof is based on the efficient algorithm that exhausts all the possibilities, while using the results for subfamilies that eventual counter-example cannot contain, which we obtained in a number of consecutive steps. Family of sets G is an FC-family if for every family F containing G there exists an element from S G that appears in at least half of the sets of F. NonFC-family is every family that is not FC. The author’s contribution is the complete classification of all families consisting of 6 or less elements into FC and NonFC-families. From the Boolean matrices theory we present our results concerning the row space cardinality. Boolean matrices are the matrices whose all components are from the set f0; 1g, while the row space of a Boolean matrix is the set of vectors that can be obtained by disjunction from the rows of a matrix. We present the set consisted of all values a from the interval [2n2 + 2n3; 2n2] such that there exists a matrix of dimension n n having the row space cardinality equal to a. For the least positive integer an for which there exists no matrix of dimension n n having the row space cardinality equal to an, the author gives a lower bound that is an improvement over the previously known results. All proofs for the main results in the dissertation are constructive. Proofs of some of them require the use of computers where there is a calculation of a great number of possibilities. For other proofs this was not necessity, though algorithms following the steps of the proofs can be implemented to obtain a graph coloring or a matrix with the desired properties. URI: http://hdl.handle.net/123456789/4661 Files in this item: 1
Disertacija_-_Bojan_Vuckovic.pdf ( 1.143Mb ) -
Ivanović, Marija (Beograd , 2022)[more][less]
Abstract: This dissertation focuses on the Roman domination problem and its two modifications. Improvements and relaxations of two integer linear programming for- mulations for the Roman domination problem from the literature are introduced, proved to be equivalent to the existing ones despite of the variables relaxation and usage of fewer number of constraints and compared by standard optimization solvers, CPLEX and Gurobi. Improved formulations can be equally used as original ones, but, as it can be seen from numerical results, for some instances, they can be more useful. Given the fact that old and new formulations can not be used for some large size instances, and that algorithms for solving Roman domination problem are mostly defined for some particular graph classes, the aim of this research was to find a new algorithm that can be used for solving Roman domination problem on all graph classes and all graph sizes. Although the Roman domination problem belongs to the class NP, presented algorithm is able to find solution value equal to optimal solution value on very large number of instances in less then a second. For the first modification of the Roman domination problem, named Restrained Roman domination problem, a new mixed integer linear programming formulation is intro- duced and, to the best of the author’s knowledge, this formulation is the first in the literature. For the second modification of the Roman domination problem, the Weak Roman domination problem, an improved integer linear programming formu- lation is presented. Improved formulation is also proved to be correct, equivalent to the existing formulation from the literature and compared using standard op- timization solvers, CPLEX and Gurobi. Numerical results showed the advantage of the improved formulation on almost all tested instances. Additionally, an im- proved linear-time algorithm for solving the Weak Roman domination problem on block graphs is introduced and, similarly to the Roman domination problem, a new algorithm, based on the variable neighborhood search method is presented. With the new variable neighborhood search based algorithm we aimed to find solution of the Weak Roman domination problem equal to the optimal on very large number of tested instances. For instances for which some solution value is found but not proved to be an optimal, presented algorithm provided the new lower-bounds. Even more, for some instances, where optimization solvers were not able to prove optimality or to find any solution, new solutions are found. URI: http://hdl.handle.net/123456789/5431 Files in this item: 1
MarijaIvanovic_ ... a_saPotpisanimIzjavama.pdf ( 1.958Mb ) -
Pejović, Tadija (Belgrade)[more][less]
-
Milić Žitnik, Ivana (Beograd , 2017)[more][less]
Abstract: The subje t of this dissertation is intera tion between the mean motion resonan- es and the Yarkovsky e e t. This intera tion o urs when an asteroid due to the hanges of its orbital semi-major axis ( aused by the Yarkovsky e e t) rea h the re- sonan e. The resonan e indu es a periodi os illations in the asteroid's semi-major axis around its enter. The Yarkovsky e e t exa tly auses the permanent (se ular) evolution of the orbital semi-major axis. As a result of their intera tion the mean semi-major axis drift speed is modi ed with respe t to the one aused solely by Yarkovsky. One of the main goals of this investigation was to study this intera tion, and to establish and de ne how the time that an asteroid spend in the resonan e depends on some hara teristi s of this resonan e, as well as of the asteroid itself. So far, the impa t of the resonan e on the semi-major axis drift speed has not been studied to that extent neither from that point of view. In order to study the afo- rementioned intera tion the orbital motion of test parti les a ross the resonan es is numeri ally simulated using ORBIT9 integrator. The most important result of this dissertation ertainly is determination of fun tional relation between on one side the time-period that obje ts spend inside a resonan e, and, on the other side, the semi- majors axis drift speed, the orbital e entri ity and the resonan e strength. In this work not only that existen e of the above-mentioned relationship is on rmed, but for the rst time it was expli itly de ned. Two the most interesting results are that the time spent in the resonan e is inversely proportional to the semi-major axis drift speed aused by the Yarkovsky e e t, and that this time is dire tly proportional to the resonan e strength. URI: http://hdl.handle.net/123456789/4651 Files in this item: 1
Milic-Zitnik_Ivana.pdf ( 40.69Mb ) -
Gardašević Filipvić, Milanka (Beograd , 2011)[more][less]
-
Milošević, Stanislav (Beograd , 2023)[more][less]
Abstract: In this dissertation, we presented galaxy mergers and the forming of stellar morphological substructures. We assume a large spiral galaxy and dwarf galaxy which is a satellite of the first one. We used N-body simulations to present different scenarios for the merger of two galaxies, and after that, we analyzed the properties of formed structures. The investigation of the formation of these structures and their properties is important for understanding the dynamics and evolution of galaxies. First, we did simulations where we investigated the influence of the properties of dwarf galaxies on the forming structures. We tested: 1. morphology of the dwarf galaxy where we used two models – dwarf with a disk and spheroidal dwarf galaxy; 2. inclination of the orbit in the case of a very radial merger, because in that case, we have the formation of the stellar shells and streams; 3. direction of rotation of the dwarf in the case of a dwarf with a disk. In each case, after the merger, we have stellar shells and streams formed. Morphology, inclination of the orbit, and direction of rotation have their influence on the properties of formed substructures, and on the timescale of disruption of the remnant of the dwarf. In the case of the merger of Andromeda galaxy (M31) and dwarf galaxy, the satellite of M31, we investigate the properties of the Giant Stellar Stream (GSS), as well as the Northeast shell (NE) and West shell (W). The orientation of the GSS, distances, and velocities of the GSS, NE, and W shells from our simulation are in agreement with the observed one. For the first time, we explained the observed metallicity distribution in these substructures. With a linearly decreasing gradient of the initial metallicity in the dwarf galaxy before the merger, using Monte Carlo (MC) simulations, we successfully explained the observed metallicity distribution in these substructures. These results are a contribution to the investigation of metallicity gradients in dwarf galaxies which is important for galaxy evolution in general. URI: http://hdl.handle.net/123456789/5583 Files in this item: 1
SM_Doktorat_18_07_2023.pdf ( 18.08Mb ) -
Jovanikić, Branko (Belgrade)[more][less]
-
Torgašev, Aleksandar (Belgrade)[more][less]
URI: http://hdl.handle.net/123456789/101 Files in this item: 1
phdAleksandarTorgasev.pdf ( 83.67Mb ) -
Kašanin, Radivoj (Belgrade)[more][less]
-
Mutavdžić, Nikola (Beograd , 2023)[more][less]
Abstract: In this PhD thesis we investigate bounds of the gradient of harmonic and harmonic quasiconformal mappings. We also discuss such bounds for functions that are harmonic with respect to the hyperbolic metric or certain other metrics. This research has been motivated by some recent results about Lipschitz-continuity of quasiconformal mappings that satisfy the Laplace gradient inequality. More precisely, the mappings we consider are solutions of the Dirichlet problem for the Poisson equation and can be considered as a generalization of harmonic mappings. Besides the ball, we also work with general domains on which solutions of the Dirichlet problem are defined, as well as general codomains. Finally, we announce new results that have been formulated for regions of C1,α-smoothness, both as the domain and the codomain. Besides presenting the main results, we give an overview of general notions from differential geometry and recall some of the properties of hyperbolic metric in an n-dimensional ball. We also state properties of harmonic and sub-harmonic functions with respect to the hyperbolic metric, which are analogous to some classical results from the theory if harmonic functions and Hardy’s theory. It turns out that the gradients of hyperbolic harmonic functions behave differently from those of euclidean harmonic functions. A similar conclusion is obtained for the family of Tα-harmonic functions. Namely, unlike the space of harmonic functions, the solution of the Dirichlet problem in the space of Tα-harmonic functions is shown to be Lipschitz-continuous when so is the boundary function. In addition, we investigate Höldercontinuity of the solution of the Dirichlet problem for the Poisson equation in the euclidean and hyperbolic metric. We will present versions of the Schwarz lemma on the boundary for pluriharmonic mappings in Hilbert and Banach spaces. These results will follow from the version of the Schwarz lemma for harmonic mappings from the unit disc to the interval (1, 1) without the assumption that the point z = 0 maps to itself. Furthermore, we show a version of the boundary Schwarz lemma for harmonic mappings from a ball to a ball, not necessarily of the same dimension. The proof uses a version of the Schwarz lemma for multivariable functions, first considered by Burget. This result is obtained by integrating the Poisson kernel over so-called polar caps. The assumption that point z = 0 maps to itself is again not needed, thus yielding a generalization of a recent result by D. Kalaj. At the end of this section, it is demonstrated that the analogous result is false in the case of hyperbolic harmonic functions. In a certain sense, this means that the Hopf lemma is not valid for hyperbolic harmonic functions. Amongst various versions of the Schwarz lemma, we have been investigating bounds of the modulus for classes of holomorphic functions f on the unit disc whose index If fulfils certain geometric conditions. These classes are a generalization of the star and α-star functions, previously investigated by B. N. Örnek. Our method is based on using Jack’s lemma and can be applied in certain more general cases. As an illustration, we derive the sharp bounds for the modulus of a holomorphic function f with index If whose codomain is a vertical strip, as well as bounds for the modulus of the derivative of f at point z = 0. Moreover, we give a bound for the rate of growth of the modulus of holomorphic functions on disk U that map point z = 0 to itself and whose codomain is a vertical strip. URI: http://hdl.handle.net/123456789/5582 Files in this item: 1
Doktorska_Disertacija_Nikola_Mutavdzic.pdf ( 939.2Kb ) -
Tošić, Dušan (Belgrade , 1984)[more][less]
-
Malešević, Jovan (Belgrade)[more][less]
-
Svetlik, Marek (Beograd , 2020)[more][less]
Abstract: In this dissertation we consider various versions of the Schwarz lemma and theSchwarz-Pick lemma for holomorphic, harmonic and harmonic quasiregular mappings. Inaddition, in order to present new results, an overview of the results that can be considered asclassical is given. As one of the most important consequence of the Schwarz-Pick lemma forholomorphic mappings, an introduction of the hyperbolic distancedΩon the simply connecteddomainsΩ C(such thatΩ6=C) is given in details, as well as the connection of that distanceand holomorphic mappings. All versions of the Schwarz lemma and the Schwarz-Pick lemma for harmonic mappingsare shown as assertions analogous to the corresponding claims for holomorphic mappings.In the proofs of these assertions, the properties of the hyperbolic distance and Euclideanproperties of hyperbolic disks are used. Firstly, we considered some versions of the Schwarzlemma for harmonic mappings from the unit disk to the interval (-1,1) and then for harmonicmappings of the unit disk into itself, without the assumption thatz= 0is mapped to itself bythe corresponding map. Thereby, the corresponding inequalities were shown to be sharp andextremal mappings were found. By using the strip and half plane method, simple proofs ofthe Schwarz-Pick lemma for real-valued harmonic mappings are given, as well as the simpleproofs of their corollaries that are formulated in terms of corresponding hyperbolic distances.For both holomorphic and harmonic mappings a version of the Schwarz lemma have beenformulated and proved in the case where the values of these mappings and values of thenorms of their differentials, at the pointz= 0, are given. Also, in that case we showed thatthe corresponding inequalities are sharp and extremal mappings were found. It has also beenshown that the same methods can be used to obtain Harnack’s inequalities for harmonicmappings, as well as for their generalizations.Furthermore, we give simple proofs of a version of the Schwarz-Pick lemma for harmonicquasiregular mappings whose codomain is a half plane or a strip. One version of the Schwarzlemma for harmonic quasiregular mappings from the unit disk into a strip is obtained thanksto the appropriate (which seems unexpected) inequality satisfied by the Euclidean and hyper-bolic distances on the strip. By using the properties of the Gaussian curvature we also showthat harmonic quasiconformal mappings of the hyperbolic domain into convex hyperbolicdomain are quasi-isometries of the corresponding metric spaces.The introduction of the hyperbolic distance is shown in two ways. The first way is classi-cal one. Starting from the hyperbolic metric on the unit disk, first we define the hyperboliclength of theC1curve and then the hyperbolic distance between two given points. Thesecond one is based on the axiomatic foundation of the absolute plane geometry. Startingfrom the theorem related to the existence and the uniqueness (up to the unit for length) ofthe distance in the absolute plane (which is in accordance with the basic geometry relations- between and congruence), we simultaneously derive the formula for that distance in twomodels of that plane. One of these models is the set of complex numbersC, observed as amodel of the Euclidean plane and the second one is the unit disk that is considered as thePoincaré disk model of hyperbolic plane. URI: http://hdl.handle.net/123456789/5091 Files in this item: 1
svetlik_marek-phd.pdf ( 1.279Mb ) -
Ilić, D. Ivana (Belgrade , 2013)[more][less]
Abstract: For the sequence of heavy-tailed and possibly dependent random variables with the missing observations the estimation of the tail-index is considered. Under minimal but verifiable assumption of ''extremal dependence'' we proved the consistency of geometric-type estimator (Brito and Freitas, 2003). We extended results from Mladenovic and Piterbarg (2008) and proved the consistency and the asymptotic normality of the Hill estimator. Illustrative examples are provided. URI: http://hdl.handle.net/123456789/2485 Files in this item: 1
Elektronska verzija.pdf ( 3.250Mb ) -
Jovanović, Milan (Beograd , 2015)[more][less]
Abstract: Early papers dealing with so-called stress-strength problems were published in the middle of the 20th century. This topic, which belongs to the reliability theory, is still very active nowadays, which can be seen through the number of published papers dealing with it - around ten each year. In this dissertation, some methods for estimation of the reliability parameter for a system with independent stress and strength are presented. Also, two new models are introduced and some estimators of the reliability parameter for each of them are derived. The dissertation is divided into four chapters. In the rst chapter, some basic terms are introduced and some examples from real life, illustrating big possibilities for application of the results from this scienti c eld, are described. Sorted based on the stress and strength distributions, a chronological overview of all research activities dealing with these topics, to the author's best knowledge, is presented. Some special func- tions, which are later used for calculations, along with their main properties are shown. The expressions for the reliability parameter for some stress and strength distributions are either derived or listed. The second chapter is devoted to different methods used for point esti- mation, as well as for interval estimation of the reliability parameter of a system. For each methods estimators of the reliability parameter for some stress and strength distributions are either derived or listed. In the third chapter, a new model is introduced. In this model, the stress has geometric, while the strength has Poisson distribution. This is one of the rst, if not the rst, appearances in the literature, where the stress and strength distributions do not belong to the same family of distributions. For this model, the reliability parameter is estimated using different methods and decision on optimal estimators for usage in practice is based on the simulations. In the fourth chapter, another model is introduced, with the stress and strength distributions which are not only from different families of distribu- tions, but also do not belong to the same type of distributions. The stress has geometric, while the strength has exponential distribution. The reliabil- ity parameter for this model is also estimated using different methods, and the decision on optimal estimators for usage in practice is once again based on the simulations. URI: http://hdl.handle.net/123456789/4352 Files in this item: 1
Jovanovic_Milan_teza.pdf ( 4.636Mb ) -
Mladenović, Pavle (Belgrade , 1985)[more][less]