In the design of a robust control system, one needs a nominal model together with a quantitative bound on the uncertainty that results from under-modeling and disturbances. In this thesis we do not intentionally seek a nominal model and a quantitative bound, instead, the uncertainty is directly parameterized so that the resulting uncertain model family can be characterized by means of a real parameter vector with at most unit length. This is an innovative approach to the control-oriented system identification, since it is not in accordance with the general philosophy of robust identification. However, it is applicable to the robust synthesis problem by taking advantage of a convex parameterization of robust controllers that simultaneously stabilize the uncertain models in the family. The robust performance problem becomes tractable since it can be converted into a quasi-convex optimization problem with Linear Matrix Inequality (LMI) constraints. The relation between the optimal robust performance and the uncertainty is studied by analyzing the explicit bounds of the maximal robust margin. Model (in)validation is a complement to system identification. In our approach it is an integral ingredient of the process of obtaining robust control-oriented system models. A single model is not invalidated if it is inside the ellipsoid, and thus the intersection of the ellipsoids is not invalidated. In order to make the unfalsified model set (the intersection) fit in our framework, we can compute an optimal ellipsoid bounding the intersection of the ellipsoids. (Abstract shortened by UMI.)

This dissertation attempts to address two important problems in systems engineering, namely, multicriteria optimization and robustness optimization. In fields ranging from engineering to the social sciences designers are very often required to make decisions that attempt to optimize several criteria or objectives at once. Mathematically this amounts to finding the Pareto optimal set of points for these constrained multiple criteria optimization problems which happen to be nonlinear in many realistic situations, particularly in engineering design. Traditional techniques for nonlinear multicriteria optimization suffer from various drawbacks. The popular method of minimizing weighted sums of the multiple objectives suffers from the deficiency that choosing an even spread of 'weights' does not yield an even spread of points on the Pareto surface and further this spread is often quite sensitive to the relative scales of the functions. A continuation/homotopy based strategy for tracing out the Pareto curve tries to make up for this deficiency, but unfortunately requires exact second derivative information and further cannot be applied to problems with more than two objectives in general. Another technique, goal programming, requires prior knowledge of feasible goals which may not be easily available for more than two objectives. Normal-Boundary Intersection (NBI), a new technique introduced in this dissertation, overcomes all of the difficulties inherent in the existing techniques by introducing a better parametrization of the Pareto set. It is rigorously proved that NBI is completely independent of the relative scales of the functions and is quite successful in producing an evenly distributed set of points on the Pareto set given an evenly distributed set of 'NBI parameters' (comparable to the 'weights' in minimizing weighted sums of objectives). Further, this method can easily handle more than two objectives while retaining the computational efficiency of continuation-type algorithms, which is an improvement over homotopy techniques for tracing the trade-off curve. Various aspects of NBI including computational issues and its relationships with minimizing convex combinations and goal programming are discussed in this dissertation. Finally some case studies from engineering disciplines are performed using NBI. The other facet of this dissertation deals with robustness optimization, a concept useful in quantifying the stability of an optimum in the face of random fluctuations in the design variables. This robustness optimization problem is presented as an application of multicriteria optimization since it essentially involves the simultaneous minimization of two criteria, the objective function value at a point and the dispersion in the function values in a neighborhood of the point. Moreover, a formulation of the robustness optimization problem is presented so that it fits the framework of constrained, nonlinear optimization problems, which is an improvement on existing formulations that deal with either unconstrained nonlinear formulations or constrained linear formulations.

The purpose of this study is to develop a general algorithm to solve in a robust, flexible and fast way the short-term scheduling problems of two different multi-reservoir hydroelectric power systems. The first system consists of four reservoirs in series, the second corresponds to that of the Ste Maurice river. / The solution method is based on the discrete maximum principle. Gradient method is used for the solution of the two-point boundary value problem. Two algorithms are suggested for dealing with difficulties posed by the state variable constraints. The first uses the augmented lagrangian method, the second is iterative. / Both algorithms give a satisfactory solution for the problem. However, the first requires more memory space than the second but converges more rapidly. The second algorithm has the advantage of executing an iteration in less CPU time. For large scale problems the second (iterative) algorithm option is recommended.

In this thesis, we address issues in the specification, simulation, and formal verification of systems that are characterzed by real-time constraints and a mix of protocol and data computation aspects. We propose a novel specification language and modeling methodology--HAAD (Hierarchical Annotated Action Diagrams). In HAAD, the interface behavior or a system is captured as a hierarchy of action diagrams. The internal behavior is modeled by an Extended Finite State Machine (EFSM). A leaf action diagram defines a behavior (a template) over a set of ports. Procedures and predicates are attached to actions in order to describe the functional aspects of the interface. / We propose algorithms and methods for the automatic generation of simulation models and response verification scripts from HAAD specifications. These models perform "on-the-fly parsing" of actions received at their I/O ports, sequencing through state transitions based on the result of this parsing, detecting incorrect, or ill-formed interface operations (bus cycles), verifying that all timing constraints at the input of the model are met, and driving the model outputs with appropriate delays. / We formalize the operational semantics of leaf action diagrams under linear timing constraints, based on the concepts of a block machine and causal block machine. We state the realizability of an action diagram in terms of the existence of a causal block machine derived from the action diagram. We examine the problem of the compatability of concurrent, communicating leaf action diagrams described by linear timing constraints and we show the inaccuracies of known methods that address this problem. We define the action diagram compatibility problem in terms of the compatibility of all the possible combinations of causal block machines derived form these action diagrams. We prove that such enumeration is not needed in answering the compatibility question. This leads to an exact and efficient compatibility verification procedure.

In the field of robotic manipulation, a key problem addressed by researchers is the presence of uncertainties in the position and orientation of objects to be manipulated by a computer controlled robotic manipulator. One approach used to reduce these uncertainties is the use of pushing operations, without the use of sensors. In this thesis we first used the quasi-static assumption made by several researchers to plan trajectories of pushed sliding objects. Quasi-static pushing refers to operations in which the mass or acceleration of the object is significantly small that inertial forces are dominated by frictional forces. The travel distance of the object, which is first pushed with one point of contact before coming into contact with a second point of the pusher, is calculated. Then, we compare these trajectories with ones predicted based on a dynamical model of the system. This new model takes inertial forces into consideration and it comprises a more realistic friction model as it also includes a velocity dependent coefficient of friction factor. Results showed that predictions in terms of object rotation direction in similar in both cases, but that predictably the object rotation rate, in the dynamical system, is slower due inertial forces, and therefore a longer travel distance is needed to reorient an object.

The goal of business modelling is the design, analysis, and simulation of an enterprise's architectural structures and their information technology components. To comprehensively support business modelling, an appropriate modelling environment with adequate visualization mechanisms is required. Such an environment may handle model information in a flexible, yet expressive way and support substitution, documentation, validation and dynamic analysis of models as well as model visualization and alternative representations. / We have developed a business modelling approach which is based on the formalism of extended, colored Petri nets. To support and validate our approach, we have engineered the Macrotec environment. Macrotec meets a list of requirements we have identified as crucial for the quality of a comprehensive modelling environment. It is conceived as a set of various tools which are seamlessly integrated. Our experience with Macrotec suggests that our concepts and environment substantially facilitate business modelling.

The object of software frameworks is to mechanize human processes in order to accomplish high-level tasks that call upon diverse software tools. This thesis describes the ELISA framework prototype which performs power-system dynamic security analysis in the operations planning environment. ELISA mechanizes routines traditionally carried out by experts that are essential to power-system dynamic security analysis, greatly accelerating the realization of complex processes. Typically, ELISA executes appropriate load-flow and transient-stability simulations (i.e. using commercially available simulation software), performs result analysis, identifies and executes changes to the input and repeats this process until a user-defined goal, such as finding transient stability transfer limits, has been achieved. / A taxonomy of dynamic security analysis in operations planning is proposed employing the semantic net, class-object-property and rule paradigms. All of these are required to cover the full spectrum of knowledge found in the high-level goals, the process details, the complex conditional structures and the acceptance criteria which characterize dynamic security analysis. This taxonomy also describes the language of operations planners, defining not only the features presently supported by ELISA, but also providing a roadmap to future enhancements. Typical sensitivity studies are presented using a 700-bus production model of the Hydro-Quebec network to illustrate the considerable leverage afforded from using ELISA-like software. / In addition, the thesis addresses the issue of how such tools can assist in performing research to improve our understanding of fundamental power systems behaviour. Using the ELISA prototype as a laboratory test bed, it is shown that the signal energy E of a network's transient response acts as a barometer to define the relative severity of any normal contingency with respect to power generation or transfer P. For a given contingency, as P is varied and the network approaches instability, signal energy increases smoothly and predictably towards an asymptote which defines the network's stability limit: This limit, in turn, permits us to compare the severity of different contingencies. This behaviour can be explained in terms of the effect of increasing power on the damping component of dominant poles, and a simple function is derived which estimates network stability limits with surprising accuracy from two or three stable simulations. / As a corollary to this, it is also shown that a network's transient response can be screened for instability using a simple frequency-domain criterion. Essentially, this criterion requires performing the Fourier transform of a network's transient voltage response at various monitoring locations: When P is varied and the network goes beyond its stability limit, the angle of the Fourier transform's polar plot fundamentally changes its behaviour, passing from a clockwise to a counterclockwise rotational behaviour about the origin. This is confirmed by results obtained from performing stability-limit searches on the Hydro-Quebec system. Used in conjunction with signal energy analysis for determining stability limit proximity, this criterion can be quite useful for mechanized security-limit-determination tools such as ELISA. / Signal energy limit estimation and the proposed stability criterion are shown to be applicable to all normal contingencies and these results hold not-withstanding the presence of many active, nonlinear elements in the network.

This work addresses the problem of robust stabilization and robust Hinfinity control of uncertain time-delay systems. The time-delays are considered to be present in the states and/or the outputs, and the uncertainties in the system representation are of the parametric norm-bounded type. Both cases of actuators, with and without saturation are studied, and the state-feedback and output-feedback control designs are presented. Two methods for analysis and synthesis of controllers are used: The first is based on the transfer function, and the second on the use of functionals. / In the context of the design method based on transfer functions, the problem of Hinfinity output feedback design for a class of uncertain linear continuous-time or discrete-time systems, with delayed states and/or outputs (only for the continuous-time case), and norm-bounded parametric uncertainties is considered. The objective is to design a linear output feedback controller such that, for the unknown state and output time-delays and all admissible norm-bounded parameter uncertainties, the feedback system remains robustly stable and the transfer function from the exogenous disturbances to the state-error outputs meets the prescribed Hinfinity norm upper-bound constraint. The output feedback structure does not depend on the time-delay. The conditions for the existence of the desired robust Hinfinity output feedback and the analytical expression of these controllers, are then characterized in terms of matrix Riccati-type inequalities. In the continuous-time context, both the time-invariant and the time-varying cases are treated. Finally, examples are presented to demonstrate the validity and the solvability of the proposed design methods. / Still in the same context, the state-feedback robust stabilization problem for neutral systems with time-varying delays and saturating actuators is addressed. The systems considered are continuous-time, with parametric uncertainties entering all the matrices in the system representation. The model used for the representation of actuator saturations is that of differential inclusions. A saturating control law is designed and a region of initial conditions is specified within which local asymptotic stability of the closed-loop system is ensured. / Finally, the robust output-feedback stabilization problem for state-delayed systems with time-varying delays and saturating actuators is addressed. The systems considered are again continuous-time, with parametric uncertainties entering all the matrices in the system representation. Two models are used for the representation of actuator saturations: sector modeling and differential inclusions. Saturating control laws are designed, and in the case of differential inclusions, a region of initial conditions is specified within which local asymptotic stability of the closed-loop system is ensured. (Abstract shortened by UMI.)

FPGAs have been among the fastest growing segments of the semiconductor industry and this growth is expected to continue over the next several years. The complexity of FPGAs makes manual designs time consuming and error prone; design automation tools are essential to the effective use of FPGAs. Routing is one part of the design process of VLSI chips that is computational complex and hence very time consuming. / The arrival of the interior-point method proposed by N. Karmarkar was an important event in linear programming and several variants have been developed since then. Recent experiments in a variety of areas report that Karmarkar's algorithm performs asymptotically faster than the traditional simplex method and is competitive with simulated annealing. / In this thesis, a global router for symmetrical FPGAs is implemented based on the dual-affine variant of Karmarkar's interior-point method. Current approaches in VLSI global routing, the model of symmetrical FPGAs and the formulation of the global-routing problem as a mixed-integer linear program are reviewed. The objective of the global router is to find a routing that balances the channel-segment densities in the symmetrical FPGA. Specific implementation details are covered in depth and experimental results for samples of random global-routing problems are reported.

Design automation has steadily contributed to improvements witnessed in the system design process. Initial applications were to address low level design concerns such as transistor layout and simulation; however the focus of tools has slowly been progressing up the design abstraction scale. The current state-of-the-art provides modelling capabilities at different levels of abstraction, but solutions for synthesis issues at the register-transfer and lower levels are the norm. The proliferation of design description languages at different abstraction levels has prompted the need for standardization (VHDL and Open-Verilog) to promote design migration and re-use. / While design automation has helped in reducing design time-lines and design churn, a major source of design difficulties is just recently being addressed and promise to be the next wave in design automation applicability. The problems arise within the architectural (or system) level of abstraction very early in the design cycle. The recent research in this field attempts to bridge the design process gap between specification and design, and provides a platform for experimenting with hardware and software trade-offs. / This dissertation studies the requirements for an environment for architectural design. In particular, an environment specific to the telecommunications domain is proposed in order to limit the potentially large design exploration space. An intermediate design language is also introduced to accommodate both high level modelling and synthesis driven by the user and environment. Finally a Design Analysis and Synthesis Environment (DASE) is described to facilitate the architectural level activities. The environment, a proof of concept, provides generic modal library, simulation, synthesis and Petri-net analysis support. Realistic design examples are explored, to illustrate architectural design activities with the environment.