Este trabalho tem o intuito, de apresentar alguns conceitos, relativos à identifi cação de sistemás, tanto lineares quanto não linearep, além da ideia de referência virtual para, em conjunto com a teoria de projeto "de controladores baseados em dados, propor uma forrha de projeto de controladores não lineares baseados em identificação de sistemas. A utilização de referência virtual para a obtenção dos sinais necessários para a caracterização do controlador ótimo de um sistema e utilizado no método VRFT (Virtual Reference Feedback Tuning). Este método serve como base para o desenvolvimento da proposta deste trabalho que, em conjunto com a teoria de identificação de sistemas não lineares, permite a obteriçãci do controlador ótimo que leva o sistema a se comportar como especificado em malha fechada. Em especial optou-se pela caracterização do controlador utilizando estrutura de modelos racional, por esta ser uma classe bastante abrangente no que - diz respeito à quantidade de sistemas reais que ela é capaz de descrever. Fara demonstrar o potencial do método proposto para projeto de controladores, são apresentados ecemplos ilustrativos em situações onde o controlador ideal consegue ser representado pela classe de modelos, e quando isso não é possível. / This work aims to present some concepts related to linear and nonlinear system identification, as well as the •concept of virtual reference that, together with data based controller design's theory, provides design framework for nonlinear controllers. The Virtual Reference Feedback Tuning method (VRFT) is used as a basis for the current proposal, where we propose to unite nonlinear system identification algorithms and virtual reference to obtain the ideal controller: the one which makes the system behave as desired in closed loop. It was choosen to model the controller as a rational model due the wide variety of practical systems that can be represented by this model structure. For rational system identification we used an iterative algorithm which, based on the signal from input and output of the pIant, allows to identify the parameters of the pre defined controller structure with the signals obtained by virtual reference. To demonstrate the operation of the proposed identification controller methodology, illustrative examples are presented in situations where the ideal controller can be represented by the class of modeIs, and also when it is not possible.

This thesis, for the first time in the literature, presents an automatic mode-matching system that uses the phase relationships between the residual quadrature and drive signals in a gyroscope to achieve and maintain the frequency matching condition, and also the system allows controlling the system bandwidth by adjusting the closed loop parameters of the sense mode controller, independently from the mechanical sensor bandwidth. There are two mode-matching methods, using the proposed mode-matching system, presented in this thesis. In the first method, the frequency matching between the resonance modes of the gyroscope is automatically accomplished by changing the proof mass potential. The main motivation behind the first method is to tune the sense mode resonance frequency with respect to the drive mode resonance frequency using the electrostatic tuning capability of the sense mode. In the second method, the mode-matched gyroscope operation is accomplished by using dedicated frequency tuning electrodes that only provides a capability of tuning the sense mode resonance frequency generating an electrostatic spring effect on the sense frame, independently from the proof mass potential. This study mainly focuses on the second method because the proof mass potential variation is not desired during the gyroscope operation since the proof mass potential directly affects the drive and sense mode dynamics of the gyroscope. Therefore, a single-mass fully-decoupled gyroscope including the dedicated frequency tuning electrodes are designed. To identify mode shapes and mode frequencies of the designed gyroscope, FEM simulations are performed. The designed gyroscopes are fabricated using SOI-based SOG process. The fabrication imperfections are clarified during the formation of the structural layer of the gyroscope. Next, the closed loop controllers are designed for the drive amplitude control, sense force-feedback, quadrature cancellation, and mode-matching regarding the phase relationship between the quadrature and drive signals. Mode-matching is achieved by using a closed loop controller that provides a DC tuning potential. The mode-matching system consisting of vacuum packaged sensor, drive amplitude control, sense force-feedback, quadrature cancellation, and mode-matching modules is implemented on a printed circuit board (PCB), and then the system level tests are performed. Tests illustrate that the mode-matching system operates in a desired manner. Test results demonstrate that the performances of the studied MEMS gyroscopes are improved up to 2.6 times in bias instability and 2 times in ARW under the mode-matched condition compared to the mismatched (~200 Hz) condition, reaching down to 0.73 &deg / /hr and 0.024 &deg / /&radic / hr, respectively. At the mode-matched gyroscope operation, the better performance is obtained to be bias instability of 0.87

This study presented the design and implementation of a spatial Hinf controller to suppress the free and forced vibrations of a cantilevered smart beam. The smart beam consists of a passive aluminum beam with surface bonded PZT (Lead-Zirconate-Titanate) patches. In this study, the PZT patches were used as the actuators and a laser displacement sensor was used as the sensor. In the first part of the study, the modeling of the smart beam by the assumed-modes method was conducted. The model correction technique was applied to include the effect of out-of-range modes on the dynamics of the system. Later, spatial system identification work was performed in order to clarify the spatial characteristics of the smart beam. In the second part of the study, a spatial Hinf controller was designed for suppressing the first two flexural vibrations of the smart beam. The efficiency of the controller was verified both by simulations and experimental implementation. As a final step, the comparison of the spatial and pointwise Hinf controllers was employed. A pointwise Hinf controller was designed and experimentally implemented. The efficiency of the both controllers was compared by simulations.

This study presents the theoretical analysis of an active flutter suppression methodology applied to a smart fin. The smart fin consists of a cantilever aluminum plate-like structure with surface bonded piezoelectric (PZT, Lead- Zirconate-Titanate) patches. A thermal analogy method for the purpose of modeling of piezoelectric actuators in MSC&reg / /NASTRAN based on the analogy between thermal strains and piezoelectric strains was presented. The results obtained by the thermal analogy were compared with the reference results and very good agreement was observed. The unsteady aerodynamic loads acting on the structure were calculated by using a linear two-dimensional Doublet-Lattice Method available in MSC&reg / /NASTRAN. These aerodynamic loads were approximated as rational functions of the Laplace variable by using one of the aerodynamic approximation schemes, Roger&amp / #8217 / s approximation, with least-squares method. These approximated aerodynamic loads together with the structural matrices obtained by the finite element method were used to develop the aeroelastic equations of motion of the smart fin in state-space form. The Hinf robust controllers were then designed for the state-space aeroelastic model of the smart fin by considering both SISO (Single-Input Single-Output) and MIMO (Multi-Input Multi-Output) system models. The verification studies of the controllers showed satisfactory flutter suppression performance around the flutter point and a significant improvement in the flutter speed of the smart fin was also observed.

The omnipresent tendency to &ldquo / live easy&rdquo / is a sign of our need for automatization. To enable for such a &ldquo / comfortable and safe&rdquo / world, the automatic systems have to be developed that satisfies the necessities of life. Biologically inspired robots, especially the humanoids, are thus the key, and research in this area focuses on the improvement of such systems. Lately, it has been shown by high dexterity examples that the humanoid robots achieved to a mature level even if there are still open issues to be improved, especially in the control and stability of the bipeds. The purpose of this thesis is to study biped locomotion in different floor conditions, such as stairs and obstacles / to improve the research done in this area / to contribute to the development of autonomous biped robots, dynamic modeling, gait planning, supervisory and guidence control, stability analysis of biped robots / and to implement new control algorithms for biped locomotion, especially by using optimization and high level intelligent control techniques. The locomotion aimed to be realized results from complex, high-dimensional, nonlinear and dynamically related interactions between the robot and its environment. The mathematical modeling of the physical system is realized based on a 5-link 7 DOF biped robot model walking on a 3D planar surface and the dynamic simulation is performed using MATLAB. In terms of control, several different methods applied, comparison and the performance of each method are given. The 3D dynamic simulation software is developed, which allows the user to operate the biped systems within a 3D virtual environment.

This thesis, for the first time in literature, introduces a comprehensive study about analog controller designs for MEMS vibratory gyroscopes. A controller of a MEMS gyroscope is mandatory for robust operation, which is insensitive to sensor parameters and ambient con- ditions. Errors in the controller design not only deteriorate transient performance, such as settling time and overshoot, but also cause performance degradation due to stability problems. Accordingly, true controller design for a gyroscope is critical work in terms of functionality and system performance. This thesis gives details for modeling, analysis of closed-loop sys- tems, and design procedure for drive and sense modes. Controller loops are implemented both with discrete components and in a CMOS technology as an integrated circuit. Simulation and test results verify the modeling, analysis, and design procedure discussed in this thesis. Drive mode system developed previously at METU is optimized by taking circuit imperfec- tions into account, which results in an improved transient performance of 50 msec settling time with no overshoot for a 4&mu / m drive mode oscillation amplitude. This system has a 60 phase margin with the help of the pole-zero cancellation technique. In addition, a new gener- iv ation and simple drive mode controller for tactical grade applications is designed and verified with a moderate transient performance. Two different sense mode controller design procedures are also developed according to a new base-band equivalent model derived for mismatch operation, as a new contribution to the literature. Firstly, a PID controller is designed for low frequency separation between the drive and sense modes of the gyroscope. Secondly, an integral controller is used for moderate and high mismatch amount. The controller system designed with the new base-band equivalent model improves the linearity, angle random walk, and bias instability by factors of 4, 9, and 3, respectively. Proposed drive and sense mode controllers are also designed and implemented using a 0.6&mu / m standard CMOS process. These chips are the first functional chips developed at METU de- signed for MEMS gyroscopes. Functionality of the proposed three systems, i.e., conventional drive mode controller, new generation drive mode controller, and sense mode controller, are verified with tests. The first prototypes result in 0.033 degree/sqrt/(hr) angle random walk and 3 degree/hr bias instability for open-loop operation, which is very promising and can be improved even further in future designs.

Este trabalho tem o intuito, de apresentar alguns conceitos, relativos à identifi cação de sistemás, tanto lineares quanto não linearep, além da ideia de referência virtual para, em conjunto com a teoria de projeto "de controladores baseados em dados, propor uma forrha de projeto de controladores não lineares baseados em identificação de sistemas. A utilização de referência virtual para a obtenção dos sinais necessários para a caracterização do controlador ótimo de um sistema e utilizado no método VRFT (Virtual Reference Feedback Tuning). Este método serve como base para o desenvolvimento da proposta deste trabalho que, em conjunto com a teoria de identificação de sistemas não lineares, permite a obteriçãci do controlador ótimo que leva o sistema a se comportar como especificado em malha fechada. Em especial optou-se pela caracterização do controlador utilizando estrutura de modelos racional, por esta ser uma classe bastante abrangente no que - diz respeito à quantidade de sistemas reais que ela é capaz de descrever. Fara demonstrar o potencial do método proposto para projeto de controladores, são apresentados ecemplos ilustrativos em situações onde o controlador ideal consegue ser representado pela classe de modelos, e quando isso não é possível. / This work aims to present some concepts related to linear and nonlinear system identification, as well as the •concept of virtual reference that, together with data based controller design's theory, provides design framework for nonlinear controllers. The Virtual Reference Feedback Tuning method (VRFT) is used as a basis for the current proposal, where we propose to unite nonlinear system identification algorithms and virtual reference to obtain the ideal controller: the one which makes the system behave as desired in closed loop. It was choosen to model the controller as a rational model due the wide variety of practical systems that can be represented by this model structure. For rational system identification we used an iterative algorithm which, based on the signal from input and output of the pIant, allows to identify the parameters of the pre defined controller structure with the signals obtained by virtual reference. To demonstrate the operation of the proposed identification controller methodology, illustrative examples are presented in situations where the ideal controller can be represented by the class of modeIs, and also when it is not possible.

Este trabalho tem o intuito, de apresentar alguns conceitos, relativos à identifi cação de sistemás, tanto lineares quanto não linearep, além da ideia de referência virtual para, em conjunto com a teoria de projeto "de controladores baseados em dados, propor uma forrha de projeto de controladores não lineares baseados em identificação de sistemas. A utilização de referência virtual para a obtenção dos sinais necessários para a caracterização do controlador ótimo de um sistema e utilizado no método VRFT (Virtual Reference Feedback Tuning). Este método serve como base para o desenvolvimento da proposta deste trabalho que, em conjunto com a teoria de identificação de sistemas não lineares, permite a obteriçãci do controlador ótimo que leva o sistema a se comportar como especificado em malha fechada. Em especial optou-se pela caracterização do controlador utilizando estrutura de modelos racional, por esta ser uma classe bastante abrangente no que - diz respeito à quantidade de sistemas reais que ela é capaz de descrever. Fara demonstrar o potencial do método proposto para projeto de controladores, são apresentados ecemplos ilustrativos em situações onde o controlador ideal consegue ser representado pela classe de modelos, e quando isso não é possível. / This work aims to present some concepts related to linear and nonlinear system identification, as well as the •concept of virtual reference that, together with data based controller design's theory, provides design framework for nonlinear controllers. The Virtual Reference Feedback Tuning method (VRFT) is used as a basis for the current proposal, where we propose to unite nonlinear system identification algorithms and virtual reference to obtain the ideal controller: the one which makes the system behave as desired in closed loop. It was choosen to model the controller as a rational model due the wide variety of practical systems that can be represented by this model structure. For rational system identification we used an iterative algorithm which, based on the signal from input and output of the pIant, allows to identify the parameters of the pre defined controller structure with the signals obtained by virtual reference. To demonstrate the operation of the proposed identification controller methodology, illustrative examples are presented in situations where the ideal controller can be represented by the class of modeIs, and also when it is not possible.

In this Thesis a Unified Discrete State-Space Model for power converters in CCM is presented. Two main approaches to arriving at the discrete model are used. The first approach involves an impulse function approximation of the duty cycle modulations of the converter switches , and this approach results in a small signal discrete model. The Second approach is direct and does not involve any approximation of the modulations , this approach yields both a large signal nonlinear discrete model and a linear small signal model. Harmonic analysis of the converter states at steady-state is done for steady-state waveform acquisition , which increases the accuracy of the model especially for finding the control to inductor current frequency response. Finally the Discrete model is verified for the Half-Bridge DC/DC topology for its three main control schemes (Asymmetric , Symmetric , DCS). A GUI platform in MATLAB is presented as a wrapper that utilizes the models and analysis presented in this thesis.

No / This paper presents a novel concept to describe the couplings among the outputs of the stochastic systems which are represented by NARMA models. Compared with the traditional coupling description, the presented concept can be considered as an extension using statistical independence theory. Based on this concept, the decoupling control in statistical sense is established with the necessary and sufficient conditions for complete decoupling. Since the complete decoupling is difficult to achieve, a control algorithm has been developed using the Cauchy-Schwarz mutual information criterion. Without modifying the existing control loop, this algorithm supplies a compensative controller to minimise the statistical couplings of the system outputs and the local stability has been analysed. In addition, a further discussion illustrates the combination of the presented control algorithm and data-based mutual information estimation. Finally, a numerical example is given to show the feasibility and efficiency of the proposed algorithm.