<p> One of the two basic philosophies underlying adaptive control is that the transfer function of the plant must be first determined and then the values of an adjustable controller varied for optimizing a given index of performance. The process of identifying the plant characteristics is popularly known as Identification Problem and constitutes a major problem in the realization of an adaptive system of this type.</p> <p> The other philosophy is that a complete knowledge of the plant is not necessary for the optimum adjustments of the parameter of control. The system is caused to measure its own performance against a figure of merit and drives its performance towards optimum. This approach is becoming popular because of the many difficulties associated with the identification problem and a number of "hill climbing" techniques have been proposed based on this philosophy.</p> <p> In this thesis, three such techniques (steepest descent, conjugate gradients and parallel tangents) have been analysed with a view to determine the most efficient and quickest way to determine the parameters of a controller for optimum performance.</p> / Thesis / Master of Engineering (MEngr)

Structural Health Monitoring (SHM) is the process of monitoring the state of a structure to determine the existence, location, and degree of damage that may exist within the entire structure. A structure’s health or level of damage can be monitored by identifying changes in structural or modal parameters. In this research, the structure’s health is monitored by identifying changes in structural stiffness. The Adaptive Least Mean Square (LMS) filtering approach is used to directly identify changes in structural stiffness for the IASC-ASCE Structural Health Monitoring Task Group Benchmark problem for both Phase I and II. The research focuses primarily on Phase II of the benchmark problem. In Phase II, modeling error and noise is introduced to the problem making the problem more realistic. The research found that the LMS filter approach can be used to detect damage and distinguish relative severity of the damage in Phase II of the benchmark problem in real time. Even though the LMS filter approach identified damage, a threshold below which damage is hard to identify exists. If the overall stiffness changes less than 10%, then identifying the presence and location of damage is difficult. But if the time of damage is known, then the presence and location can be determined. The research is of great interest to those in the structural health monitoring community, structural engineers, and inspection practitioners who deal with structural damage identification problems.

A complete chain of analysis and synthesis system identification tools for detectability assessment and adaptive identification of parameters with physical interpretation that can be found commonly in control-oriented powertrain models is presented. This research is motivated from the fact that future powertrain control and monitoring systems will depend increasingly on physically oriented system models to reduce the complexity of existing control strategies and open the road to new environmentally friendly technologies. At the outset of this study a physics-based control-oriented dynamic model of a complete transient engine testing facility, consisting of a single cylinder engine, an alternating current dynamometer and a coupling shaft unit, is developed to investigate the functional relationships of the inputs, outputs and parameters of the system. Having understood these, algorithms for identifiability analysis and adaptive identification of parameters with physical interpretation are proposed. The efficacy of the recommended algorithms is illustrated with three novel practical applications. These are, the development of an on-line health monitoring system for engine dynamometer coupling shafts based on recursive estimation of shaft’s physical parameters, the sensitivity analysis and adaptive identification of engine friction parameters, and the non-linear recursive parameter estimation with parameter estimability analysis of physical and semi-physical cyclic engine torque model parameters. The findings of this research suggest that the combination of physics-based control oriented models with adaptive identification algorithms can lead to the development of component-based diagnosis and control strategies. Ultimately, this work contributes in the area of on-line fault diagnosis, fault tolerant and adaptive control for vehicular systems.

A complete chain of analysis and synthesis system identification tools for detectability assessment and adaptive identification of parameters with physical interpretation that can be found commonly in control-oriented powertrain models is presented. This research is motivated from the fact that future powertrain control and monitoring systems will depend increasingly on physically oriented system models to reduce the complexity of existing control strategies and open the road to new environmentally friendly technologies. At the outset of this study a physics-based control-oriented dynamic model of a complete transient engine testing facility, consisting of a single cylinder engine, an alternating current dynamometer and a coupling shaft unit, is developed to investigate the functional relationships of the inputs, outputs and parameters of the system. Having understood these, algorithms for identifiability analysis and adaptive identification of parameters with physical interpretation are proposed. The efficacy of the recommended algorithms is illustrated with three novel practical applications. These are, the development of an on-line health monitoring system for engine dynamometer coupling shafts based on recursive estimation of shaft’s physical parameters, the sensitivity analysis and adaptive identification of engine friction parameters, and the non-linear recursive parameter estimation with parameter estimability analysis of physical and semi-physical cyclic engine torque model parameters. The findings of this research suggest that the combination of physics-based control oriented models with adaptive identification algorithms can lead to the development of component-based diagnosis and control strategies. Ultimately, this work contributes in the area of on-line fault diagnosis, fault tolerant and adaptive control for vehicular systems.

Distributed parameter systems (DPS) comprise an important class of engineering systems ranging from "traditional" such as tubular reactors, to cutting edge processes such as nano-scale coatings. DPS have been studied extensively and significant advances have been noted, enabling their accurate simulation. To this end a variety of tools have been developed. However, extending these advances for systems design is not a trivial task . Rigorous design and operation policies entail systematic procedures for optimisation and control. These tasks are "upper-level" and utilize existing models and simulators. The higher the accuracy of the underlying models, the more the design procedure benefits. However, employing such models in the context of conventional algorithms may lead to inefficient formulations. The optimisation and control of DPS is a challenging task. These systems are typically discretised over a computational mesh, leading to large-scale problems. Handling the resulting large-scale systems may prove to be an intimidating task and requires special methodologies. Furthermore, it is often the case that the underlying physical phenomena span various temporal and spatial scales, thus complicating the analysis. Stiffness may also potentially be exhibited in the (nonlinear) models of such phenomena. The objective of this work is to design reliable and practical procedures for the optimisation and control of DPS. It has been observed in many systems of engineering interest that although they are described by infinite-dimensional Partial Differential Equations (PDEs) resulting in large discretisation problems, their behaviour has a finite number of significant components , as a result of their dissipative nature. This property has been exploited in various systematic model reduction techniques. Of key importance in this work is the identification of a low-dimensional dominant subspace for the system. This subspace is heuristically found to correspond to part of the eigenspectrum of the system and can therefore be identified efficiently using iterative matrix-free techniques. In this light, only low-dimensional Jacobians and Hessian matrices are involved in the formulation of the proposed algorithms, which are projections of the original matrices onto appropriate low-dimensional subspaces, computed efficiently with directional perturbations.The optimisation algorithm presented employs a 2-step projection scheme, firstly onto the dominant subspace of the system (corresponding to the right-most eigenvalues of the linearised system) and secondly onto the subspace of decision variables. This algorithm is inspired by reduced Hessian Sequential Quadratic Programming methods and therefore locates a local optimum of the nonlinear programming problem given by solving a sequence of reduced quadratic programming (QP) subproblems . This optimisation algorithm is appropriate for systems with a relatively small number of decision variables. Inequality constraints can be accommodated following a penalty-based strategy which aggregates all constraints using an appropriate function , or by employing a partial reduction technique in which only equality constraints are considered for the reduction and the inequalities are linearised and passed on to the QP subproblem . The control algorithm presented is based on the online adaptive construction of low-order linear models used in the context of a linear Model Predictive Control (MPC) algorithm , in which the discrete-time state-space model is recomputed at every sampling time in a receding horizon fashion. Successive linearisation around the current state on the closed-loop trajectory is combined with model reduction, resulting in an efficient procedure for the computation of reduced linearised models, projected onto the dominant subspace of the system. In this case, this subspace corresponds to the eigenvalues of largest magnitude of the discretised dynamical system. Control actions are computed from low-order QP problems solved efficiently online.The optimisation and control algorithms presented may employ input/output simulators (such as commercial packages) extending their use to upper-level tasks. They are also suitable for systems governed by microscopic rules, the equations of which do not exist in closed form. Illustrative case studies are presented, based on tubular reactor models, which exhibit rich parametric behaviour.

La stimulation électrique fonctionnelle (SEF) a le potentiel de fournir une amélioration active aux blessés médullaires en termes de mobilité, de stabilité et de prévention des effets secondaires.Dans le domaine des système SEF pour les membres inférieurs, le couple articulaire adéquat doit être fournie de façon appropriée pour effectuer le mouvement prévu et maintenir l'équilibre postural. Toutefois, les changements d'état du muscle tels que la fatigue musculaire est une cause majeure qui dégrade ses performances. En outre, la plupart des patients, dont la blessure médullaire est complète, n'ont pas le retour sensorielle qui permet de détecter la fatigue et les capteurs de couples in-vivo ne sont pas disponible à l'heure actuelle. Les systèmes conventionnels de commande SEF sont soit en boucle ouverte ou pas assez robustes aux changements d'état du muscle. L'objectif de cette thèse est le développement de la prédiction du couple articulaire et la commande en boucle fermée afin d'améliorer les performances de la commande SEF en termes de précision, de robustesse et de sécurité pour les patients.Afin de prédire le couple articulaire induit de la SEF, l'électromyographie (EMG) induit est utilisé pour corréler l'activité musculaire électrique et mécanique. Bien que la fatigue musculaire représente une variation dans le temps, une dépendance aux sujets et aux protocoles, la méthode proposée d'identification adaptative, basée sur le filtre de Kalman, est capable de prédire le couple articulaire variant dans le temps de manière systématique. La robustesse de la prédiction du couple articulaire a été évaluée lors d'une tâche de suivi de la fatigue en expérimentation chez des sujets blessés médulaires.Les résultats montrent une bonne performance de suivi des variations d'état des muscles en présence de fatigue et face à d'autres perturbations. Basé sur les performances de précision de la méthode prédictive proposée, une nouvelle stratégie de commande basée sur le retour EMG, «EMG-Feedback Predictive Control» (EFPC), est proposée afin de contrôler de manière adaptative les séquences de stimulation en compensant la variation dans le temps de l'état du muscle. De plus, cette stratégie de commande permet explicitement d'éviter d'appliquer une stimulation excessive aux patients, et de générer les séquences de stimulation appropriées pour obtenir la trajectoire désirée des couples articulaires. / Functional electrical stimulation (FES) has the potential to provide active improvement to spinal cord injured (SCI) patients in terms of mobility, stability and side-effect prevention. In the domain of lower limb FES system, elicited muscle force must be provided appropriately to perform intended movement and the torque generation by FES should be accurate not to lose the posture balance. However, muscle state changes such as muscle fatigue is a major cause which degrades its performance. In addition, most of the complete SCI patients don't have sensory feedback to detect the fatigue and in-vivo joint torque sensor is not available yet. Conventional FES control systems are either in open-loop or not robust to muscle state changes. This thesis aims at a development of joint torque prediction and feedback control in order to enhance the FES performance in terms of accuracy, robustness, and safety to the patients.In order to predict FES-induced joint torque, evoked-Electromyography (eEMG) has been applied to correlate muscle electrical activity and mechanical activity. Although muscle fatigue represents time-variant, subject-specific and protocol-specific characteristics, the proposed Kalman filter-based adaptive identification was able to predict the time-variant torque systematically. The robustness of the torque prediction has been investigated in a fatigue tracking task in experiment with SCI subjects. The results demonstrated good tracking performance for muscle variations and against some disturbances.Based on accurate predictive performance of the proposed method, a new control strategy, EMG-Feedback Predictive Control (EFPC), was proposed to adaptively control stimulation pattern compensating to time-varying muscle state changes. In addition, this control strategy was able to explicitly avoid overstimulation to the patients, and conveniently generate appropriate stimulation pattern for desired torque trajectory.