Robust regression is a commonly required approach in industrial studies like data mining, quality control and improvement, and finance areas. Among the robust regression methods / Least Median Squares, Least Trimmed Squares, Mregression, MM-method, Least Absolute Deviations, Locally Weighted Scatter Plot Smoothing and Multivariate Adaptive Regression Splines are compared under contaminated normal distributions with each other and Ordinary Least Squares with respect to the multiple outlier detection performance measures. In this comparison / a simulation study is performed by changing some of the parameters such as outlier density, outlier locations in the x-axis, sample size and number of independent variables. In the comparison of the methods, multiple outlier detection is carried out with respect to the performance measures detection capability, false alarm rate and improved mean square error and ratio of improved mean square error. As a result of this simulation study, the three most competitive methods are compared on an industrial data set with respect to the coefficient of multiple determination and mean square error.

In this study, we address the long-term project pricing problem for a company that operates in the defense industry. The pricing problem is a bid project pricing problem which includes various technical and financial uncertainties, such as estimations of workhour content of the project and exchange &amp / inflation rates. We propose a Robust Optimization (RO) approach that can deal with the uncertainties during the project lifecycle through the identification of several discrete scenarios. The bid project&rsquo / s performance measures, other than the monetary measures, for R&amp / D projects are identified and the problem is formulated as a multi-attribute utility project pricing problem. In our RO approach, the bid pricing problem is decomposed into two parts which are v solved sequentially: the Penalty-Model, and the RO model. In the Penalty-Model, penalty costs for the possible violations in the company&rsquo / s workforce level due to the bid project&rsquo / s workhour requirements are determined. Then the RO model searches for the optimum bid price by considering the penalty cost from the Penalty-Model, the bid project&rsquo / s performance measures, the probability of winning the bid for a given bid price and the deviations in the bid project&rsquo / s cost. Especially for the R&amp / D type projects, the model tends to place lower bid prices in the expected value solutions in order to win the bid. Thus, due to the possible deviations in the project cost, R&amp / D projects have a high probability of suffering from a financial loss in the expected value solutions. However, the robust solutions provide results which are more aware of the deviations in the bid project&rsquo / s cost and thus eliminate the financial risks by making a tradeoff between the bid project&rsquo / s benefits, probability of winning the bid and the financial loss risk. Results for the probability of winning in the robust solutions are observed to be lower than the expected value solutions, whereas expected value solutions have higher probabilities of suffering from a financial loss.

The purpose of this dissertation is to study the robust control for the smart flexible linkage mechanism. The control of flexible linkage induced inertia force under high-speed rotation is taken into consideration with the system parameter uncertainties such as modeling error, truncation error, and both of control spillover and observation spillover due to the residual modes of structural control problem. Based on the principles of LQ, optimal model following (OMF) and frequency shaping, this study proposes some sufficient conditions of stability criteria for the design of robust controller, respectively. These techniques guarantee that the controlled plant, under both bounded parameter perturbations and model truncation, could remain stable. Meanwhile, searching for the optimal locating positions of sensor and actuator by applying Taguchi method and genetic algorithm (GA) combined technique is further studied. The system is modeled through employing finite element method (FEM), and the limited lower frequency part modes are separated into controlled modes and residual modes. In time domain, at first we design a Luenberger-observer-based robust controller for the finite-dimensional mode plant keeping stability in a specified region. And then, a robust controller with the OMF is designed for the controlled system to achieve the performance as those of the specified optimum model. From the view of frequency domain, the robust controller could also be deigned according to the frequency shaping principle to suppress the exciting effect of higher frequency residual modes, and similarly avoid that the system might be destabilized. Finally, the combination of Taguchi method and GA technique to search the optimal locations for placing actuator and sensor to coincide with the stability and performance requirement is also done. From the computer simulation, the middle point movement of the linkage is obviously well controlled; hence, the designed robust controllers can certainly suppress the affection of parameter uncertainties and the spillover stimulation of residual modes, and can produce satisfactory results.

Abstract In this dissertation, the eigenvalue analysis and decentralized robust controller design of uncertain multi-time-scale system with parametrical perturbations are considered. Because the eigenvalues of the multi-time-scale systems cluster in some difference regions of the complex plane, we can use the singular perturbation method to separate the systems into some subsystems. These subsystems are independent to each other. We can discuss the properties of eigenvalues and design controller for these subsystem respectively, then we composite these controllers to a decentralized controller. The eigenvalue positions dominate the stability and the performance of the dynamic system. However, we cannot obtain the precise position of the eigenvalues from the influence of parametrical perturbations. The sufficient conditions of the eigenvalues clustering for the multi-time-scale systems will be discussed. The uncertainties consider as unstructured and structured perturbations are taken into considerations. The design algorithm provides for designing a decentralized controller that can assign the poles to our respect regions. The specified regions are half-plane and circular disk. Furthermore, the concepts of decentralized control and optimal control are used to design the linear quadratic regulator (LQR) controller and linear quadratic Gaussian (LQG) controller for the perturbed multi-time-scale systems. That is, the system can get the optimal robust performance. The bound of the singular perturbation parameter would influence the robust stability of the multi-time-scale systems. Finally, the sufficient condition to obtain the upper bound of the singular perturbation parameter presented by the Lyapunov method and matrix norm. The condition also extends for the pole assignment in the specified regions of each subsystem respectively. The illustrative examples are presented behind each topic. They show the applicability of the proposed theorems, and the results are satisfactory.

In this thesis, the multi-carrier (MC) code division multiple access (CDMA) system is considered in Rayleigh fading channel. The main concern of this thesis is to devise a new direct linearly constrained constant modulus (LCCM) inverse QRD-RLS algorithm for multiple access interference (MAI) cancellation and the problem due to the mismatch of the channel estimator. In the conventional approach, two significant detectors are applied to the system for multiuser interference suppression, one is the blind adaptation algorithm and the other is adaptive linearly constrained PLIC approach. However, the mirror effect may occur when the blind adaptation algorithm is employed. It might affect the performance in terms of bit error rate (BER), although the desired signal to interference (due to other users) improvement is still acceptable. Moreover, in case that the channel coefficients could not be estimated perfectly, the mismatch problem may occur to degrade the performance of the adaptive linearly constrained PLIC approach with the LMS or RLS algorithm. To overcome the mismatch problem, the conventional approach is to use the LCCM criterion with gradient algorithm. However, the convergence rate of the gradient algorithm is too slow to be implemented in real-time wireless communication system. In this thesis, to have fast convergence rate and to circumvent the mismatch problem, the robust LCCM-IQRD algorithm is devised and applied to the MC-CDMA system in Rayleigh fading channel. The proposed robust LCCM-IQRD algorithm has shown to be more effective in terms of MAI cancellation and the mismatch due to imperfect channel estimator. The performance, in terms of BER, of the proposed algorithm is superior to that of the conventional PLIC based algorithms, the blind adaptation algorithm, and the conventional LCCM gradient algorithm.

In this paper, the gap metric is introduced to study the robustness of the stability of feedback systems. A relation between the gap metric and coprime fractions is also investigated. It is shown that the stability radius of the controller in the gap metric is equal to the stability margin of the controller. In the loop-shaping design procedure in the £h-gap metric, it is practically hard to formulate an ideal controller. Finally, this paper studied the conservatism of the gap metric, and proposed some properties that can help for control design and analysis.

In this thesis, the problem of multiple access interference (MAI) suppression for the multi-carrier (MC) code division multiple access (CDMA) system, based on the wavelet-based (WB) multi-carrier modulation, associated with the combining process is investigated for Rayleigh fading channel. The main concern of this thesis is to derive a new scheme, based on the linearly constrained constant modulus (LCCM) criterion with the robust inverse QR decomposition (IQRD) recursive least squares (RLS) algorithm to improve the performance of the conventional MC-CDMA system with combining process. To verify the merits of the new algorithm, the effect due to imperfect channel parameters estimation and frequency offset are investigated. We show that the proposed robust LCCM IQRD-RLS algorithm outperforms the conventional LCCM-gradient algorithm [6], in terms of output SINR, improvement percentage index (IPI), and bit error rate (BER) for MAI suppression under channel mismatch environment. Also, the performance of the WB MC-CDMA system is superior to the one with conventional MC-CDMA system. It is more robust to the channel mismatch and frequency offset. Moreover, the WB MC-CDMA system with robust LCCM IQRD-RLS algorithm does have better performance over other conventional approaches, such as the LCCM-gradient algorithm, maximum ratio combining (MRC), blind adaptation algorithm and partitioned linear interference canceller (PLIC) approach with LMS algorithm, in terms of the capability of MAI suppression and bit error rate (BER).

This thesis presents strict LMI conditions for the bounded real lemma of discrete descriptor systems. Compared with existing nonstrict LMI conditions, the proposed new conditions are more tractable and reliable in numerical computations, in the sense that they can be tested easily by using the LMI Control Toolbox of Matlab. Based on the strict LMI conditions, the state feedback design for H-infinite control problem is also addressed. A sufficient LMI condition is derived so that the constructed feedback gain matrix from its solution will meet the design criteria of the closed-loop systems. Furthermore, we can probe into the problems of robust H-infinite control and pole-clustering in a disk for uncertain discrete descriptor systems subject to time-invariant norm-bounded uncertainty and convex polytopic uncertainty in the state matrix, respectively. Some sufficient LMI conditions are derived for analysis and design of these problems as well. Numerical examples are included to illustrate the results.

In recent years, there has been a growing interest in wireless ad hoc network. One of the major issues in wireless network is developing efficient routing protocol. Based on the concept of designing protocol model such as OSI model, the designers distilled the process of transmitting data to its most fundamental elements and identified which networking functions had related uses and collected those functions into discrete groups that became the layers. It is not suitable to design wireless ad hoc routing protocol based on OSI model conception because the OSI model is developed from the view point of wired network and there are many different characteristics between wired and wireless environment. The main different characteristics between wired and wireless are the mobility of mobile host and the transmission medium. Such differences have great effect on network performance. Due to the differences between wired and wireless characteristics, we present a comprehensive conception of designing wireless ad hoc routing protocol. In this context, we provide a wireless ad hoc routing protocol based on physical layer characteristics, ex: bit error rate, robust link. Our routing protocol will find out a route in good transmission environment and it is efficient to improve network throughput. Furthermore, our routing protocol will decrease the number of route request packets, the amount of retransmissions, link breakage rate, and increase throughput.

The Rapid Thermal Chemical Vapor Deposition (RTCVD) system is an emerging and promising technology in semiconductor manufacturing which possess advantages of rapidly increasing wafer temperature and reducing the thermal budget over traditional batch processing. In recent years, the growth of thin films in the manufacture of semiconductor devices has been widely employed in the industry. Because the influences of processing variables on RTCVD systems may lead to spatial wafer temperature non-uniformity, the precise control of wafer temperature is an important issue up to the present. In this paper the complementary sensitivity function shaping based on H-infinite control theory is applied to design robust controllers for the single-input/single-output (SISO) model of the RTCVD system, the multi-input/multi-output (MIMO) model of the RTCVD system, and the MIMO model with multiplicative uncertainties. Through control the power of the tungsten-halogen lamps, it can achieve the temperature tracking with good uniformity. Finally, the computer simulation results are obviously that the performance of the proposed controllers is satisfactory.