No description available.

No description available.

The present study examined 8 -10 month-old and 11 -13 month-old infants’ capabilities to infer constraints imposed by an explicit visual pathway on object movement of object. In a preferential looking paradigm, infants observed a ball rolling down a U– or V–shaped path. In the U-shaped path, infants observed a ball rolling from beginning to the end of the path (possible), and a ball rolling down from the beginning and stopping midway of the path (impossible). In the V-shaped path, infants observed a ball rolling from beginning and stopping midway (possible), and a ball rolling down the beginning to end of the path (impossible). Analyses of looking times showed a marginal effect on path, with longer looking times towards the possible V-shaped path by both age groups. Overall, infants had weak representations for constraints induced by an explicitly presented path on movement of the object.

The work presented here is an experimental investigation of the effect of cylindrical geometry on electrical conductivity, in which single-crystal samples of cadmium at the temperature of liquid helium are used, with the diameter on the order of the electron mean free path.

This thesis develops a theory for approximate quantum time-correlation functions, Matsubara dynamics, that rigorously describes how to combine quantum statistics with classical dynamics. Matsubara dynamics is based on Feynman's path integral formulation of quantum mechanics and is expected to describe the physics of any system that satisfies quantum Boltzmann statistics and exhibits rapid quantum decoherence, e.g. liquid water at room temperature. Having derived the Matsubara dynamics theory and explored the symmetry properties that it shares with the quantum Kubo time-correlation function, we demonstrate that two heuristic computational methods, Centroid Molecular Dynamics and Ring Polymer Molecular Dynamics, are based on quantifiable approximations to the Matsubara dynamics time-correlation function. This provides these methods with a stronger theoretical foundation and helps to explain their strengths and shortcomings. We then apply the Matsubara dynamics theory to a recently developed computational method of Poulsen et al. called the planetary model. We show that the planetary model is based on a harmonic approximation to Matsubara dynamics that is engineered to maintain the conservation of the quantum Boltzmann distribution, so quantum statistics and classical dynamics remain harmonised. By making practical modifications to the planetary model, we were able to calculate infrared absorption spectra for a point charge model of condensed-phase water over a range of thermodynamic conditions. We find that this harmonic approximation to Matsubara dynamics provides a good description of bending and vibrational motions and is expected to be a useful tool for future spectroscopic studies of more complex, polarisable models of water.

Our related research review on propagation models reveals six factors that are significant in last mile connectivity via LAP: path loss, elevation angle, LAP altitude, coverage area, power consumption, operation frequency, interference, and antenna type. These factors can help with monitoring system performance, network planning, coverage footprint, receivers' line-of-sight, quality of service requirements, and data rates which may all vary in response to geomorphology characteristics. Several competing propagation models have been proposed over the years but whilst they collectively raise many shortcomings such as limited altitude up to few tens of meters, lack of cover across different environments, low perdition accuracy they also exhibit several advantages. Four propagation models, which are representatives of their types, have been selected since they exhibit advantages in relation to high altitude, wide coverage range, adaption across different terrains. In addition, all four have been extensively deployed in the past and as a result their correction factors have evolved over the years to yield extremely accurate results which makes the development and evaluation aspects of this research very precise. The four models are: ITU-R P.529-3, Okumura, Hata-Davidson, and ATG. The aim of this doctoral research is to design a new propagation model for last-mile connectivity using LAPs technology as an alternative to aerial base station that includes all six factors but does not exhibit any of the shortcomings of existing models. The new propagation model evolves from existing models using machine learning. The four models are first adapted to include the elevation angle alongside the multiple-input multiple-output diversity gain, our first novelty in propagation modelling. The four adapted models are then used as input in a Neural Network framework and their parameters are clustered in a Self-Organizing-Map using a minimax technique. The framework evolves an optimal propagation model that represents the main research contribution of this research. The optimal propagation model is deployed in two proof-of-concept applications, a wireless sensor network, and a cellular structure. The performance of the optimal model is evaluated and then validated against that of the four adapted models first in relation to predictions reported in the literature and then in the context of the two proof-of-concept applications. The predictions of the optimised model are significantly improved in comparison to those of the four adapted propagation models. Each of the two proof-of-concept applications also represent a research novelty.

Shortest path problems arise in a variety of applications ranging from transportation planning to network routing among others. One group of these problems involves finding shortest paths in graphs where the edge weights are defined by probability distributions. While some research has addressed the problem of finding a single shortest path, no research has been done on finding multiple paths in such graphs. This thesis addresses the problem of finding paths for multiple robots through a graph in which the edge weights represent the probability that each edge will fail. The objective is to find paths for n robots that maximize the probability that at least k of them will arrive at the destination. If we make certain restrictions on the edge weights and topology of the graph, this problem can be solved in O(n log n)time. If we restrict only the topology, we can find approximate solutions which are still guaranteed to be better than the single most reliable path.

The expected yield to maturity on a defaultable obligation equals the nominal yield less expected default losses. However, in a mark-to-market world, one doesn't have the luxury of reporting one's performance on the basis of yield to maturity. Total return is calculated for an arbitrary holding period, and must reflect any mark-to-market gains or losses as at the close of the period-gains or losses that can be triggered by the bond's upgrade or downgrade. Thus to estimate expected total return, one must estimate not only expected default losses, but also the impact on capital price of expected ratings transitions. This paper begins with the observation that a bond which is blessed by more favourable transition characteristics is likely to produce a higher total return, and poses the question of how that benefit can be quantified. How much is it worth? To answer the question, I start by specifying a formal bond-pricing model reflective of ratings transitions. I survey various statistical methods and past research efforts to identify the ratings-transition matrix which best parametrises the model, and propose a novel test for selecting between competing matrices. Using this approach, I replicate several important studies of ratings transitions. I also use it to examine new published and unpublished data, testing for (and finding) ratings path-dependency, and otherwise exploring the effect of ratings changes on different bond sectors. I then turn to the question of whether it is possible to estimate bond-specific transition probabilities, and propose a way to do so. I combine these efforts into the specifications for a pricing model capable of answering the question: How much is it worth?

The conventional information used to guide automobile drivers in selecting their driving routes is the shortest-distance path (SDP). As several researchers have pointed out, driver route selection is a multiple criteria decision process. This research proposes a multiobjective path optimisation (MOPO) decision model to make a more precise simulation of the decision-making behaviour of driver route selection. Seven single-objective path optimisation (SOPO) decision models are taken into account to establish the MOPO decision model. They relate to travel time, travel cost, cumulative distance, roadway capacity, roadway grade, passed intersections and number of turns. To solve the MOPO problem, a two-stage technique which incorporates shortest path (SP) algorithms and techniques for solving the multiobjective programming problem and a path genetic algorithm (PGA) are proposed. In addition, algorithms such as Dijkstra, A* and GA are reviewed and algorithms that are applicable for solving the MOPO problems are suggested. Furthermore, new algorithms for solving least-node path (LNP) problem, corresponding to the objective of passed intersections, as well as minimum-turn path (MTP) problem, corresponding to the objective of number of turns, are developed. To conduct the empirical study, a software tool - the multiobjective path optimisation analysis tool (MOPOAT) - was implemented. It contains tools for constructing a road network and its corresponding network topology, the environment of coding techniques for solving the MOPO problems and tools for the manipulation, statistics, analysis and display of experimental results. The purpose of implementing the MOPOAT software is to provide more efficient, convenient and user-friendly tools for solving MOPO and SOPO problems so that an empirical study of real road networks can be carried out more easily. To demonstrate the advantages of the proposed model in supporting more diverse information to drivers to assist in route selection, several experiments were conducted utilising three real road networks with different roadway types and numbers of nodes and links. Techniques and algorithms such as the two-stage approach, Dijkstra and the PGA for solving the MOPO problem, and the Dijkstra, LNP and MTP algorithms for solving the SOPO problems were applied. Finally, to deal with improvements in computational efficiency for identifying SPs in a large road network and for population initialisation of the PGA, the critical-section (CS) approach and the seed-path expansion (SPE) approach are proposed. To compare the run time between the conventional SP and CS algorithms as well as the PGA and the SPE algorithms, tools were implemented with commercial GIS, and experimental tests were conducted using road networks with a large amount of nodes and links and different roadway types. Through these theoretical and empirical studies, several useful contributions and conclusions were obtained. Some of the most significant findings are: 1. The experimental results demonstrate the advantages of integration with commercial GIS packages in supporting both spatial and attribute data displays. It can be safely said that, assisted by the MOPOAT software, it is easy for automobile drivers to obtain the optimal paths of the SDP, LNP, MTP and MOPO problems in seconds, despite these problems being highly complex and difficult to resolve manually. 2. According to the experimental results, the proposed LNP, MTP and MOPO decision models give automobile drivers richer information for choosing their driving routes in a more diverse way. 3. It is shown by the experimental results that the SDP and LNP mostly locate different paths in both radial-circumferential and grid-type road networks, and that the total passed intersections by the SDP are greater than passed by the LNP. Moreover, it is revealed that ambiguous turns might occur in both radial-circumferential and grid-type road networks. 4. It is found that the number of nodes of the SDP is in general greater than the number of nodes of the LNP and MTP irrespective of the type of road network. 5. A sensitivity analysis for weights shows that as the weighting value of the SDP objective incrementally increases by 0.1 units, the corresponding SDP??s objective value varies either low or high. The same results also occur for the LNP and MTP objectives. This verifies the fact that the weighting coefficients do not reflect proportionally the relative importance of the objectives. Moreover, the MTP objective has the higher sensitivity in comparison with the other two objectives. 6. Despite utilising Dijkstra or PGA algorithms for solving the MOPO problem, the LNP and MTP algorithms have to be employed to solve the non-commeasurable problem, whereby the standardisation objective value can be obtained. In addition, without any assisting information the PGA might fail to reach the best-compromise solution. 7. It is found that the total run time for solving the MOPO problem by applying the Dijkstra algorithm is much faster than by the PGA. However, if the run time excludes the time needed for population initialisation, the PGA is much faster than the Dijkstra algorithm. 8. Based on calculated bottlenecks, the proposed CS approach partitions a SP into many critical sections in advance, with the result that a long SP can be obtained by combining all SPs of all CSs. The experimental results show that the run time of the CS algorithm is much faster than Dijkstra??s algorithm. Moreover, the test result for the P-pointer indicates that if the total number of nodes of a SP grows the computational efficiency of the CS algorithm becomes significantly better than the Dijkstra algorithm, and that the CS approach has the best performance. 9. The experimental result for the E-pointer reveals that the computational efficiency of the CS algorithm will decrease gradually as the number of selected CSs increases. Therefore, the total percentage of selected CSs suggested by the experimental result is no more than 30 percent. 10. According to the experimental results, the performance order among SDP, LNP and MTP algorithms from fast to slow is SDP, MTP and LNP, and the LNP algorithm requires much more time than the other two algorithms. 11. As the total nodes of a path increase, most of the run time for SDP and LNP also increases. However, there are still some paths that violate the above rule. This result verifies that the run time needed for solving SDP and LNP is not only affected by the node numbers but also depends on the network topology. 12. Run time for solving the MOPO problem by applying the PGA is faster than applying the Dijkstra algorithm, if the run time of the former algorithm does not take into account the population initialisation time. Nevertheless, if the run time of the former algorithm does take into account the population initialisation time, the latter algorithm is much faster than the former algorithm. 13. In comparing the run time for population initialisation, the run time of the evolution process by applying the PGA is quite small, and the bottleneck of the run time for solving MOPO problem by applying the PGA is the population initialisation. 14. The population initialisation time is reduced significantly by applying the SPE algorithm, and increases at a very slow rate as the number of nodes of a path increases. As the total nodes of a path grow ever larger, the computing time is reduced more noticeably.

This thesis presents an indoor obstacle avoidance system for car-like mobile robot. The system consists of stereo vision, map building, and path planning. Stereo vision is performed on stereo images to create a geometric map of the environment. A fast sparse stereo approach is employed. For different areas of the image there are different optimal values of disparity range. A multi-pass method to combine results at different disparity range is proposed. To reduce computational complexity the matching is limited to areas that are likely to generate useful data. The stereo vision system outputs a more complete disparity map. Abstract Map building involves converting the disparity map into map coordinates using triangulation and generating a list of obstacles. Occupancy grids are built to aid a hierarchical collision detection. The fast collision detection method is used by the path planner. Abstract A steering set path planner calculates a path that can be directly used by a car-like mobile robot. An adaptive approach using occupancy grid information is proposed to improve efficiency. Using a non-fixed steering set the path planner spends less computation time in areas away from obstacles. The path planner populates a discrete tree to generate a smooth path. Two tree population methods were trialled to execute the path planner. The methods are implemented and experimented on a real car-like mobile robot.