Sound power measurements of acoustic sources are generally made in reverberation or anechoic chambers using acoustic pressure measurements as outlined in specific ISO or other standards. A reverberation chamber produces an approximate diffuse-field condition, wherein the sound power is determined from the spatially averaged squared pressure. An anechoic chamber produces an approximate free-field condition, wherein the sound power is estimated from squared pressure over an enveloping measurement surface. However, in many cases it is desirable to estimate sound power within nonideal semi-reverberant spaces. In these environments, both direct and reverberant energies may contribute significantly to the total acoustic field. This paper introduces two measurement methods that utilize a weighted combination of potential and kinetic energy densities, known as generalized acoustic energy density, to estimate sound power in nonideal semi-reverberant rooms. The first method employs a generalized sound power formulation, which is an adaptation to an equation developed in 1948 for semi-reverberant spaces. The second, called the two-point in situ method, is a technique based on the generalized sound power formulation for quick and accurate in situ sound power estimates. Since the generalized acoustic energy density is more spatially uniform than the squared acoustic pressure in an enclosed field, these methods have the advantage of achieving the same accuracy in sound power determination with fewer measurement positions. This thesis explores the possibility of using these new methods in place of methods outlined in current ISO standards by describing analytical, numerical, and experimental results.

This master thesis uses applied mathematicalstatistics to analyse purchase behaviour based on customer data of the Swedishbrand Indiska. The aim of the study is to build a model that can helppredicting the sales quantities of different product classes and identify whichfactors are the most significant in the different models and furthermore, tocreate an algorithm that can provide suggested product combinations in thepurchasing process. Generalized linear models with a Negative binomial distributionare applied to retrieve the predicted sales quantity. Moreover, conditionalprobability is used in the algorithm which results in a product recommendationengine based on the calculated conditional probability that the suggestedcombinations are purchased.From the findings, it can be concluded that all variables considered in themodels; original price, purchase month, colour, cluster, purchase country andchannel are significant for the predicted outcome of the sales quantity foreach product class. Furthermore, by using conditional probability andhistorical sales data, an algorithm can be constructed which createsrecommendations of product combinations of either one or two products that canbe bought together with an initial product that a customer shows interest in. / Matematisk statistik tillämpas i denna masteruppsats för att analysera köpbeteende baserat på kunddata från det svenska varumärket Indiska. Syftet med studien är att bygga modeller som kan hjälpa till att förutsäga försäljningskvantiteter för olika produktklasser och identifiera vilka faktorer som är mest signifikanta i de olika modellerna och därtill att skapa en algoritm som ger förslag på rekommenderade produktkombinationer i köpprocessen. Generaliserade linjära modeller med en negativ binomialfördelning utvecklades för att beräkna den förutspådda försäljningskvantiteten för de olika produktklasserna. Dessutom används betingad sannolikhet i algoritmen som resulterar i en produktrekommendationsmotor som baseras på den betingade sannolikheten att de föreslagna produktkombinationerna är inköpta.Från resultaten kan slutsatsen dras att alla variabler som beaktas i modellerna; originalpris, inköpsmånad, produktfärg, kluster, inköpsland och kanal är signifikanta för det predikterade resultatet av försäljningskvantiteten för varje produktklass. Vidare är det möjligt att, med hjälp av betingad sannolikhet och historisk försäljningsdata, konstruera en algoritm som skapar rekommendationer av produktkombinationer av en eller två produkter som kan köpas tillsammans med en produkt som en kund visar intresse för.

For many decades, engineers and students have heavily depended on simulation packages such as Pspice to run transit and steady-state simulation for their circuits. The majority of these circuits, such as soft switching cells, contain complicated modes of operations that require the Pspice simulation to run for a long time and, finally, it may not reach a convergent solution for these kinds of circuits. Also, there is a need for an educational tool that provides students with a better understanding of circuit modes of operation through state-plan figures and steady-state switching waveforms. The unified steady-state computer aided model proposes a simulation block that covers common unified soft-switching cells operations and can be used in topologies simulation. The simulation block has a simple interface that enables the user to choose the switching cell type and connects the developed simulation model in the desired topology configuration. In addition to the measured information that can be obtained from the circuitry around the unified simulation model, the simulation block includes some additional nodes (other than the inputs and outputs) that make internal switching cell information, such as switching voltages and currents, easy to access and debug. The model is based on mathematical equations, resulting in faster simulation times, smaller file size and greatly minimized simulation convergence problems. The Unified Model is based on the generalized analysis: Chapter 1 discusses the generalized equation concept along with a detailed generalization example of one switching cell, which is the zero current switching quasi-resonant converter ZCS-QRC. Chapter 2 presents a detailed discussion of the unified model concept, the unified model flow chart and the unified model implementation in Pspice. Chapter 3 presents the unified model applications; generating the switching cell inductor current and the switching cell capacitor voltage steady-state waveforms, the State-Plane Diagram , the feedback design using the unified model, and the chapter concludes with how the model can be used with different topologies. Finally, chapter 4 presents the summary and the future work

Statistics show that signalized intersections are among the most dangerous locations of a roadway network. Different approaches including crash frequency and severity models have been used to establish the relationship between crash occurrence and intersection characteristics. In order to model crash occurrence at signalized intersections more efficiently and eventually to better identify the significant factors contributing to crashes, this dissertation investigated the temporal, spatial, and site correlations for total, rear-end, right-angle and left-turn crashes. Using the basic regression model for correlated crash data leads to invalid statistical inference, due to incorrect test statistics and standard errors based on the misspecified variance. In this dissertation, the Generalized Estimating Equations (GEEs) were applied, which provide an extension of generalized linear models to the analysis of longitudinal or clustered data. A series of frequency models are presented by using the GEE with a Negative Binomial as the link function. The GEE models for the crash frequency per year (using four correlation structures) were fitted for longitudinal data; the GEE models for the crash frequency per intersection (using three correlation structures) were fitted for the signalized intersections along corridors; the GEE models were applied for the rear-end crash data with temporal or spatial correlation separately. For right-angle crash frequency, models at intersection, roadway, and approach levels were fitted and the roadway and approach level models were estimated by using the GEE to account for the "site correlation"; and for left-turn crashes, the approach level crash frequencies were modeled by using the GEE with a Negative Binomial link function for most patterns and using a binomial logit link function for the pattern having a higher proportion of zeros and ones in crash frequencies. All intersection geometry design features, traffic control and operational features, traffic flows, and crashes were obtained for selected intersections. Massive data collection work has been done. The autoregression structure is found to be the most appropriate correlation structure for both intersection temporal and spatial analyses, which indicates that the correlation between the multiple observations for a certain intersection will decrease as the time-gap increase and for spatially correlated signalized intersections along corridors the correlation between intersections decreases as spacing increases. The unstructured correlation structure was applied for roadway and approach level right-angle crashes and also for different patterns of left-turn crashes at the approach level. Usually two approaches at the same roadway have a higher correlation. At signalized intersections, differences exist in traffic volumes, site geometry, and signal operations, as well as safety performance on various approaches of intersections. Therefore, modeling the total number of left-turn crashes at intersections may obscure the real relationship between the crash causes and their effects. The dissertation modeled crashes at different levels. Particularly, intersection, roadway, and approach level models were compared for right-angle crashes, and different crash assignment criteria of "at-fault driver" or "near-side" were applied for disaggregated models. It shows that for the roadway and approach level models, the "near-side" models outperformed the "at-fault driver" models. Variables in traffic characteristics, geometric design features, traffic control and operational features, corridor level factor, and location type have been identified to be significant in crash occurrence. In specific, the safety relationship between crash occurrence and traffic volume has been investigated extensively at different studies. It has been found that the logarithm of traffic volumes per lane for the entire intersection is the best functional form for the total crashes in both the temporal and spatial analyses. The studies of right-angle and left-turn crashes confirm the assumption that the frequency of collisions is related to the traffic flows to which the colliding vehicles belong and not to the sum of the entering flows; the logarithm of the product of conflicting flows is usually the most significant functional form in the model. This study found that the left-turn protection on the minor roadway will increase rear-end crash occurrence, while the left-turn protection on the major roadway will reduce rear-end crashes. In addition, left-turn protection reduces Pattern 5 left-turn crashes (left-turning traffic collides with on-coming through traffic) specifically, but it increases Pattern 8 left-turn crashes (left-turning traffic collides with near-side crossing through traffic), and it has no significant effect on other patterns of left-turn crashes. This dissertation also investigated some other factors which have not been considered before. The safety effectiveness of many variables identified in this dissertation is consistent with previous studies. Some variables have unexpected signs and a justification is provided. Injury severity also has been studied for Patterns 5 left-turn crashes. Crashes were located to the approach with left-turning vehicles. The "site correlation" among the crashes occurred at the same approach was considered since these crashes may have similar propensity in crash severity. Many methodologies and applications have been attempted in this dissertation. Therefore, the study has both theoretical and implementational contribution in safety analysis at signalized intersections.

The development in power conversion technology is in constant demand of high power efficiency and high power density. The DC-DC power conversion is an indispensable stage for numerous power supplies and energy related applications. Particularly, in PV micro-inverters and front-end converter of power supplies, great challenges are imposed on the power performances of the DC-DC converter stage, which not only require high efficiency and density but also the capability to regulate a wide variation range of input voltage and load conditions. The resonant DC-DC converters are good candidates to meet these challenges with the advantages of achieving soft switching and low EMI. Among various resonant converter topologies, the LLC converter is very attractive for its wide gain range and providing ZVS for switches from full load to zero load condition. The operation of the LLC converter is complicated due to its multiple resonant stage mechanism. A literature review of different analysis methods are presented, and it shows that the study on the LLC is still incomplete. Therefore, an operation mode analysis method is proposed, which divides the operation into six major modes based on the occurrence of resonant stages. The resonant currents, voltages and the DC gain characteristics for each mode is investigated. To obtain a thorough view of the converter behavior, the boundaries of every mode are studied, and mode distribution regarding the gain, load and frequency is presented and discussed. As this operation mode model is a precise model, an experimental prototype is designed and built to demonstrate its accuracy in operation waveforms and gain prediction. iv Since most of the LLC modes have no closed-form solutions, simplification is necessary in order to utilize this mode model in practical design. Some prior approximation methods for converter’s gain characteristics are discussed. Instead of getting an entire gain-vs.-frequency curve, we focus on peak gains, which is an important design parameters indicating the LLC’s operating limit of input voltage and switching frequency. A numerical peak gain approximation method is developed, which provide a direct way to calculate the peak gain and its corresponding load and frequency condition. The approximated results are compared with experiments and simulations, and are proved to be accurate. In addition, as PO mode is the most favorable operation mode of the LLC, its operation region is investigated and an approximation approach is developed to determine its boundary. The design optimization of the LLC has always been a difficult problem as there are many parameters affecting the design and it lacks clear design guidance in selecting the optimal resonant tank parameters. Based on the operation mode model, three optimization methods are proposed according to the design scenarios. These methods focus on minimize the conduction loss of resonant tank while maintaining the required voltage gain level, and the approximations of peak gains and PO mode boundary can be applied here to facilitate the design. A design example is presented using one of the proposed optimization methods. As a comparison, the L-C component values are reselected and tested for the same design specifications. The experiments show that the optimal design has better efficiency performance. Finally, a generalized approach for resonant converter analysis is developed. It can be implemented by computer programs or numerical analysis tools to derive the operation waveforms and DC characteristics of resonant converters

In many applications, the quality of process outputs is described by more than one characteristic variable. These quality variables usually follow a multivariate normal (MN) distribution. This dissertation discusses the monitoring of the mean vector and the covariance matrix of MN processes. The first part of this dissertation develops a statistical process control (SPC) chart based on a generalized likelihood ratio (GLR) statistic to monitor the mean vector. The performance of the GLR chart is compared to the performance of the Hotelling Χ² chart, the multivariate exponentially weighted moving average (MEWMA) chart, and a multi-MEWMA combination. Results show that the Hotelling Χ² chart and the MEWMA chart are only effective for a small range of shift sizes in the mean vector, while the GLR chart and some carefully designed multi-MEWMA combinations can give similarly better overall performance in detecting a wide range of shift magnitudes. Unlike most of these other options, the GLR chart does not require specification of tuning parameter values by the user. The GLR chart also has the advantage in process diagnostics: at the time of a signal, estimates of change-point and out-of-control mean vector are immediately available to the user. All these advantages of the GLR chart make it a favorable option for practitioners. For the design of the GLR chart, a series of easy to use equations are provided to users for calculating the control limit to achieve the desired in-control performance. The use of this GLR chart with a variable sampling interval (VSI) scheme has also been evaluated and discussed. The rest of the dissertation considers the problem of monitoring the covariance matrix. Three GLR charts with different covariance matrix estimators have been discussed. Results show that the GLR chart with a multivariate exponentially weighted moving covariance (MEWMC) matrix estimator is slightly better than the existing method for detecting any general changes in the covariance matrix, and the GLR chart with a constrained maximum likelihood estimator (CMLE) gives much better overall performance for detecting a wide range of shift sizes than the best available options for detecting only variance increases. / Ph. D.

The objective of this dissertation is to investigate GLR control charts based on a sequential sampling scheme (SS GLR charts). Phase II monitoring is considered and the goal is to quickly detect a wide range of changes in the univariate normal process mean parameter and/or the variance parameter. The performance of the SS GLR charts is evaluated and design guidelines for SS GLR charts are provided so that practitioners can easily apply the SS GLR charts in applications. More specifically, the structure of this dissertation is as follows: We first develop a two-sided SS GLR chart for monitoring the mean μ of a normal process. The performance of the SS GLR chart is evaluated and compared with other control charts. The SS GLR chart has much better performance than that of the fixed sampling rate GLR chart. It is also shown that the overall performance of the SS GLR chart is better than that of the variable sampling interval (VSI) GLR chart and the variable sampling rate (VSR) CUSUM chart. The SS GLR chart has the additional advantage that it requires fewer parameters to be specified than other VSR charts. The optimal parameter choices are given, and regression equations are provided to find the limits for the SS GLR chart. If detecting one-sided shifts in μ is of interest, the above SS GLR chart can be modified to be a one-sided chart. The performance of this modified SS GLR chart is investigated. Next we develop an SS GLR chart for simultaneously monitoring the mean μ and the variance 𝜎² of a normal process. The performance and properties of this chart are evaluated. The design methodology and some illustrative examples are provided so that the SS GLR chart can be easily used in applications. The optimal parameter choices are given, and the performance of the SS GLR chart remains very good as long as the parameter choices are not too far away from the optimized choices. / Ph. D.

The Design of an experiment involves selection of levels of one or more factor in order to optimize one or more criteria such as prediction variance or parameter variance criteria. Good experimental designs will have several desirable properties. Typically, one can not achieve all the ideal properties in a single design. Therefore, there are frequently several good designs and choosing among them involves tradeoffs. This dissertation contains three different components centered around the area of optimal design: developing a new graphical evaluation technique, discussing designs for non-regular regions for first order models with interaction for the two- and three-factor case, and using the standard designs in the case of generalized linear models (GLM). The Fraction of Design Space (FDS) technique is proposed as a new graphical evaluation technique that addresses good prediction. The new technique is comprised of two tools that give the researcher more detailed information by quantifying the fraction of design space where the scaled predicted variance is less than or equal to any pre-specified value. The FDS technique complements Variance Dispersion Graphs (VDGs) to give the researcher more insight about the design prediction capability. Several standard designs are studied with both methods: VDG and FDS. Many Standard designs are constructed for a factor space that is either a p-dimensional hypercube or hypersphere and any point inside or on the boundary of the shape is a candidate design point. However, some economic, or practical constraints may occur that restrict factor settings and result in an irregular experimental region. For the two- and three-factor case with one corner of the cuboidal design space excluded, three sensible alternative designs are proposed and compared. Properties of these designs and relative tradeoffs are discussed. Optimum experimental designs for GLM depend on the values of the unknown parameters. Several solutions to the dependency of the parameters of the optimality function were suggested in the literature. However, they are often unrealistic in practice. The behavior of the factorial designs, the well-known standard designs of the linear case, is studied for the GLM case. Conditions under which these designs have high G-efficiency are formulated. / Ph. D.

Nowadays, outlier detection has attracted growing interest. Unlike typical outlier detection problems, relational outlier detection focuses on detecting abnormal patterns in datasets that contain relational implications within each data point. Furthermore, different from the traditional outlier detection that focuses on only numerical data, modern outlier detection models must be able to handle data in various types and structures. Detecting relational outliers should consider (1) Dependencies among different data types, (2) Data types that are not continuous or do not have ordinal characteristics, such as binary, categorical or multi-label, and (3) Special structures in the data. This thesis focuses on the development of relational outlier detection methods and real-world applications in datasets that contain non-numerical, mixed-type, and special structure data in three tasks, namely (1) outlier detection in mixed-type data, (2) categorical outlier detection in music genre data, and (3) outlier detection in categorized time series data. For the first task, existing solutions for mixed-type data mostly focus on computational efficiency, and their strategies are mostly heuristic driven, lacking a statistical foundation. The proposed contributions of our work include: (1) Constructing a novel unsupervised framework based on a robust generalized linear model (GLM), (2) Developing a model that is capable of capturing large variances of outliers and dependencies among mixed-type observations, and designing an approach for approximating the analytically intractable Bayesian inference, and (3) Conducting extensive experiments to validate effectiveness and efficiency. For the second task, we extended and applied the modeling strategy to a real-world problem. The existing solutions to the specific task are mostly supervised, and the traditional outlier detection methods only focus on detecting outliers by the data distributions, ignoring the input-output relation between the genres and the extracted features. The proposed contributions of our work for this task include: (1) Proposing an unsupervised outlier detection framework for music genre data, (2) Extending the GLM based model in the first task to handle categorical responses and developing an approach to approximate the analytically intractable Bayesian inference, and (3) Conducting experiments to demonstrate that the proposed method outperforms the benchmark methods. For the third task, we focused on improving the outlier detection performance in the second task by proposing a novel framework and expanded the research scope to general categorized time-series data. Existing studies have suggested a large number of methods for automatic time series classification. However, there is a lack of research focusing on detecting outliers from manually categorized time series. The proposed contributions of our work for this task include: (1) Proposing a novel semi-supervised robust outlier detection framework for categorized time-series datasets, (2) Further extending the new framework to an active learning system that takes user insights into account, and (3) Conducting a comprehensive set of experiments to demonstrate the performance of the proposed method in real-world applications. / Doctor of Philosophy / In recent years, outlier detection has been one of the most important topics in the data mining and machine learning research domain. Unlike typical outlier detection problems, relational outlier detection focuses on detecting abnormal patterns in datasets that contain relational implications within each data point. Detecting relational outliers should consider (1) Dependencies among different data types, (2) Data types that are not continuous or do not have ordinal characteristics, such as binary, categorical or multi-label, and (3) Special structures in the data. This thesis focuses on the development of relational outlier detection methods and real-world applications in datasets that contain non-numerical, mixed-type, and special structure data in three tasks, namely (1) outlier detection in mixed-type data, (2) categorical outlier detection in music genre data, and (3) outlier detection in categorized time series data. The first task aims on constructing a novel unsupervised framework, developing a model that is capable of capturing the normal pattern and the effects, and designing an approach for model fitting. In the second task, we further extended and applied the modeling strategy to a real-world problem in the music technology domain. For the third task, we expanded the research scope from the previous task to general categorized time-series data, and focused on improving the outlier detection performance by proposing a novel semi-supervised framework.

Expressiveness, and more recently, succinctness, are two central concerns of finite model theory and descriptive complexity theory. Succinctness is particularly interesting because it is closely related to the complexity-theoretic trade-off between parallel time and the amount of hardware. We develop new bounds on the expressiveness and succinctness of first-order logic with two variables on finite words, present a related result about the complexity of the satisfiability problem for this logic, and explore a new approach to the generalized star-height problem from the perspective of logical expressiveness. We give a complete characterization of the expressive power of first-order logic with two variables on finite words. Our main tool for this investigation is the classical Ehrenfeucht-Fra¨ıss´e game. Using our new characterization, we prove that the quantifier alternation hierarchy for this logic is strict, settling the main remaining open question about the expressiveness of this logic. A second important question about first-order logic with two variables on finite words is about the complexity of the satisfiability problem for this logic. Previously it was only known that this problem is NP-hard and in NEXP. We prove a polynomialsize small-model property for this logic, leading to an NP algorithm and thus proving that the satisfiability problem for this logic is NP-complete. Finally, we investigate one of the most baffling open problems in formal language theory: the generalized star-height problem. As of today, we do not even know whether there exists a regular language that has generalized star-height larger than 1. This problem can be phrased as an expressiveness question for first-order logic with a restricted transitive closure operator, and thus allows us to use established tools from finite model theory to attack the generalized star-height problem. Besides our contribution to formalize this problem in a purely logical form, we have developed several example languages as candidates for languages of generalized star-height at least 2. While some of them still stand as promising candidates, for others we present new results that prove that they only have generalized star-height 1.