Genetic algorithm optimization of adaptive multi-scale GLCM features

We introduce a new second-order method of texture analysis called Adaptive Multi-Scale Grey Level Co-occurrence Matrix (AMSGLCM), based on the well-known Grey Level Co-occurrence Matrix (GLCM) method. The method deviates significantly from GLCM in that features are extracted, not via a fixed 2D weighting function of co-occurrence matrix elements, but by a variable summation of matrix elements in 3D localized neighborhoods. We subsequently present a new methodology for extracting optimized, highly discriminant features from these localized areas using adaptive Gaussian weighting functions. Genetic Algorithm (GA) optimization is used to produce a set of features whose classification worth is evaluated by discriminatory power and feature correlation considerations. We critically appraised the performance of our method and GLCM in pairwise classification of images from visually similar texture classes, captured from Markov Random Field (MRF) synthesized, natural, and biological origins. In these cross-validated classification trials, our method demonstrated significant benefits over GLCM, including increased feature discriminatory power, automatic feature adaptability, and significantly improved classification performance.

A tutorial on the cross-entropy method

The cross-entropy (CE) method is a new generic approach to combinatorial and multi-extremal optimization and rare event simulation. The purpose of this tutorial is to give a gentle introduction to the CE method. We present the CE methodology, the basic algorithm and its modifications, and discuss applications in combinatorial optimization and machine learning. combinatorial optimization

Application of the cross-entropy method to the buffer allocation problem in a simulation-based environment

The buffer allocation problem (BAP) is a well-known difficult problem in the design of production lines. We present a stochastic algorithm for solving the BAP, based on the cross-entropy method, a new paradigm for stochastic optimization. The algorithm involves the following iterative steps: (a) the generation of buffer allocations according to a certain random mechanism, followed by (b) the modification of this mechanism on the basis of cross-entropy minimization. Through various numerical experiments we demonstrate the efficiency of the proposed algorithm and show that the method can quickly generate (near-)optimal buffer allocations for fairly large production lines.

Microstructure of MmM(5)/Mg multi-layer films prepared by magnetron sputtering

Microstructure of MmNi(3.5)(CoAlMn)(1.5)/Mg (here Mm denotes La-rich mischmetal) multi-layer hydrogen storage thin films prepared by direct current magnetron sputtering was investigated by cross-sectional transmission electron microscopy (XTEM). It was shown that the MMM5 layers are composed of two regions: an amorphous region with a thickness of similar to 4nm at the bottom of the layers and a randomly orientated nanocrystallite region on the top of the amorphous region and the Mg layers consist of typical columnar crystallite with their [001] direction nearly parallel to the growth direction. The mechanism for the formation of the above microstructure characteristics in the multi-layer thin films has been proposed. Based on the microstructure feature of the multi-layer films, mechanism for the apparent improvement of hydrogen absorption/desorption kinetics was discussed. (c) 2005 Elsevier B.V. All rights reserved.

Optimization of slitlike carbon nanopores for storage of hythane fuel at ambient temperatures

Carbons with slitlike pores can serve as effective host materials for storage of hythane fuel, a bridge between the petrol combustion and hydrogen fuel cells. We have used grand canonical Monte Carlo simulation for the modeling of the hydrogen and methane mixture storage at 293 K and pressure of methane and hydrogen mixture up to 2 MPa. We have found that these pores serve as efficient vessels for the storage of hythane fuel near ambient temperatures and low pressures. We find that, for carbons having optimized slitlike pores of size H congruent to 7 angstrom ( pore width that can accommodate one adsorbed methane layer), and bulk hydrogen mole fraction >= 0.9, the volumetric stored energy exceeds the 2010 target of 5.4 MJ dm(-3) established by the U. S. FreedomCAR Partnership. At the same condition, the content of hydrogen in slitlike carbon pores is congruent to 7% by energy. Thus, we have obtained the composition corresponding to hythane fuel in carbon nanospaces with greatly enhanced volumetric energy in comparison to the traditional compression method. We proposed the simple system with added extra container filled with pure free/adsorbed methane for adjusting the composition of the desorbed mixture as needed during delivery. Our simulation results indicate that light slit pore carbon nanomaterials with optimized parameters are suitable filling vessels for storage of hythane fuel. The proposed simple system consisting of main vessel with physisorbed hythane fuel, and an extra container filled with pure free/adsorbed methane will be particularly suitable for combustion of hythane fuel in buses and passenger cars near ambient temperatures and low pressures.

Optimization of microcavity OLED by varying the thickness of multi-layered mirror

We optimized the emission efficiency from a microcavity OLEDs consisting of widely used organic materials, N,N'-di(naphthalene-1-yl)-N,N'-diphenylbenzidine (NPB) as a hole transport layer and tris (8-hydroxyquinoline) (Alq(3)) as emitting and electron transporting layer. LiF/Al was considered as a cathode, while metallic Ag anode was used. TiO2 and Al2O3 layers were stacked on top of the cathode to alter the properties of the top mirror. The electroluminescence emission spectra, electric field distribution inside the device, carrier density, recombination rate and exciton density were calculated as a function of the position of the emission layer. The results show that for certain TiO2 and Al2O3 layer thicknesses, light output is enhanced as a result of the increase in both the reflectance and transmittance of the top mirror. Once the optimum structure has been determined, the microcavity OLED devices can be fabricated and characterized, and comparisons between experiments and theory can be made.

Physical and MAC layer techniques for next generation satellite communications

In this thesis, we consider four different scenarios of interest in modern satellite communications. For each scenario, we will propose the use of advanced solutions aimed at increasing the spectral efficiency of the communication links. First, we will investigate the optimization of the current standard for digital video broadcasting. We will increase the symbol rate of the signal and determine the optimal signal bandwidth. We will apply the time packing technique and propose a specifically design constellation. We will then compare some receiver architectures with different performance and complexity. The second scenario still addresses broadcast transmissions, but in a network composed of two satellites. We will compare three alternative transceiver strategies, namely, signals completely overlapped in frequency, frequency division multiplexing, and the Alamouti space-time block code, and, for each technique, we will derive theoretical results on the achievable rates. We will also evaluate the performance of said techniques in three different channel models. The third scenario deals with the application of multiuser detection in multibeam satellite systems. We will analyze a case in which the users are near the edge of the coverage area and, hence, they experience a high level of interference from adjacent cells. Also in this case, three different approaches will be compared. A classical approach in which each beam carries information for a user, a cooperative solution based on time division multiplexing, and the Alamouti scheme. The information theoretical analysis will be followed by the study of practical coded schemes. We will show that the theoretical bounds can be approached by a properly designed code or bit mapping. Finally, we will consider an Earth observation scenario, in which data is generated on the satellite and then transmitted to the ground. We will study two channel models, taking into account one or two transmit antennas, and apply techniques such as time and frequency packing, signal predistortion, multiuser detection and the Alamouti scheme.

Representing spray zone with cross flow as a well-mixed compartment in a high shear granulator

The spray zone is an important region to control nucleation of granules in a high shear granulator. In this study, a spray zone with cross flow is quantified as a well-mixed compartment in a high shear granulator. Granulation kinetics is quantitatively derived at both particle-scale and spray zone-scale. Two spatial decay rates, DGSDR (droplet-granule spatial decay rate) ζDG and DPSDR (droplet-primary particle spatial decay rate) ζDP, which are functions of volume fraction and diameter of particulate species within the powder bed, are defined to simplify the deduction. It is concluded that in cross flow, explicit analytical results show that the droplet concentration is subject to exponential decay with depth which produces a numerically infinite depth of spray zone in a real penetration process. In a well-mixed spray zone, the depth of the spray zone is 4/(ζDG + ζDP) and π2/3(ζDG + ζDP) in cuboid and cylinder shape, respectively. The first-order droplet-based collision rates of, nucleation rate B0 and rewetting rate RW0 are uncorrelated with the flow pattern and shape of the spray zone. The second-order droplet-based collision rate, nucleated granule-granule collision rate RGG, is correlated with the mixing pattern. Finally, a real formulation case of a high shear granulation process is used to estimate the size of the spray zone. The results show that the spray zone is a thin layer at the powder bed surface. We present, for the first time, the spray zone as a well-mixed compartment. The granulation kinetics of a well-mixed spray zone could be integrated into a Population Balance Model (PBM), particularly to aid development of a distributed model for product quality prediction.

Resource management and optimization in wireless mesh networks

A wireless mesh network is a mesh network implemented over a wireless network system such as wireless LANs. Wireless Mesh Networks(WMNs) are promising for numerous applications such as broadband home networking, enterprise networking, transportation systems, health and medical systems, security surveillance systems, etc. Therefore, it has received considerable attention from both industrial and academic researchers. This dissertation explores schemes for resource management and optimization in WMNs by means of network routing and network coding.^ In this dissertation, we propose three optimization schemes. (1) First, a triple-tier optimization scheme is proposed for load balancing objective. The first tier mechanism achieves long-term routing optimization, and the second tier mechanism, using the optimization results obtained from the first tier mechanism, performs the short-term adaptation to deal with the impact of dynamic channel conditions. A greedy sub-channel allocation algorithm is developed as the third tier optimization scheme to further reduce the congestion level in the network. We conduct thorough theoretical analysis to show the correctness of our design and give the properties of our scheme. (2) Then, a Relay-Aided Network Coding scheme called RANC is proposed to improve the performance gain of network coding by exploiting the physical layer multi-rate capability in WMNs. We conduct rigorous analysis to find the design principles and study the tradeoff in the performance gain of RANC. Based on the analytical results, we provide a practical solution by decomposing the original design problem into two sub-problems, flow partition problem and scheduling problem. (3) Lastly, a joint optimization scheme of the routing in the network layer and network coding-aware scheduling in the MAC layer is introduced. We formulate the network optimization problem and exploit the structure of the problem via dual decomposition. We find that the original problem is composed of two problems, routing problem in the network layer and scheduling problem in the MAC layer. These two sub-problems are coupled through the link capacities. We solve the routing problem by two different adaptive routing algorithms. We then provide a distributed coding-aware scheduling algorithm. According to corresponding experiment results, the proposed schemes can significantly improve network performance.^

Pitting and porous layer formation on n-InP anodes

Surface pitting occurs when InP electrodes are anodized in KOH electrolytes at concentrations in the range 2 - 5 mol dm-3. The process has been investigated using atomic force microscopy (AFM) and the results correlated with cross-sectional transmission electron microscopy (TEM) and electroanalytical measurements. AFM measurements show that pitting of the surface occurs and the density of pits is observed to increase with time under both potentiodynamic and potentiostatic conditions. This indicates a progressive pit nucleation process and implies that the development of porous domains beneath the surface is also progressive in nature. Evidence for this is seen in plan view TEM images in which individual domains are seen to be at different stages of development. Analysis of the cyclic voltammograms of InP electrodes in 5 mol dm-3 KOH indicates that, above a critical potential for pit formation, the anodic current is predominantly time dependent and there is little differential dependence of the current on potential. Thus, pores continue to grow with time when the potential is high enough to maintain depletion layer breakdown conditions.

Potential Shifts in Canadian High Arctic Sedimentary Organic Matter Composition With Permafrost Active Layer Detachments

Increased temperature and precipitation in Arctic regions have led to deeper thawing and structural instability in permafrost soil. The resulting localized disturbances, referred to as active layer detachments (ALDs), may transport organic matter (OM) to more biogeochemically active zones. To examine this further, solid state cross polarization magic angle spinning 13C nuclear magnetic resonance (CPMAS NMR) and biomarker analysis were used to evaluate potential shifts in riverine sediment OM composition due to nearby ALDs within the Cape Bounty Arctic Watershed Observatory, Nunavut, Canada. In sedimentary OM near ALDs, NMR analysis revealed signals indicative of unaltered plant-derived material, likely derived from permafrost. Long chain acyclic aliphatic lipids, steroids, cutin, suberin and lignin occurred in the sediments, consistent with a dominance of plant-derived compounds, some of which may have originated from permafrost-derived OM released by ALDs. OM degradation proxies for sediments near ALDs revealed less alteration in acyclic aliphatic lipids, while constituents such as steroids, cutin, suberin and lignin were found at a relatively advanced stage of degradation. Phospholipid fatty acid analysis indicated that microbial activity was higher near ALDs than downstream but microbial substrate limitation was prevalent within disturbed regions. Our study suggests that, as these systems recover from disturbance, ALDs likely provide permafrost-derived OM to sedimentary environments. This source of OM, which is enriched in labile OM, may alter biogeochemical patterns and enhance microbial respiration within these ecosystems.

Effects of the thickness of cross-laminated timber (CLT) panels made from Irish Sitka spruce on mechanical performance in bending and shear

An investigation was carried out on CLT panels made from Sitka spruce in order to establish the effect of the thickness of CLT panels on the bending stiffness and strength and the rolling shear. Bending and shear tests on 3-layer and 5-layer panels were performed with loading in the out-of-plane and in-plane directions. ‘Global’ stiffness measurements were found to correlate well with theoretical values. Based on the results, there was a general tendency that both the bending strength and rolling shear decreased with panel thickness. Mean values for rolling shear ranged from 1.0 N/mm2 to 2.0 N/mm2 .

A comparison of sound localisation techniques using cross-correlation and spiking neural networks for mobile robotics

This paper outlines the development of a crosscorrelation algorithm and a spiking neural network (SNN) for sound localisation based on real sound recorded in a noisy and dynamic environment by a mobile robot. The SNN architecture aims to simulate the sound localisation ability of the mammalian auditory pathways by exploiting the binaural cue of interaural time difference (ITD). The medial superior olive was the inspiration for the SNN architecture which required the integration of an encoding layer which produced biologically realistic spike trains, a model of the bushy cells found in the cochlear nucleus and a supervised learning algorithm. The experimental results demonstrate that biologically inspired sound localisation achieved using a SNN can compare favourably to the more classical technique of cross-correlation.

Multi-physics modelling for design optimization and manufacturing on high-performance parallel systems

Abstract not available

A finite-volume shallow layer method for the MHD instabilities in an aluminium production cell

An electrolytic cell for Aluminium production contains molten metal and molten electrolyte, which are subject to high dc-currents and magnetic fields. Lorentz forces arising from the cross product of current and magnetic field may amplify natural gravity waves at the interface between the two fluids, leading to short circuits in extreme cases. The external magnetic field and current distribution in the production cell is computed through a detailed finite element analysis at Torino Polytechnic. The results are then used to compute the magnetohydrodynamic and thermal effects in the aluminium/electrolyte bath. Each cell has lateral dimensions of 6m x 2m, whilst the bath depth is only 30cm. the electrically resistive electrolyte path, which is critical in the operation of the cell, has layer depth of only a few centimetres below each carbon anode. Because the shallow dimensions of the liquid layer a finite-volume shallow-layer technique has been used at Greenwich to compute the resulting flow-field and interface perturbations. The information obtained from this method, i.e. depth averaged velocities and aluminium/electrolyte interface position is then embedded in the three-dimensional finite volume code PHYSICA and will be used to compute the heat transfer and phase change in the cell.