In this thesis, we study brand choice problem via the following three perspectives: a company's market share management, introduction of customers with different perspectives, and an analysis of an application domain which is illustrative of these issues. Our contributions following these perspectives include: (1) development of a stochastic differential-jump game (SDJG) model for brand competition in a specific situation wherein market share is modeled by a jump-diffusion process, (2) a robust hierarchical logit/probit model for market heterogeneity, and (3) applications of logit/probit model to the dynamic pricing problem occurring in production-inventory systems with jump events. Our research explores the use of quantitative method of operations research to control the dynamics of market share and provides a precise estimation method to integrate more detail information in discrete brand choice models.

Efficient integrated/robust planning and recovery models were studied. The research focus was to integrate fleet assignment and crew scheduling, and, in addition, to provide solutions robust to real time operations. The contributions include: (1) To understand how schedule development and fleet assignment stages influence crew scheduling performance, schedule analysis methods were proposed to evaluate the crew friendliness of a schedule for a given fleet. (2) To meet the computational challenges of crew scheduling in integrated planning, a duty flow model was proposed which can efficiently find suboptimal legal pairing solutions. (3) A new robust crew scheduling method based on spoke purity was proposed. Computational results indicated that with little or no extra cost, a more robust crew pairing solutions can be expected. (4) By imposing station purity, an integrated and robust planning model which integrates fleet assignment and crew connections was proposed. The impact of crew base purities and fleet purities on FAM profit, crew scheduling, and computational efficiency were investigated. (5) Airline integrated recovery method was studied. A recovery scope for integrated recovery was proposed to limit the ripple effect caused by disruptions. Based on the defined recovery scope, a new integrated recovery model and Bender¡¯s decomposition solution approach was studied.

The human auditory perception system is much more noise-robust than any state-of the art automatic speech recognition (ASR) system. It is expected that the noise-robustness of speech feature can be improved by employing the human auditory based feature extraction procedure. In this thesis, we investigate modifying the commonly-used feature extraction process for automatic speech recognition systems. A novel frequency masking curve, which is based on modeling the basilar membrane as a cascade system of damped simple harmonic oscillators, is used to replace the critical-band masking curve to compute the masking threshold. We mathematically analyze the coupled motion of the oscillator system (basilar membrane) when they are driven by short-time stationary (speech) signals. Based on the analysis, we derive the relation between the amplitudes of neighboring oscillators, and accordingly insert a masking module in the front-end signal processing stage to modify the speech spectrum. We evaluate the proposed method on the Aurora 2.0 noisy-digit speech database. When combined with the commonly-used cepstral mean subtraction post-processing, the proposed auditory front-end module achieves a significant improvement. The method of correlational masking effect curve combine with CMS can achieves relative improvements of 25.9% over the baseline respectively. After applying the methods iteratively, the relative improvement improves from 25.9% to 30.3%.

Sliding mode controllers (SMC) with time delay and a rotor position observer are designed for the sensorless speed control of permanent magnet synchronous motor (PMSM) are proposed in this paper. Based on field-oriented principle, a flux SMC is designed to achieve quick flux control. And then a speed SMC with time delay is presented and compared with PI controller in the direct torque control framework. The effectiveness of the proposed control scheme under the load disturbance and parameter uncertainties is verified by simulation results.

In this paper, we investigate rescaling of energy features for noise-robust speech recognition. The performance of the speech recognition system will degrade very quickly by the influence of environmental noise. As a result, speech robustness technique has become an important research issue for a long time. However, many studies have pointed out that the impact of speech recognition under the noisy environment is enormous. Therefore, we proposed the data-driven energy features rescaling (DEFR) to adjust the features. The method is divided into three parts, that are voice activity detection (VAD), piecewise log rescaling function and parameter searching algorithm. The purpose is to reduce the difference of noisy and clean speech features. We apply this method on Mel-frequency cepstral coefficients (MFCC) and Teager energy cepstral coefficients (TECC), and we compare the proposed method with mean subtraction (MS) and mean and variance normalization (MVN). We use the Aurora 2.0 and Aurora 3.0 databases to evaluate the performance. From the experimental results, we proved that the proposed method can effectively improve the recognition accuracy.

Considered in this research is a framework for effective formation control of multirobot systems in dynamic environments. The basic formation control involves two important considerations: (1) Real-time trajectory generation algorithms for distributed control based on nominal agent models, and (2) robust tracking of reference trajectories under model uncertainties. Proposed is a two-layer hierarchical architecture for collectivemotion control ofmultirobot nonholonomic systems. It endows robotic systems with the ability to simultaneously deal with multiple tasks and achieve typical complex formation missions, such as collisionfree maneuvers in dynamic environments, tracking certain desired trajectories, forming suitable patterns or geometrical shapes, and/or varying the pattern when necessary. The study also addresses real-time formation tracking of reference trajectories under the presence of model uncertainties and proposes robust control laws such that over each time interval any tracking errors due to system uncertainties are driven down to zero prior to the commencement of the subsequent computation segment. By considering a class of nonlinear systems with favorable finite-time convergence characteristics, sufficient conditions for exponential finite-time stability are established and then applied to distributed formation tracking controls. This manifests in the settling time of the controlled system being finite and no longer than the predefined reference trajectory segment computing time interval, thus making tracking errors go to zero by the end of the time horizon over which a segment of the reference trajectory is generated. This way the next segment of the reference trajectory is properly initialized to go into the trajectory computation algorithm. Consequently this could lead to a guarantee of desired multi-robot motion evolution in spite of system uncertainties. To facilitate practical implementation, communication among multi-agent systems is considered to enable the construction of distributed formation control. Instead of requiring global communication among all robots, a distributed communication algorithm is employed to eliminate redundant data propagation, thus reducing energy consumption and improving network efficiency while maintaining connectivity to ensure the convergence of formation control.

Stability is a critical issue in gain-scheduled control problems in that the closed loop system may not be stable during the transitions between operating conditions despite guarantees that the gain-scheduled controller stabilizes the plant model at fixed values of the scheduling variable. For Linear Parameter Varying (LPV) model representations, a controller interpolation method using Youla parameterization that guarantees stability despite fast transitions in scheduling variables is proposed. By interconnecting an LPV plant model with a Local Controller Network (LCN), the proposed Youla parameterization based controller interpolation method allows the interpolation of controllers of different size and structure, and guarantees stability at fixed points over the entire operating region. Moreover, quadratic stability despite fast scheduling is also guaranteed by construction of a common Lyapunov function, while the characteristics of individual controllers designed a priori at fixed operating condition are recovered at the design points. The efficacy of the proposed approach is verified with both an illustrative simulation case study on variation of a classical MIMO control problem and an experimental implementation on a multi-evaporator vapor compression cycle system. The dynamics of vapor compression systems are highly nonlinear, thus the gain-scheduled control is the potential to achieve the desired stability and performance of the system. The proposed controller interpolation/switching method guarantees the nonlinear stability of the closed loop system during the arbitrarily fast transition and achieves the desired performance to subsequently improve thermal efficiency of the vapor compression system.

The purpose of this thesis is to study the active robust control for the flexible slider-crank linkage mechanism with piezoelectric films using genetic algorithms. The instability caused by the inertia force that is induced by the high-speed rotation flexible slider-crank linkage, mode truncation, parameter uncertainties, and spillover effect due to the residual modes of structure. For the application of the mechanical structure system, a mathematical model for a slider-crank linkage mechanism with piezoelectric films is developed in conjunction with finite element method (FEM), and the lower frequency modes are separated into controlled modes and residual modes. For the robustness of the system, a robust stability condition and genetic algorithms are employed to ensure the stability of the system under the parameter uncertainties and spillover effect simultaneously. Numerical simulation is performed to the control methodology with genetic algorithms can suppress deform of flexible slider-crank linkage mechanism operating at high speeds, and the instability caused by the parameter uncertainties and spillover effect can be avoided.

In this thesis, pairwise multiple comparisons and multiple comparisons with a control are studied when the observations have short-tailed symmetric distributions. Under non-normality, the testing procedure is given and Huber estimators, trimmed mean with winsorized standard deviation, modified maximum likelihood estimators and ordinary sample mean and sample variance used in this procedure are reviewed. Finally, robustness properties of the stated estimators are compared with each other and it is shown that the test based on the modified maximum likelihood estimators has better robustness properties under short-tailed symmetric distribution.

This study describes the design and implementation of the pitch and roll autopilots for a fixed wing unmanned vehicle. A Tactical Unmanned Aerial Vehicle (TUAV), which is designed at the Middle East Technical University (METU), is used as a platform. This work combines development of the classical and robust controllers, which are used for the pitch and roll autopilots. One of the important steps in the thesis is development of the non-linear dynamic model of the UAV, which is developed in MATLAB/Simulink environment. Two different strategies of the controller design imply development of the PID and controllers. Simulation results illustrate the performances of the designed controllers. Simulation is performed for the nominal model of the UAV and for the model that includes uncertainties and sensor noises.