Search

The proton exchange membrane water electrolyser (PEMWE) is a promising technology for the production of hydrogen from water. The oxygen evolution reaction (OER) has a high over potential cf. with the hydrogen evolution reaction and is one of the main reasons for the high energy demand of the electrolyser. RuO₂ and IrO₂ are the most active catalyst for OER, but are costly, making the electrolyser system expensive. In general, it is important to use stable, active and cheap catalysts in order to make a cost efficient electrolyser system. Supporting the active catalyst on a high surface area conducting support material is one of the approaches to reduce the precious metal loading on the electrode. Antimony tin oxide (ATO) and indium tin oxide (ITO) were studied as possible support materials for IrO₂ in the PEMWE anode prepared by the Adams method. The effect of the support material on the surface area, electronic conductivity, particle size and agglomeration were investigated. The IrO₂ showed highest conductivity (4.9 S cm-¹) and surface area (112 m2 g-¹) and decreased with the decrease in the IrO₂ loading. Using the catalysts in the membrane electrode assemblies (MEA) with Nafion®-115 membranes, at 80°C showed that the catalyst with better dispersion and conductivity gave better performance. The unsupported IrO₂ and 90% IrO₂ supported on ATO and ITO showed the best performance among all the catalysts tested, achieving a cell voltage of 1.73 V at 1 A cm-². A lower IrO₂ loading decreased the conductivity and surface area. The IrO₂ particle size and bulk conductivity of the supported catalyst significantly influenced the MEA performance. Overall, it is important to maintain a conductive network of IrO₂ on the non-conducting support to maintain the bulk conductivity and thus reduce the Ohmic potential drop. Although RuO₂ is the most active catalyst for OER, it lacks stability on long term operation. RuxNb1-xO₂ and IrxNb1-xO₂ catalysts were synthesized and characterized, to try to develop stable electrodes for PEMWE. However the Adams method of catalyst synthesis formed a sodium–niobium complex making it unsuitable for preparation of Nb based catalysts. In both Adams and hydrolysis methods of synthesis, the addition of Nb ₂O₅ decreased the anodic charge and electronic conductivity of the catalyst due to the dilution of the active RuO₂. The RuO₂ catalyst showed the best performance in MEA evaluation compared to the bimetallic catalyst (1.62 V and 1.75 V @1 A cm-² for RuO₂(A) and RuO₂(H) respectively). A higher stability for bimetallic catalyst compared to the monometallic catalysts was obtained from the continuous CV cycling and MEA stability test.

In this thesis growth on patterned substrates has been studied for the lateral bandgap control of the quantum well (QW) structures, utilising indium migration from the side facets onto the adjoining (100) surfaces, leading to the possibility of integration of multi- functional optoelectronic devices. InGaAs/GaAs/AlGaAs single quantum well (SQW) lasers and InGaAs/InAlAs QW heterostructures were grown by molecular beam epitaxy (MBE) on (100) GaAs and InP substrates respectively, patterned to produce (100) mesa top surfaces with angled side facets. Chemical beam epitaxy (CBE) was used for the growth of InGaAs/InP heterostructures over InP substrates, patterned into undercut mesas with (100) top surfaces using chemically assisted ion beam etching (CAIBE). Indium migration behaviour was compared by growing two InGaAs/GaAs/AlGaAs SQW graded index separate confinement heterostructure (GRINSCH) lasers by MBE over patterned GaAs substrates. The first laser was grown using As2 throughout, whereas the active region of the second laser was grown using As4. It is observed that the use of As4 facilitates the migration process whilst the use of As2 completely stops it. However, the broad area devices of both lasers exhibit extremely low threshold current densities and very high external quantum efficiencies. Split contact devices were made by growing InGaAs/GaAs/AlGaAs SQW GRINSCH lasers by MBE over variable step width mesas patterned on GaAs substrates using a newly designed mask. Electrical measurements on partially pumped devices showed bistability or pulsation behaviour depending on the bias conditions on the unpumped section. Growth of InGaAs/InP heterostructures by CBE on undercut mesas showed facetting behaviour producing atomically flat (111)B planes. Complete triangular shape structures bounded by very smooth (111)B facets were produced on mesas as narrow as 2.0 μm. This shows the strong possibility of growing one dimensional (1D) quantum wire structures, produced completely in-situ as a direct result of the growth process.

This thesis reports some results from an experiment using a high energy muon beam incident upon a liquid hydrogen target performed at the Fermi National Accelerator Laboratory, Illinois, USA. It is solely concerned with the inclusice muon scattering distributions. The results reported are derived from the analysis of data taken at two incident momenta, 96 GeV/c and 147 GeV/c.

This thesis is composed of two separate parts. Part A deals with the investigation of the residual shear strength of stiff clays under very large dis-placement in a new ring shear apparatus. Residual shear strength from these tests was round to be lower than that obtained from conventional residual shear tests (e.g. the reversal direct shear box tests). It was found that some clays exhibit marked variation of residual shear strength, with the magnitude of effective normal stress, and that tie residual strength is independant of stress history. Some optical studios on thin sections of shear zones produced in the ring shear apparatus have also been presented. A drained tension test apparatus was developed and. it has been shown that the intact London clay has a tensile strength which is in good agreement with the modified Griffith failure criteria. The effect of sample size on the consolidation characteristics of fissured. blue London clay has been discussed in part 73 of the thesis. The results of conventional oedometer tests; dissipation tests on 1½ in., 4 in. and 12 in. diameter undisturbed samples and in-situ constant head permeability tests have been com-pared. It has been shown that the field permeability (and the field rate of consolidation) at shallow depths, where the degree of fissuring is high, can be seriously underestimated by one dimensional consolidation tests in the laboratory.

Tumor associated carbohydrate antigen (TACA) GM2 is a tetrasaccharide which is present on the surface of various cancer cells such as human melanomas, breast, ovarian, prostate and neuroblastomas which makes it an attractive target for making a carbohydrate based anti-cancer vaccine. Its chemical structure consists of N-acetyl Neuraminic acid (sialic acid), galactosamine and lactose attached to a ceramide which is in turn embedded into the surface of cancer cells. It’s chemically abbreviated as (Neu5Acα2-3[GalNAcβ1-4]Galβ1-4Glcβ1-cer). In general, TACAs are poorly immunogenic since they cannot induce strong IgG antibody responses. Therefore, they are conjugated to carrier proteins such as Keyhole limpet hemocyanin (KLH), Tetanus Toxoid (TTox) via linkers. However, these carriers and linkers are immunogenic leading to antigen suppression. In the past, GM2 has been conjugated to KLH and administered along with QS-21 as an adjuvant to patients with primary melanoma in the randomized phase III trials. However, it was stopped due to futility in terms of relapse free-survival and harmful results regarding overall survival.In the quest for alternatives to TACA-protein conjugates, our group has been making tremendous progress in making carbohydrate immunogens by conjugating TACAs to an immune stimulating zwitterionic polysaccharide PS A1 with a linker free site-specific oxime ligation strategy. We have shown Tn-PS A1, STn-PS A1, and Globo-H-PS A1 have induced robust immune responses with high antigen specific IgG antibodies exhibiting complement dependent as well as tumor cell lysis. This works describes the total synthesis of GM2 with a conjugation-ready aminooxy chemical handle at the reducing end of the carbohydrate moiety which can be conjugated to aldehyde functionalized PS A1 to make GM2-PS A1 conjugate. The key steps of the synthesis are 1) the trifluoroacetic acid one pot deprotection of the acid-labile isopropilidene acetal as well as the anomeric trimethylsilyl ethyloxy group on the trisaccharide intermediate, 2) late-stage installation of the thiophenyl leaving group for the facile glycosylation to yield the final tetrasaccharide, and 3) two-step deprotection strategy to yield β-Aminooxy GM2.

Polysaccharides represent natural biopolymers with numerous advantages over their synthetic counterparts. The application of polysaccharide-based materials in smart drug delivery systems has attracted significant interest, primarily due to their non-toxic, biocompatible, biodegradable, and water-soluble properties. A significant portion of this thesis focused on acetal modification of polysaccharides, aiming to create an acid-responsive polysaccharide-based material with potential applications in drug delivery. Moreover, this study extends to the development of smart drug delivery systems for anticancer treatment through the exploration of block copolymers formed by combining acetalated polysaccharides with various blocks, including both synthetic polymers and other polysaccharides. In Chapter 2, we describe the acetal modification of amylopectin to produce an acid-responsive hydrophobic material that enables particle formulation and encapsulation of glucoamylase enzyme by a double emulsion method. The pH-dependent acetal hydrolysis and subsequent particle degradation lead to the release of glucoamylase under acidic conditions, thereby demonstrating the pH-responsiveness of this system. Furthermore, the native amylopectin in the solution underwent enzymatic hydrolysis by glucoamylase to produce free glucose. This system demonstrated efficient glucose production under acidic conditions while preserving its stability and minimizing glucose production under physiological conditions. Similarly, Chapter 3 presents the formulation of a spermine-modified acetalated dextran-based nanocarrier for the encapsulation of NFkB decoy oligodeoxynucleotides (ODN) using a double emulsion method. Due to the cationic nature of the polymer, we observed high encapsulation efficiency along with pH-dependent particle degradation and controlled release of NFkB decoy ODN under acidic conditions. These formulations demonstrated significant inhibition of three key cancer characteristics: cell proliferation, migration, and colony formation. Furthermore, we developed polysaccharide-based amphiphilic block copolymers that respond to specific stimuli for targeted delivery of anticancer drugs. Specifically, in Chapter 4, we synthesized an acid-responsive block copolymer consisting of acetalated dextran and poly(2-ethyl-2-oxazoline), which self-assembled into micellar nanoparticles. A reactive oxygen species (ROS)-activatable dimeric prodrug of camptothecin (CPT) containing a cinnamaldehyde (CA)-based thioacetal linker was encapsulated into this particle system with excellent drug loading. This formulation has shown a pH-dependent particle disassembly and release of prodrugs. Moreover, the activation of prodrugs by elevated ROS level in cancer cells generates free CA that induces intracellular ROS production. This results in a self-amplification cycle that further activates the prodrug and enhances the therapeutic efficacy of CPT. Lastly, in Chapter 5, the concept of the amphiphilic block copolymer was extended to a fully polysaccharide-based dual-responsive block copolymer. Here, an acetalated dextran was conjugated with hyaluronic acid via a disulfide linkage. The self-assembly of this block copolymer exhibits pH and reduction-dependent particle degradation, enabling controlled drug release. Additionally, we observed an increase in toxicity of doxorubicin-loaded particles towards cancer cells with CD44 overexpression compared to non-cancer cells with low expression, highlighting the targeting capability of our micellar nanoparticles. Overall, this thesis demonstrates the development of different polysaccharide-based stimulus-responsive materials for smart drug delivery applications.

Children experiencing trauma and entering child protective services have been continuously increasing. Problems associated with childhood trauma, such as neurodevelopmental disorder, trauma and stress-related disorders, personality disorders, substance use disorder, externalizing and internalizing disorders, academic problems, relational difficulties, and delinquent behaviors, have been found increasing despite advances in trauma and translational research. Children’s trauma is mostly interpersonal in nature and nested in their immediate environment. There is a need for a change in focus from helping children to overcome challenges and adversities to strengthening the resilience-building process by utilizing functional strengths in the environment to achieve sustainable outcomes. This study’s goal was to investigate how ecological community-oriented variables can help strengthen resilience-building processes of adaptive abilities and skills based on cognitive, behavioral, and motivational principles and moderate the progression of risks in children, adolescents, and young adults ages 10 and 21. The results of this study revealed that the ecological models comprising several community-oriented variables were statistically significant in influencing the expected variance on the resilience-building adaptive abilities of children, adolescents, and young adults.

While majority of today’s oil wells employ directional drilling technology (deviated, extended reach and horizontal wells), a thorough understanding of the drill string dynamics is necessary to increase the drilling efficiency. Wellbore in such wells spans long horizontal distances through the shale to extract oil and natural gas effectively. Very long slender drill pipes transmit the required torque and cutting force through miles of distance from the earth’s surface to the drill bit. Drill string is subjected to different loads and torques which can cause coupled random excitations and failure of its components (drill pipes, bit, sensor tools and wellbore) eventually. If left unnoticed, these vibrations can cause stuck pipe and reduced rate of penetration, both of which are heavily cost dependent. Identifying the conditions causing harmful vibrations hence would significantly reduce cost and time. Controlling the drill string and bottom hole assembly is one way of mitigating the dynamic instability, which is currently done by means of controlling the rotational speed, torque applied and axial force applied to the drill string. This article presents modeling of horizontal drilling and the comparison of horizontal and vertical drill string dynamics. Drill string components are discretized into lumped elements based on their curvature. A vertical wellbore structure with same drill string components is considered for comparison. The computations are performed in MATLAB. Results and discussions are presented in the later part of the report.

The purpose of this thesis was to investigate regional corneal topographic responses in overnight orthokeratology (OK) and their influence on treatment zone (TZ) decentration. In all studies reported in this thesis corneal topography was assessed using the Medmont E300 corneal topographer. In the initial study, eyelid photographs of healthy volunteers were captured using the Nikon D5000 digital SLR camera. Eyelid parameters were determined from the photographs using i-Metrics software. The performance of BE spherical OK lenses on minimally and moderately toric corneas was evaluated in retrospective and prospective studies. When evaluating the performance of toric OK lenses on moderately toric corneas on a small number of participants the DreamLite TRZ toric periphery OK lens design was used. In general, corneal asymmetry was found to be significant between horizontal sectors but not in vertical sectors in minimally toric corneas. Several corneal sectors showed a difference between East Asian and non-East Asian eyes with clinically significant differences apparent only in the superior corneal sector. Upper eyelid curvature in non-East Asian eyes showed a stronger positive association with corneal spherical equivalent power than in East Asian eyes. During spherical OK lens wear on minimally toric corneas, a tendency of TZ decentration to move away from the most prolate corneal region was found. In the same cohort, it was shown that TZ decentration caused non-uniform corneal curvature changes. Significant central temporal flattening and para-central steepening were consistent with temporal TZ decentration. When the TZ decentration during spherical OK was compared between minimally and moderately toric corneas, decentration was significantly greater in moderately toric corneas. Alternative methods of fitting OK lenses with slightly deeper sag heights did not show any benefit in controlling TZ decentration in moderately toric corneas. Toric OK lens fitting on moderately toric corneas showed greater TZ decentration compared to spherical OK lenses, but the TZ margins were well defined giving good visual outcomes. In conclusion, normal corneal shape asymmetry in minimally toric corneas and magnitude of corneal toricity were shown to influence TZ decentration. Toric OK lenses may provide improved visual outcomes in moderately toric corneas.

A significant number of individuals with disabilities are denied powered mobility because they lack the visual, motor, or cognitive skills required to operate a powered wheelchair safely. The Drive-Safe System (DSS) is an add-on, distributed, shared control navigation assistance system for powered wheelchairs, intended to provide safe and independent mobility to these individuals. The DSS is a human-machine system in which the user and machine share navigation control. The user is responsible for high-level control of the system, such as choosing the destination, path planning, and some navigation actions, while the DSS overrides unsafe maneuvers through autonomous collision avoidance, automatic wall following, and door crossing. This dissertation reports the design and development of the DSS, followed by results from rigorous engineering and clinical evaluations. The engineering tested technical aspects of the DSS such as sensor coverage, maximum safe speed, maximum detection distance, and power consumption. Clinical evaluations included testing the DSS with Orientation & Mobility (O&M) specialists, ambulatory and non-ambulatory visually impaired individuals, and able-bodied controls. We compared the performance of the DSS with conventional navigation aids such as canes that are commonly used in conjunction with wheelchairs based on measures such as time for task completion and number of collisions. Additionally, we collected data with the NASA-TLX to gain insight into users' subjective experience with the DSS. Results indicate that the DSS was able to provide a uniform and reliable sensor coverage field around the wheelchair and could successfully detect obstacles as small as 3 inches in height to overhanging obstacles at a height of 55 inches. The DSS significantly reduced the number of collisions compared to using a cane. Users rated the DSS favorably despite the fact they took longer to navigate the same obstacle course than they would using a cane. Visually impaired participants reported experiencing less physical demand, and had to exert less effort in order to achieve better performance when using the DSS, compared to using a cane. These findings suggest that the DSS can be a viable solution for powered mobility in populations with visual impairment.

The recent 1994 Northridge and 1995 Hyogo-Ken earthquakes made the structural engineers to look for an alternate connection type to the fully welded connections, which failed to perform miserably under the nature’s seismic forces. This quest for an alternate connection type led the structural engineers to revisit the “Partially Restrained Connections (PRC)”, which exhibit sufficient ductility to dissipate the earthquake forces. Top and Seat Angle connection is one among the many PRC’s widely accepted and researched. The present study focuses on the various currently available prying models used to predict the capacity of angle connections. These include the prying model adopted by AISC design code (2005), the prying model used in Eurocode (2003), and Modified Struik Model by Swanson (2002). A comparison study is carried out to arrive at the prying model that predicts the capacity of the connection close to that obtained by performing a physical test on it and modification to the prying models is recommended to better the prediction of the capacity for a given angle connection. The experimental results on Top and Seat Angle connections reported by Abolmaali (1999) have been used in this study. In addition to the available experimental information on the Top and Seat Angle connections, static finite element analysis was performed on candidate selected angle connection configurations. This additional information on the connection behavior from the finite element analysis helped in substantiating the comparison study. Parameters defining the connection geometry were chosen and the test cases for performing the finite element analysis were formulated by looking at the different possible combinations of these geometric parameters, while taking into consideration the currently accepted design and fabrication requirements. A finite element software package ANSYS (6.1) was used to perform the finite element analysis on the connections.

Humans often organize information by encoding it in structures that link together entities such as concepts, objects, properties etc. Among the various structures possible, hierarchies are commonly used. For instance, taxonomies of categories commonly employ hierarchies to indicate that one category “is a” type of another. The Yahoo! Web Directory and the Open Directory Project are two examples of large taxonomies where topics are hierarchically arranged. Hierarchies are also used to recursively decompose composite objects into their constituent parts. Examples of this are webpages that can be parsed and then represented as DOM-trees, where the DOM nodes correspond to sections of the webpages. In this thesis we argue that these hierarchical relationships between entities can be exploited to facilitate common data mining tasks defined upon them, like automated classification. Specifically, we show that the information encoded in these hierarchies can be reduced to constraints on class membership scores that can then be enforced as a post-processing step to enhance the accuracy of classification. We demonstrate our ideas and algorithms on three real-world tasks. First, we tackle the problem of classification into hierarchical taxonomies. We show how different taxonomy structures can be translated into constraints on the outputs of classifiers learned at the nodes of the hierarchy. In addition, we give algorithms to optimally enforce these constraints and show that this results in improved classification accuracy. In cases where the taxonomies are not available, we give an approach to automatically derive hierarchical relationships amongst a flat set of categories. Next, we work on the problem of detecting noisy (templated) parts of webpages. We give algorithms that rate each section of a webpage in terms of how templated it is. Then we show that smoothing the output of these template classifiers over the DOM-tree hierarchy improves the template detection performance of our system. Finally, we investigate the task of segmenting websites into topically cohesive regions. We define a framework and within it a set of measures that characterize good segmentations, and give an efficient algorithm to find the best segmentation within this framework. We formalize the problem of enforcing constraints on the outputs of classifiers as regularized isotonic or unimodal regression on rooted trees; these are generalizations of the classic isotonic regression problem. The nature of the constraints as well as the cost functions is different in each of the applications mentioned above. For all these formulations we give efficient algorithms to optimally smooth the classifier outputs. These novel formulations and algorithms might be of interest independent of the applications in this thesis.

Independent mobility is critical to individuals of any age. While the needs of many individuals with disabilities can be satisfied with power wheelchairs, some members of the disabled community find it difficult or impossible to operate a standard power wheelchair. This population includes, but is not limited to, individuals with low vision, visual field neglect, spasticity, tremors, or cognitive deficits. To meet the needs of this population, our group is involved in developing cost effective modularly designed Smart Wheelchairs. Our objective is to develop an assistive navigation system which will seamlessly integrate into the lifestyle of individual with disabilities and provide safe and independent mobility and navigation without imposing an excessive physical or cognitive load. The Smart Wheelchair Component System (SWCS) can be added to a variety of commercial power wheelchairs with minimal modification to provide navigation assistance. Previous versions of the SWCS used acoustic and infrared rangefinders to identify and avoid obstacles, but these sensors do not lend themselves to many desirable higher-level behaviors. To achieve these higher level behaviors we integrated a Continuously Adapted Mean Shift (CAMSHIFT) target tracking algorithm into the SWCS, along with the Minimal Vector Field Histogram (MVFH) obstacle avoidance algorithm. The target tracking algorithm provides the basis for two distinct operating modes: (1) a "follow-the-leader" mode, and (2) a "move to stationary target" mode.The ability to track a stationary or moving target will make smart wheelchairs more useful as a mobility aid, and is also expected to be useful for wheeled mobility training and evaluation. In addition to wheelchair users, the caregivers, clinicians, and transporters who provide assistance to wheelchair users will also realize beneficial effects of providing safe and independent mobility to wheelchair users which will reduce the level of assistance needed by wheelchair users.

Implementation of homogeneous charge compression ignition (HCCI) combustion in gasoline engines, while operating with improved fuel economy and reduced NOx emissions, is limited by the inability to control combustion phasing. Spark-assisted compression ignition (SACI) is a mechanism that controls HCCI combustion phasing, using spark ignition to advance the onset of compression ignition (CI). SACI was employed at two light-load operating points (200 kPa and 300 kPa IMEP at fueling rates of 7 and 10 mg per cycle, respectively) to advance the phasing of a marginally-stable HCCI operating point. At the lower load, stratification by means of direct injection of 3 mg fuel at the time of spark ignition was required to advance phasing using SACI. At the higher load, robust SACI operation required less stratification (late injection of 1 mg). NOx emissions increased two- and four-fold in SACI relative to HCCI operation at the lower and higher load, respectively; this increase was shown to result primarily from the stratification. This study identified two issues confounding SACI, only observed at the lower load: cyclic variability in CI combustion phasing and inability to sufficiently advance phasing. To address these issues, new metrics were created in this study to identify four stages of the SACI combustion process: spark discharge, early kernel growth (EKG), flame propagation, and CI. Experiments were performed using an optical HCCI engine to simultaneously acquire steady-state emissions, in-cylinder pressure data, spark-discharge current and voltage waveforms, images of the plasma channel, and images of combustion chemiluminescence from two orthogonal views at sub-crank-angle resolution; these measurements were used to identify the four stages for each engine cycle and thereby the cyclic variability of each stage. It was concluded that cyclic variability in the EKG, the period of low flame growth-rate and weak chemiluminescence, was the dominant cause of cyclic variability in combustion phasing. Also, findings indicated only a weak correlation between the creation of large kernels and motion of the plasma channel away from the spark gap. At the higher load, robust flame propagation immediately followed the spark discharge. It was inferred that local fuel/air distribution affected the transition to flame propagation.

Cluster Ensembles is a framework for combining multiple partitionings obtained from separate clustering runs into a final consensus clustering without accessing the original features of the data or the algorithms that determined these partitions. This framework was first proposed by Strehl and Ghosh [31] who also provided three techniques to solve the problem. Since then there have been numerous attempts to solve cluster ensembles using approaches such as Maximum Likelihood using EM, Bipartite Graph Partitioning, Genetic algorithms, and Voting-Merging. Most of this work has focused on devising approaches that aceept hard clusterings as input. Also, there has been no comparison of combining accuracy on soft vs hard cluster ensembles. In this thesis we will show experimentally as well as intuitively that using soft clusterings as input does offer signficant advantages, especially when dealing with vertically partitioned data. We modify many of the above mentioned algorithms to accept soft clusterings and experiment over multiple real-life datasets

Analysis and optimization of cross-docking systems through simulation and analytical modeling

This dissertation is about the exact and approximate small sample sampling distributions of the estimators of Zellner's Seemingly Unrelated Regression Model. Chapter I is introductory. Chapter II is about exact and approximate distributions of Zellner's two stage estimator, with 'unrestricted' least square residuals used to estimate the covariance structure. The exact distribution is obtained by relying on a numerical multiple integration procedure by which all the points required to plot the probability density functions for a two equation model are calculated simultaneously. The exact densities are then used to evaluate the accuracy of the approximate first order asymptotic densities and the approximate second order asymptotic densities developed by T. J. Rothenberg for various 'experiments'. It turns out that the second order densities are more accurate than the first order densities, particularly for smaller sample sizes. Chapter III involves exact and approximate distributions of a particular test statistic of the SUR model. After showing that the second order densities are more accurate than the first order densities, the fact that the former can be written in a closed form expression is used to develop Edgeworth corrected critical points. These critical points are further away from the origin than the critical points obtained from the asymptotic distributions. For more than two equation SUR systems the second order approximations, unlike the exact distributions, can still be easily obtained. The fact that they can be used to compare the precision of the SUR estimators of the coefficients with the single equation estimators and also to get the Edgeworth corrected critical points for the above mentioned test statistic should be useful in empirical work. Chapter IV has the conclusion.

The chemical Process Industries rely heavily on pattern based information utilization. Process trends and patterns contain virtually all the information about the process and provide the operators a basis to determine the process condition. It is based on changes in these ratterns that operators make the necessary adjustments to the process operating conditions, depending on their interpretation. In general, process condition can confidently be judged only by monitoring multiple signals simultaneously and by context dependent pattern recognition. A simple example of a exothermic CSTR can be used to illustrate this. When the rector temperature rises and the coolant flow rate remains constant or decreases, the operator may take it as a premonition of impending doom, but if the coolant flow rate also goes up to accommodate for this increase in reactor temperature, the operator might take no prophylactic measures. It can easily be concluded that good signal processing techniques are necessary to develop the next generation of automated process monitoring techniques. Sensor signals samples are periodically collected by data collecting techniques that usually introduce sensor noise to the already noisy raw signal from the process. This raw signal has to be converted to a useful form before it can actually be used further. The challenge is to provide a signal processing technique that can process a signal on-line and feed it to a process monitoring technique. Howev~r, the catch is that all signal processing cause signal distortion at least to some extent. This cannot be tolerated, especially distortions towards the end ofthe signal, samples that correspond to the most recent juncture in time. This is because it is these samples that trigger off a process monitoring technique into predicting the process as normal or abnormal. Prediction of a normal state as abnormal can probably tolerated but predicting an abnormal state as normal can prove catastrophic. In this work a novel signal processing method using Wavelets is discussed. Wavelet Transforms are similar to the age old Fourier Transforms, but outmatches the Fourier Transforms in many desired attributes. Wavelets not only provide a powerful signal processing/analysis technique, they are also capable of providing significant data compression. In this work data compression's ofthe order of 90% were achieved without significant loss in the information content of the signals. This work to our knowledge is the first to emphasize the importance of signal extension techniques for an applications like pattern recognition. A new signal extension technique (NET) is described and its superiority over other signal extension techniques is demonstrated.