京都大学 / 0048 / 新制・課程博士 / 博士(情報学) / 甲第19111号 / 情博第557号 / 新制||情||98(附属図書館) / 32062 / 京都大学大学院情報学研究科知能情報学専攻 / (主査)教授 松山 隆司, 教授 美濃 導彦, 准教授 梁 雪峰 / 学位規則第4条第1項該当 / Doctor of Informatics / Kyoto University / DFAM

Every entity in the real world can be described uniquely by it’s attributes. It is possible to rank similar entities based on these attributes, i.e. a professor can be ranked by his/her number of publications, citations etc. A community is formed by a group of connected entities. Individual ranking of an entity plays an important role in the quality of a community. Skyline community in a network represents the highest ranked communities in the network. But how do we define this ranking? Ranking system in some model considers only a single attribute [16], whereas the other [15] [23] considers multiple attributes. Intuitively multiple attributes represent a community better and produce good results. We propose a novel community discovery model, which considers multiple attribute when ranking the community and is efficient in terms of computation time and result size. We use a progressive (can produce re- sults gradually without depending on the future processing) algorithm to calculate the community in an order such that a community is guaranteed not to be dominated by those generated after it. And to verify the dominance relationship between two communities, we came up with a range based comparison where the dominance rela- tionship is decided by the set of nodes each group dominates. If domination list of a group is a subset of another group, we say the second group dominates the first. Because a groups domination list contains it’s member along with the nodes they dominate. So in the example, the second group dominates every node of the first group. / Graduate

No description available.

No description available.

No description available.

Quad-rotor helicopters are agile aircraft which are lifted and propelled by four rotors. Unlike traditional helicopters, they do not require a tail-rotor to control yaw, but can use four smaller fixed-pitch rotors. However, without an intelligent control system it is very difficult for a human to successfully fly and manoeuvre such a vehicle. Thus, most of recent research has focused on small unmanned aerial vehicles, such that advanced embedded control systems could be developed to control these aircrafts. Vehicles of this nature are very useful when it comes to situations that require unmanned operations, for instance performing tasks in dangerous and/or inaccessible environments that could put human lives at risk. This research demonstrates a consistent way of developing a robust adaptive controller for quad-rotor helicopters, using fuzzy-neural networks; creating an intelligent system that is able to monitor and control the non-linear multi-variable flying states of the quad-rotor, enabling it to adapt to the changing environmental situations and learn from past missions. Firstly, an analytical dynamic model of the quad-rotor helicopter was developed and simulated using Matlab/Simulink software, where the behaviour of the quad-rotor helicopter was assessed due to voltage excitation. Secondly, a 3-D model with the same parameter values as that of the analytical dynamic model was developed using Solidworks software. Computational Fluid Dynamics (CFD) was then used to simulate and analyse the effects of the external disturbance on the control and performance of the quad-rotor helicopter. Verification and validation of the two models were carried out by comparing the simulation results with real flight experiment results. The need for more reliable and accurate simulation data led to the development of a neural network error compensation system, which was embedded in the simulation system to correct the minor discrepancies found between the simulation and experiment results. Data obtained from the simulations were then used to train a fuzzy-neural system, made up of a hierarchy of controllers to control the attitude and position of the quad-rotor helicopter. The success of the project was measured against the quad-rotor’s ability to adapt to wind speeds of different magnitudes and directions by re-arranging the speeds of the rotors to compensate for any disturbance. From the simulation results, the fuzzy-neural controller is sufficient to achieve attitude and position control of the quad-rotor helicopter in different weather conditions, paving way for future real time applications.

This work is aimed for proposing acceptable indexing of moving objects. With the enlargement of mobile computing it is needed to manage large sets of spatiotemporal data. We introduce the problem of spatiotemporal data and basic general approaches of indexing these data. Further, we show support of spatial data in Oracle. The movement is typically represented as trajectory in two dimensional space with temporal component in third dimension. The thesis contains experiments performed in database Oracle on artificially generate data.

The environment is made up of composition of small particles. Hence, particle simulation is an important tool in many scientific and engineering research fields to simulate the real life processes of the environment. Because of the enormous amount of data in such simulations, data management, storage and processing are very challenging tasks. Spatial Distance Histogram (SDH) is one of the most popular queries being used in this field. In this thesis, we are interested in investigating the performance of improvement of an existing algorithm for computing SDH. The algorithm already being used is using a conceptual data structure called density map which is implemented via a quad tree index. An algorithm having density maps implemented via binary tree is proposed in this thesis. After carrying out many experiments and analysis of the data, we figure out that although the binary tree approach seems efficient in earlier stage, it is same as the quad tree approach in terms of time complexity. However, it provides an improvement in computing time by a constant factor for some data inputs. The second part of this thesis is dedicated to an approach that can potentially reduce the computational time to a great extent by taking advantage of regions where data points are uniformly distributed.

Due to recent legislation aimed at reducing carbon emissions into the environment through weight reduction, the automotive and aerospace industries are using light alloys such as aluminium silicon (Al–Si) and aluminium zirconium (Al–Zr) instead of steel due to their excellent mechanical properties and low weight to strength ratio. In order to further improve mechanical and metallurgical properties in these alloys, grain refinement is usually used in industry. However, the current and most widely used grain refiner Al–5Ti–B is unable to refine Al–Si alloys with silicon content greater than 3 wt.%., and Al–Zr alloys due to poisoning of the refiner by silicon and zirconium. The Al–5Ti–B refiner also contains larger Al3Ti particles and agglomerates of TiB2 which affect its efficiency and suitability in industrial applications where thin sheets are required. In this study, a new technique which improves the microstructure and efficiency of the Al–5Ti–B refiner has been developed. This involves the reaction of potassium tetrafluoroborate (KBF4) and potassium hexafluorotitanate (K2TiF6) salts at shorter reaction time before ultrasonic processing during solidification. This leads to the formation of a new Al3Ti morphology and de-agglomeration of TiB2 particles which enhances its grain refinement efficiency by 20%. Secondly, through phase diagram analysis of Al grain refining systems and crystallography studies, it was observed that Al3Ti and Al3Nb display similar lattice parameters with atomic misfit of 4.2% and would undergo a peritectic reaction with α-Al at low contact angles. Based on this, and using the duplex nucleation theory and poisoning by Si and Zr, a new quaternary grain refiner containing aluminium, titanium, niobium and boron (Al–4Ti–Nb–B) has been developed. This novel grain refiner has been found to be efficient in Al–Si alloys and Al–Zr, both at laboratory and industrial scales, and to improve the mechanical properties of the alloys despite the presence of Ti in the alloy. It was observed that the addition of Nb to an Al–Ti–B system leads to the formation of solid solution phases of Al3Ti1-xNbx, Al3Nb1-xTx, and (Ti1-xNbx)B2 which prevents poisoning by Si and Zr. Experimental simulations showed that Al3Nb1-xSix rather than Ti(Al1-xSix)3 are formed in Al–Si alloys, and Al3(Ti1-xNbx) and (Al3Ti1-xNbx)B2 phases are formed in Al–Zr alloys rather than Al3(Zrx,Ti1-x), B2(Zrx,Ti1-x) or ZrB2 phases. A new grain refining mechanism, ‘The Quad Nucleation Theory’ based on four nucleation events in Al–4Ti–Nb–B has been proposed. Other newly developed quaternary and ternary novel grain refiners capable of refining aluminium silicon alloys are also presented in this thesis. This includes a novel method of refining Al–Si alloys using phosphorus and niobium.

Particle simulation has become an important research technique in many scientific and engineering fields in latest years. However, these simulations will generate countless data, and database they required would therefore deal with very challenging tasks in terms of data management, storage, and query processing. The two-body correlation function (2-BCFs), a statistical learning measurement to evaluate the datasets, has been mainly utilized to measure the spatial distance histogram (SDH). By using a straightforward method, the process of SDH query takes quadratic time. Recently, a novel algorithm has been proposed to compute the SDH based on the concept of density map (DM), and it reduces the running time to ϴ(N(3/2)) for two-dimensional data and ϴ (N(5/3) ) for three-dimensional data, respectively. In the DM-SDH algorithm, there are two types of DMs that can be plugged in for computation: Quad-tree (Oct-tree for three-dimensional data) and k-d tree data structure. In this thesis paper, by using the geometric method, we prove the unre- solvable ratios on the k-d tree. Further, we analyze and compare the difference in the performance in each potential case generated by these DM-SDH algorithms. Experimental results confirm our analysis and show that the k-d tree structure has better performance in terms of time complexity in all cases. However, our qualitative analysis shows that the Quad-tree (Oct-tree) has an advantage over the k-d tree on aspect of space complexity.