Despite the popularity and success of neural networks in research, the number of resulting commercial or industrial applications have been limited. A primary cause of this lack of adoption is due to the fact that neural networks are usually implemented as software running on general-purpose processors. Hence, a hardware implementation that can take advantage of the inherent parallelism in neural networks is desired. This thesis investigates how the Restricted Boltzmann machine, a popular type of neural network, can be effectively mapped to a high-performance hardware architecture on FPGA platforms. The proposed, modular framework is designed to reduce the time complexity of the computations through heavily customized hardware engines. The framework is tested on a platform of four Xilinx Virtex II-Pro XC2VP70 FPGAs running at 100MHz through a variety of different configurations. The maximum performance was obtained by instantiating a Restricted Boltzmann Machine of 256x256 nodes distributed across four FPGAs, which results in a computational speed of 3.13 billion connection-updates-per-second and a speed-up of 145-fold over an optimized C program running on a 2.8GHz Intel processor.

A new general method is developed for the optimization of HPLC ternary or pseudo-quaternary mobile phases which are represented by the trilinear coordinate system. This method can predict the global optimum of the mobile phase composition. The global optimum composition along each edge of the triangle and the corresponding selectivity factor of the worst-separated peak pair(s) are used in this method. This method is named the weighted pattern comparison optimization method (WPCO) and is applicable for both known and unknown samples. The WPCO method is simpler than those currently in use. The WPCO method was tested by using 68 literature data sets whose separation response surfaces are different. Results of the WPCO method agree with the results obtained by the minimum α plot method and by the grid search method, and do so with substantially fewer experimental measurements. Compared with the 5% (in eluent composition) step size grid-search procedure, the WPCO method using the same step size reduces the experimental work by 75%. For further reducing the experimental work, the original WPCO method is simplified. In an ordinary HPLC separation, the separation factor and resolution are approximately proportional to the logarithm of the selectivity factor. Based on this, the separation factor replaces the logarithm of selectivity factor in the original WPCO method. This further reduces the experimental work and avoids the error introduced in the measurement of the column dead volume. The simplified WPCO method has been tested in the normal-phase and reversed-phase chromatography separation cases. The simplified WPCO method has been tested by using 27 literature data sets whose separation response surfaces are different. Results of the simplified and original WPCO methods are nearly identical when the capacity factors of the solutes of the worst-separated peak pairs are greater than 5. When the capacity factors are less than 5, the simplified WPCO method is satisfactory in less complex, less critical applications. Two new metabolites of nefopam have been separated from greyhound urine. In the separation process, flash chromatography is used for cleaning up and preseparating the samples in a single step. Compared with other techniques, experimental work is reduced. The structure of one of the newly discovered metabolites is determined using MS and NMR. The most probable structure of the other metabolite is determined using MS. The main metabolic pathways at different doses in greyhounds are studied. / Graduation date: 1991

Despite the popularity and success of neural networks in research, the number of resulting commercial or industrial applications have been limited. A primary cause of this lack of adoption is due to the fact that neural networks are usually implemented as software running on general-purpose processors. Hence, a hardware implementation that can take advantage of the inherent parallelism in neural networks is desired. This thesis investigates how the Restricted Boltzmann machine, a popular type of neural network, can be effectively mapped to a high-performance hardware architecture on FPGA platforms. The proposed, modular framework is designed to reduce the time complexity of the computations through heavily customized hardware engines. The framework is tested on a platform of four Xilinx Virtex II-Pro XC2VP70 FPGAs running at 100MHz through a variety of different configurations. The maximum performance was obtained by instantiating a Restricted Boltzmann Machine of 256x256 nodes distributed across four FPGAs, which results in a computational speed of 3.13 billion connection-updates-per-second and a speed-up of 145-fold over an optimized C program running on a 2.8GHz Intel processor.

The corrosion of the reinforcing steel in concrete by de-icing salts is one of the major issues concerning the durability of reinforced concrete. Different methods have been used to protect the reinforcing steel, but still corrosion of reinforced structures continues to be a big problem causing enormous costs in their restoration and rehabilitation. The continuity of the pores of concrete plays a crucial role in the corrosion of the reinforcing steel. The ingress of corrosive species, such as chloride ions, oxygen and water, through the pores of the concrete cover cause the breakdown of the passive layer formed on the steel by the high pH of the concrete. The use of supplementary cementitious materials (SCM) in the production of high performance concrete (HPC) improves its resistance to corrosive species as a result of the pozzolanic reaction which forms more calcium silicate hydrates (C-S-H). Most of the studies about the corrosion of the reinforcement in HPC have been carried out in sound concrete. However, very few works have been reported on the corrosion of steel in cracked concrete. The crack pattern on HPC is very distinct from that formed on ordinary portland cement (OPC) concrete, which may result in different corrosion mechanisms of the reinforcing steel. The objective of the present work consisted in the evaluation of the corrosion of reinforcing steel in cracked HPC and OPC concrete under different exposure and loading conditions. For that purpose, two sets of beams of HPC (containing fly ash or slag) and two sets of OPC concrete were cast. The difference between the OPC concretes was the date of casting. Three sets of reinforcing steel probes were embedded in each beam at different locations. All the beams were cracked at midspan by the four-point method. Eight beams of each concrete were coupled in pairs and partially immersed in a solution of de-icing salts every two weeks. In this way, one set of the corrosion probes was non-submerged (top) while the other two were completely submerged (one at the crack level and the other at the bottom). Two pairs of beams were subjected to static loading whereas the other two were under cyclic loading. The corrosion potentials readings were taken daily by a data acquisition system, whereas the corrosion rates were determined by the Linear Polarization technique using a corrosion monitoring system. According to the results obtained, the corrosion rates of the submerged and non-submerged probes are very low. This behaviour is observed for the four concretes and for both loading conditions. The type of loading did not influence the corrosion rates of these probes, which were in the same range for all the concretes. On the other hand, the probes close to the crack showed higher corrosion rates, especially those under cyclic loading. In general, the OPC concrete cast during the winter presented the highest corrosion rates for both loading conditions, followed by the OPC concrete cast in the summer (as were both HPCs), then by HPC-Slag and HPC-Fly Ash, which showed the lowest values. In most of the cases there was a good agreement between the corrosion potentials and the corrosion rates, so that the OPC concretes exhibited the most negative values. The lower corrosion of the probes in the HPC-Fly Ash and HPC-Slag beams was ascribed to the continued pozzolanic activity, which may result in the self-healing of the crack with time. The probes close to the crack in the dynamically loaded beams experienced higher corrosion than those in the static beams. In some cases the corrosion rate reached values above 100 mm/year. The lower corrosion of the probes in the static beams was attributed to the self-healing of the crack. The formation of additional microcracks in the dynamic beams during cyclic loading may be responsible for their higher corrosion. The corrosion potential of the rebar cage shifted to more negative values during cyclic and static loading. This change in the potential was associated with stress concentration of the reinforcement surface, making it more active. Although the shift in the potential was not really significant, this may have important consequences in practice where the concrete is subjected to higher loads.

The work presented in this thesis deals with the development of self-consolidating high performance concrete (SCHPC) incorporating rice husk ash (RHA) as a supplementary cementing material. Various SCHPCs were produced using the water-binder (W/B) ratios of 0.30, 0.35, 0.40 and 0.50, and RHA content in the range of 0 to 30% of cement by weight. In addition, a number of pastes and mortars formulated from the concretes were prepared and tested for the filling ability. The paste and mortar filling abilities were tested with respect to flow time and flow spread, respectively, at various dosages of high-range water reducer (HRWR). Also, the mortars were tested for the air content at various dosages of air-entraining admixture (AEA). It was observed that the flow time of the pastes increased with lower W/B ratio and higher RHA content, whereas the flow spread of the mortars decreased with higher W/B ratio and greater RHA content. Both paste and mortar filling abilities increased with higher HRWR dosages. In addition, the air content of the mortars decreased with lower W/B ratio and higher RHA content for given AEA dosages. The fresh SCHPCs were tested for filling ability, passing ability, air-void stability, segregation resistance, unit weight and air content. The filling ability was determined with respect to slump and slump flow, inverted slump cone flow time and spread, and orimet flow time and spread. The passing ability was measured with regard to slump and slump flow with J-ring, inverted slump cone flow spread with J-ring, and orimet flow spread with J-ring. The air-void stability in several fresh SCHPC mixtures was investigated with respect to re-mixing of concrete and subsequent measurement of air content at different test stages. The test results obtained for the fresh properties showed that the inverted slump cone and orimet flow times increased with lower W/B ratio and greater RHA content. In addition, the slump flow, inverted slump cone flow spread, and orimet flow spread with and without J-ring increased considerably with lower W/B ratio and greater RHA content. However, the increases in slump with and without J-ring at lower W/B ratio and higher RHA content were not significant. The unit weight of concrete slightly decreased with higher W/B ratio and greater RHA content, and with higher air content. Achieving the target air content required greater AEA dosages for lower W/B ratio and higher RHA content. However, the presence of RHA had no adverse effect on the air-void stability of concrete. The segregation resistance of various SCHPCs was investigated by visual inspection of concrete in mixer pan, and during and after different flow tests. Slight bleeding and a thick layer of paste were noticed in mixer pan for several concretes. The dynamic segregation in the form of discontinuity or blockage of flow did not occur during the orimet and inverted slump cone flow tests for any concrete. No aggregate pile appeared in the slump flow, and orimet and inverted slump cone flow spreads of any concrete. But minor to severe mortar halos were noticed in the periphery of the flow spread of several concretes, particularly in the presence of high RHA content. The results of visual inspection suggest that both lower W/B ratio and greater RHA content improved the dynamic segregation resistance of concrete. In contrast, the higher RHA content resulted in a lower static segregation resistance, which was overcome in the presence of viscosity-enhancing admixture (VEA). The static segregation resistance of several SCHPCs was quantitatively determined by sieve and column apparatus. The segregation index given by the sieve increased with lower W/B ratio and higher RHA content, thus indicating a reduced static segregation resistance. In contrast, the segregation factor given by the column apparatus decreased with lower W/B ratio suggesting an increased static segregation resistance. However, the segregation factor increased with higher RHA content, and thus revealed a reduction in static segregation resistance. In the presence of VEA, both segregation index and segregation factor decreased significantly, indicating an improvement in the static segregation resistance of concrete. The hardened SCHPCs were tested for compressive strength, ultrasonic pulse velocity, water absorption, total porosity and electrical resistivity. Test results revealed that the compressive strength, ultrasonic pulse velocity and true electrical resistivity increased, whereas the water absorption and total porosity decreased with lower W/B ratio and higher RHA content. The entrained air-voids decreased the compressive strength, ultrasonic pulse velocity, water absorption and total porosity, but slightly increased the electrical resistivity of concrete. In general, the hardened properties indicated good durability of the concretes. The empirical models for the filling ability (slump flow) and compressive strength of SCHPC were derived and verified with test data from this study and other data taken from the literature. The slump flow and compressive strength computed from the models were coherent with the measured values. Both filling ability and strength models were useful to develop a mixture design method for SCHPC with and without RHA.

In communication systems, the need for high bandwidth interconnects and efficient distribution of large amount of data is very essential. This thesis work addresses all-optical packet switching issues in the field of reconfigurable optical interconnection networks for high performance embedded systems. The recent research conducted at the Halmstad University, on high performance embedded systems, focuses on the optical interconnection techniques to achieve ultra high throughputs and reconfigurability at the system level. Recent research in the field of optical interconnection networks for applications like switches and routers for data and telecommunication industry and parallel computing architectures for embedded signal processing use optical to electrical conversion to switch packets. This conversion scales down the enormous bandwidth capacity of the optical communication channels to electronic processing rates. To maintain the high throughputs all over the interconnection networks, the optical packets need to be maintained in optical state and switched to different part of the interconnection network. To achieve this goal, all-optical packet switching architectures are studied. The study is concluded with a positive outlook towards alloptical switching technologies, and it will play a very important role in the near future in the field of optical communication, telecommunication and embedded systems.

The corrosion of the reinforcing steel in concrete by de-icing salts is one of the major issues concerning the durability of reinforced concrete. Different methods have been used to protect the reinforcing steel, but still corrosion of reinforced structures continues to be a big problem causing enormous costs in their restoration and rehabilitation. The continuity of the pores of concrete plays a crucial role in the corrosion of the reinforcing steel. The ingress of corrosive species, such as chloride ions, oxygen and water, through the pores of the concrete cover cause the breakdown of the passive layer formed on the steel by the high pH of the concrete. The use of supplementary cementitious materials (SCM) in the production of high performance concrete (HPC) improves its resistance to corrosive species as a result of the pozzolanic reaction which forms more calcium silicate hydrates (C-S-H). Most of the studies about the corrosion of the reinforcement in HPC have been carried out in sound concrete. However, very few works have been reported on the corrosion of steel in cracked concrete. The crack pattern on HPC is very distinct from that formed on ordinary portland cement (OPC) concrete, which may result in different corrosion mechanisms of the reinforcing steel. The objective of the present work consisted in the evaluation of the corrosion of reinforcing steel in cracked HPC and OPC concrete under different exposure and loading conditions. For that purpose, two sets of beams of HPC (containing fly ash or slag) and two sets of OPC concrete were cast. The difference between the OPC concretes was the date of casting. Three sets of reinforcing steel probes were embedded in each beam at different locations. All the beams were cracked at midspan by the four-point method. Eight beams of each concrete were coupled in pairs and partially immersed in a solution of de-icing salts every two weeks. In this way, one set of the corrosion probes was non-submerged (top) while the other two were completely submerged (one at the crack level and the other at the bottom). Two pairs of beams were subjected to static loading whereas the other two were under cyclic loading. The corrosion potentials readings were taken daily by a data acquisition system, whereas the corrosion rates were determined by the Linear Polarization technique using a corrosion monitoring system. According to the results obtained, the corrosion rates of the submerged and non-submerged probes are very low. This behaviour is observed for the four concretes and for both loading conditions. The type of loading did not influence the corrosion rates of these probes, which were in the same range for all the concretes. On the other hand, the probes close to the crack showed higher corrosion rates, especially those under cyclic loading. In general, the OPC concrete cast during the winter presented the highest corrosion rates for both loading conditions, followed by the OPC concrete cast in the summer (as were both HPCs), then by HPC-Slag and HPC-Fly Ash, which showed the lowest values. In most of the cases there was a good agreement between the corrosion potentials and the corrosion rates, so that the OPC concretes exhibited the most negative values. The lower corrosion of the probes in the HPC-Fly Ash and HPC-Slag beams was ascribed to the continued pozzolanic activity, which may result in the self-healing of the crack with time. The probes close to the crack in the dynamically loaded beams experienced higher corrosion than those in the static beams. In some cases the corrosion rate reached values above 100 mm/year. The lower corrosion of the probes in the static beams was attributed to the self-healing of the crack. The formation of additional microcracks in the dynamic beams during cyclic loading may be responsible for their higher corrosion. The corrosion potential of the rebar cage shifted to more negative values during cyclic and static loading. This change in the potential was associated with stress concentration of the reinforcement surface, making it more active. Although the shift in the potential was not really significant, this may have important consequences in practice where the concrete is subjected to higher loads.

The work presented in this thesis deals with the development of self-consolidating high performance concrete (SCHPC) incorporating rice husk ash (RHA) as a supplementary cementing material. Various SCHPCs were produced using the water-binder (W/B) ratios of 0.30, 0.35, 0.40 and 0.50, and RHA content in the range of 0 to 30% of cement by weight. In addition, a number of pastes and mortars formulated from the concretes were prepared and tested for the filling ability. The paste and mortar filling abilities were tested with respect to flow time and flow spread, respectively, at various dosages of high-range water reducer (HRWR). Also, the mortars were tested for the air content at various dosages of air-entraining admixture (AEA). It was observed that the flow time of the pastes increased with lower W/B ratio and higher RHA content, whereas the flow spread of the mortars decreased with higher W/B ratio and greater RHA content. Both paste and mortar filling abilities increased with higher HRWR dosages. In addition, the air content of the mortars decreased with lower W/B ratio and higher RHA content for given AEA dosages. The fresh SCHPCs were tested for filling ability, passing ability, air-void stability, segregation resistance, unit weight and air content. The filling ability was determined with respect to slump and slump flow, inverted slump cone flow time and spread, and orimet flow time and spread. The passing ability was measured with regard to slump and slump flow with J-ring, inverted slump cone flow spread with J-ring, and orimet flow spread with J-ring. The air-void stability in several fresh SCHPC mixtures was investigated with respect to re-mixing of concrete and subsequent measurement of air content at different test stages. The test results obtained for the fresh properties showed that the inverted slump cone and orimet flow times increased with lower W/B ratio and greater RHA content. In addition, the slump flow, inverted slump cone flow spread, and orimet flow spread with and without J-ring increased considerably with lower W/B ratio and greater RHA content. However, the increases in slump with and without J-ring at lower W/B ratio and higher RHA content were not significant. The unit weight of concrete slightly decreased with higher W/B ratio and greater RHA content, and with higher air content. Achieving the target air content required greater AEA dosages for lower W/B ratio and higher RHA content. However, the presence of RHA had no adverse effect on the air-void stability of concrete. The segregation resistance of various SCHPCs was investigated by visual inspection of concrete in mixer pan, and during and after different flow tests. Slight bleeding and a thick layer of paste were noticed in mixer pan for several concretes. The dynamic segregation in the form of discontinuity or blockage of flow did not occur during the orimet and inverted slump cone flow tests for any concrete. No aggregate pile appeared in the slump flow, and orimet and inverted slump cone flow spreads of any concrete. But minor to severe mortar halos were noticed in the periphery of the flow spread of several concretes, particularly in the presence of high RHA content. The results of visual inspection suggest that both lower W/B ratio and greater RHA content improved the dynamic segregation resistance of concrete. In contrast, the higher RHA content resulted in a lower static segregation resistance, which was overcome in the presence of viscosity-enhancing admixture (VEA). The static segregation resistance of several SCHPCs was quantitatively determined by sieve and column apparatus. The segregation index given by the sieve increased with lower W/B ratio and higher RHA content, thus indicating a reduced static segregation resistance. In contrast, the segregation factor given by the column apparatus decreased with lower W/B ratio suggesting an increased static segregation resistance. However, the segregation factor increased with higher RHA content, and thus revealed a reduction in static segregation resistance. In the presence of VEA, both segregation index and segregation factor decreased significantly, indicating an improvement in the static segregation resistance of concrete. The hardened SCHPCs were tested for compressive strength, ultrasonic pulse velocity, water absorption, total porosity and electrical resistivity. Test results revealed that the compressive strength, ultrasonic pulse velocity and true electrical resistivity increased, whereas the water absorption and total porosity decreased with lower W/B ratio and higher RHA content. The entrained air-voids decreased the compressive strength, ultrasonic pulse velocity, water absorption and total porosity, but slightly increased the electrical resistivity of concrete. In general, the hardened properties indicated good durability of the concretes. The empirical models for the filling ability (slump flow) and compressive strength of SCHPC were derived and verified with test data from this study and other data taken from the literature. The slump flow and compressive strength computed from the models were coherent with the measured values. Both filling ability and strength models were useful to develop a mixture design method for SCHPC with and without RHA.

Economic and political environment are rapidly changing and increasingly becoming fierce international competition in recent years. So companies must change their organizational structure to adapt to the competitive environment and the pursuit of high-profit governance policies. Study has shown that financial neighboring services (which is also called leasing industry) layoffs by organizational changes are rising from the uncertainty and psychological contract breach and damage. There for high-performance doctrine of internal staff and high turnover attitude show Leasing Services-oriented by leaps and bounds that impact competitive development, especially in the service competitiveness of Mainland China leasing industry¡C According to different personal attributes that explore the differences of employees in all aspects of this study. The Financial service workers that where interview for this research over the years, leasing and verify the research aspects of banking sector progress of human relationship. In conclusion with this research I try to understand how people work. We are divided to how each of us work and how long we stay to get the job done¡C

A methodology to develop an easy-to-use toolkit for the preliminary design of high performance schools in hot and humid climates was presented. The toolkit proposed in this research will allow decision makers without simulation knowledge easily to evaluate accurately energy efficient measures for K-5 schools, which would contribute to the accelerated dissemination of energy efficient design. For the development of the toolkit, first, a survey was performed to identify high performance measures available today being implemented in new K-5 school buildings. Then an existing case-study school building in a hot and humid climate was selected and analyzed to understand the energy use pattern in a school building and to be used in developing a calibrated simulation. Based on the information from the previous step, an as-built and calibrated simulation was then developed. To accomplish this, five calibration steps were performed to match the simulation results with the measured energy use. The five steps include: 1) Using an actual 2006 weather file with measured solar radiation, 2) Modifying lighting & equipment schedule using ASHRAE's RP-1093 methods, 3) Using actual equipment performance curves (i.e., scroll chiller), 4) Using the Winkelmann's method for the underground floor heat transfer, and 5) Modifying the HVAC and room setpoint temperature based on the measured field data. Next, the calibrated simulation of the case-study K-5 school was compared to an ASHRAE Standard 90.1-1999 code-compliant school. In the next step, the energy savings potentials from the application of several high performance measures to an equivalent ASHRAE Standard 90.1-1999 codecompliant school. The high performance measures applied included the recommendations from the ASHRAE Advanced Energy Design Guides (AEDG) for K- 12 and other high performance measures from the literature review as well as a daylighting strategy and solar PV and thermal systems. The results show that the net energy consumption of the final high performance school with the solar thermal and a solar PV system would be 1,162.1 MMBtu, which corresponds to the 14.9 kBtu/sqft-yr of EUI. The calculated final energy and cost savings over the code compliant school are 68.2% and 69.9%, respectively. As a final step of the research, specifications for a simplified easy-to-use toolkit were then developed, and a prototype screenshot of the toolkit was developed. The toolkit is expected to be used by non-technical decision-maker to select and evaluate high performance measures for a new school building in terms of energy and cost savings in a quick and easy way.