This thesis considers the application of information inequalities, Cramér-Rao type bounds, based on Fisher information, to linear systems. These tools are used to study the trade-offs between learning and performance in two application areas: adaptive control and control systems security. In the first part of the thesis, we study stochastic adaptive control of linear quadratic regulators (LQR). Here, information inequalities are used to derive instance-dependent  regret lower bounds. First, we consider a simplified version of LQR, a memoryless reference tracking model, and show how regret can be linked to a cumulative estimation error. This is then exploited to derive a regret lower bound in terms of the Fisher information generated by the experiment of the optimal policy. It is shown that if the optimal policy has ill-conditioned Fisher information, then so does any low-regret policy. This is combined with a Cramér-Rao bound to give a regret lower bound on the order of magnitude square-root T in the time-horizon  for a class of instances we call uninformative. The lower bound holds for all policies which depend smoothly on the underlying parametrization. Second, we extend these results to the general LQR model, and to arbitrary affine parametrizations of the instance parameters. The notion of uninformativeness is generalized to this situation to give a structure-dependent rank condition for when logarithmic regret is impossible. This is done by reduction of regret to a cumulative Bellman error. Due to the quadratic nature of LQR, this Bellman error turns out to be a quadratic form, which again can be interpreted as an estimation error. Using this, we prove a local minimax regret lower bound, of which the proof relies on relating the minimax regret to a Bayesian estimation problem, and then using Van Trees' inequality. Again, it is shown that an appropriate information quantity of any low regret policy is similar to that of the optimal policy and that any uninformative instance suffers local minimax regret at least on the order of magnitude square-root T. Moreover, it shown that the notion of uninformativeness when specialized to certain well-understood scenarios yields a tight characterization of square-root-regret. In the second part of this thesis, we study control systems security problems from a Fisher information point of view. First, we consider a secure state estimation problem and characterize the maximal impact an adversary can cause by means of least informative distributions -- those which maximize the Cramér-Rao bound. For a linear measurement equation, it is shown that the least informative distribution, subjected to variance and sparsity constraints, can be solved for by a semi-definite program, which becomes mixed-integer in the presence of sparsity constraints. Furthermore, by relying on well-known results on minimax and robust estimation, a game-theoretic interpretation for this characterization of the maximum impact is offered. Last, we consider a Fisher information regularized minimum variance control objective, to study the trade-offs between parameter privacy and control performance. It is noted that this can be motivated for instance by learning-based attacks, in which case one seeks to leak as little information as possible to a system-identification adversary. Supposing that the feedback law is linear, the noise distribution minimizing the trace of Fisher information subject to a state variance penalty is found to be conditionally Gaussian. / <p>QC 20210310</p><p>QC 20210310</p>

Les techniques d’usinage avancées sont un élément indispensable du développement des industries manufacturières. L’une de ces techniques, l’usinage à grande vitesse, constitue le sujet principal de cette thèse de doctorat. Ainsi, l’objectif majeur des travaux vise à améliorer la précision de contour dans le contexte de l’usinage multiaxes à grande vitesse de surfaces de forme libre, en agissant directement au niveau des boucles de commande d’axe. Pour cela, une première étape consiste à élaborer une stratégie permettant d’estimer le plus précisément possible l’erreur de contour pour différentes configurations de l’outil. Cette erreur de contour est ensuite minimisée grâce à l’adaptation hors ligne, pour un profil de pièce donné, des gains proportionnel et d’anticipation des régulateurs des boucles d’asservissement de la position de chaque axe. L’adaptation de ces gains est réalisée via un algorithme d’optimisation à l’aide d’un modèle non-linéaire du comportement de la machine, en considérant en particulier les frottements sur chacun des axes. L’optimisation permettant d’obtenir les gains des correcteurs des boucles de régulation tient compte des contraintes en termes de limitations cinématiques des axes (vitesse, accélération et jerk), de stabilité des boucles d’asservissement et de limites au niveau des courants des moteurs. Afin d’en faciliter la mise en oeuvre dans un cadre industriel, les stratégies développées s’avèrent directement implantables au sein des commandes numériques actuellement sur le marché, exploitant toutes les possibilités de la structure de commande classique de l’entraînement d’axe. / The advanced machining techniques are always the backbone of the manufacturing industries. Among such techniques, high speed machining is the main subject of this PhD thesis. Indeed, the main objective of this work is to improve the contouring accuracy in multi-axis high speed machining of free-form surfaces, directly acting inside the axis control loops. To do that, a first step aims at elaborating a strategy to estimate as accurately as possible the contour error for different tool configurations. This contour error is then minimized by means of an off-line adaptation for a given profile of the proportional and feedforward gains of the axis position loop controllers. This gain adaptation is performed via an optimization algorithm that considers a nonlinear model of the machine behaviour, in particular including friction related to each axis. This optimization leading to the controllers gains takes into account several constraints, including the axis kinematic (velocity, acceleration and jerk) limitations, the stability of the controlled loops and the motor current limits. Finally, to help their integration within an industrial framework, the developed strategies can be directly implemented in commercial CNC, by exploiting all possibilities of the classical control structure of axis drive.

Indiana University-Purdue University Indianapolis (IUPUI) / The main contributions in this thesis include the selectively decentralized method in solving multi-agent reinforcement learning problems and the discretized Markov-decision-process (MDP) algorithm to compute the sub-optimal learning policy in completely unknown learning and control problems. These contributions tackle several challenges in multi-agent reinforcement learning: the unknown and dynamic nature of the learning environment, the difficulty in computing the closed-form solution of the learning problem, the slow learning performance in large-scale systems, and the questions of how/when/to whom the learning agents should communicate among themselves. Through this thesis, the selectively decentralized method, which evaluates all of the possible communicative strategies, not only increases the learning speed, achieves better learning goals but also could learn the communicative policy for each learning agent. Compared to the other state-of-the-art approaches, this thesis’s contributions offer two advantages. First, the selectively decentralized method could incorporate a wide range of well-known algorithms, including the discretized MDP, in single-agent reinforcement learning; meanwhile, the state-of-the-art approaches usually could be applied for one class of algorithms. Second, the discretized MDP algorithm could compute the sub-optimal learning policy when the environment is described in general nonlinear format; meanwhile, the other state-of-the-art approaches often assume that the environment is in limited format, particularly in feedback-linearization form. This thesis also discusses several alternative approaches for multi-agent learning, including Multidisciplinary Optimization. In addition, this thesis shows how the selectively decentralized method could successfully solve several real-worlds problems, particularly in mechanical and biological systems.

No description available.

<div>Insects and hummingbirds exhibit extraordinary flight capabilities and can simultaneously master seemingly conflicting goals: stable hovering and aggressive maneuvering, unmatched by small-scale man-made vehicles. Given a sudden looming visual stimulus at hover, a hummingbird initiates a fast backward translation coupled with a 180-degree yaw turn, which is followed by instant posture stabilization in just under 10 wingbeats. Considering the wingbeat frequency of 40Hz, this aggressive maneuver is accomplished in just 0.2 seconds. Flapping Wing Micro Air Vehicles (FWMAVs) hold great promise for closing this performance gap given its agility. However, the design and control of such systems remain challenging due to various constraints.</div><div><br></div><div>First, the design, optimization and system integration of a high performance at-scale biologically inspired tail-less hummingbird robot is presented. Designing such an FWMAV is a challenging task under the constraints of size, weight, power, and actuation limitations. It is even more challenging to design such a vehicle with independently controlled wings equipped with a total of only two actuators and be able to achieve animal-like flight performance. The detailed systematic solution for the design is presented, including system modeling and analysis of the wing-actuation system, body dynamics, and control and sensing requirements. Optimization is conducted to search for the optimal system parameters, and a hummingbird robot is built and validated experimentally.</div><div><br></div><div>An open-source high fidelity dynamic simulation for FWMAVs is developed to serve as a testbed for the onboard sensing and flight control algorithm, as well as design, and optimization of FWMAVs. For simulation validation, the hummingbird robot was recreated in the simulation. System identification was performed to obtain the dynamics parameters. The force generation, open-loop and closed-loop dynamic response between simulated and experimental flights were compared and validated. The unsteady aerodynamics and the highly nonlinear flight dynamics present challenging control problems for conventional and learning control algorithms such as Reinforcement Learning.</div><div><br></div><div>For robust transient and steady-state flight performance, a robust adaptive controller is developed to achieve stable hovering and fast maneuvering. The model-based nonlinear controller can stabilize the system and adapt to system parameter changes such as wear and tear, thermo effect on the actuator or strong disturbance such as ground effect. The controller is tuned in simulation and experimentally verified by hovering, point-to-point fast traversing, and following by rapid figure-of-eight trajectory. The experimental result demonstrates the state-of-the-art performance of the FWMAV in stationary hovering and fast trajectory tracking tasks, with minimum transient and steady-state error.</div><div><br></div><div>To achieve animal level maneuvering performance, especially the hummingbirds' near-maximal performance during rapid escape maneuvers, we developed a hybrid flight control strategy for aggressive maneuvers. The proposed hybrid control policy combines model-based nonlinear control with model-free reinforcement learning. The model-based nonlinear control stabilizes the system's closed-loop dynamics under disturbance and parameter variation. With the stabilized system, a model-free reinforcement learning policy trained in simulation can be optimized to achieve the desirable fast movement by temporarily "destabilizing" the system during flight. Two test cases were demonstrated to show the effectiveness of the hybrid control method: 1)a rapid escape maneuver observed in real hummingbird, 2) a drift-free fast 360-degree body flip. Direct simulation-to-real transfers are achieved, demonstrating the hummingbird-like fast evasive maneuvers on the at-scale hummingbird robot.</div>

No description available.

A project report submitted to the Faculty of Engineering, University of the Witwatersrand, Johannesburg, in partial fulfillment of the requirements for 'the degree of Master of Science in Engineering. Johannesburg 1991. / The report describes the design of an algorithm which can be used for automatic controller tuning purposes. It uses an on-line parameter estimator and a pole assignrnent design method. The resulting control law is formulated to approximate a proportional-integral (PI) industrial controller. The development ofthe algorithm is based on the delta-operator, Some implementation aspects such as covariance resetting, dead zone, and signal conditioning are also discussed. Robust stability and performance are two issues that govern the design approach. Additionally transient and steady state system response criteria are utilized from the time and frequency domains. The design work is substantiated with the use of simulation and real plant tests. / AC2017

La presente tesis tiene como principal objetivo el diseño de un sistema de control para la navegación automática de vehí-culos subacuáticos autónomos y teleoperados, asegurando propiedades de amplia maniobrabilidad y de alta performance de control en los 6 grados de libertad de movimiento, ante incertidumbres y variaciones temporales de la dinámica y bajo el efecto de perturbaciones externas del entorno y de cable. El control desarrollado es del tipo adaptivo y esta basado en el método de gradiente de velocidad con proyección dinámica suave, apto para una clase general y casi arbitraria de cambios paramétricos comunes a operacio-nes en la Ingeniería Oceánica. Se incluye en el diseño del sistema la dinámica parásita de los propulsores mediante la utilización de observadores de estados y disturbios para establecer la entrada óptima de los actuadores. Esta modifica-ción produce una diferencia entre la fuerza de propulsión real y la fuerza ideal requerida por la acción de control, la cual es tratada como una perturbación endógena. Para este diseño se analiza en detalle la convergencia de los errores de trayecto-ria espacial y cinemática, la acotabilidad de las variables del lazo de control y la performance transitoria.También se realizó el modelado del sistema barco-cable-vehículo para corrientes estacionarias y olas de componentes de baja y media frecuen-cia. Los resultados sugieren que la magnitud de la perturba-ción del cable (denominada perturbación exógena) puede ser controlada mediante la regulación del largo del cable. Para ambos tipos de perturbaciones (endógena y exógena) se demuestra mediante teoremas que el controlador diseñado es totalmente estable. Esto significa que el error de seguimiento de trayectorias permanece acotado, alrededor del punto de equilibrio del sistema no perturbado, para perturbaciones y condiciones iniciales acotadas. El orden del error depende de la magnitud de las perturbaciones. Los resultados perseguidos se orientaron a la aplicación en operaciones planificadas de muestreo y rastreo sobre el fondo marino, reduciendo eventualmente los tiempos de navegación a un mínimo sobre las trayectorias especificadas. Para la realización de esto último se diseñó un algoritmo de optimización del tiempo de recorrido de trayectorias de referencia geométricas. El algo-ritmo se incluyó en el esquema de control adaptivo demos-trándose las propiedades de convergencia para el sistema completo. / The present Thesis has as principal objective the design of a control system for the automatic navigation of autonomous and teleoperated underwater vehicles, assuring properties of high maneuverability and high control performance in the six degrees of freedom, in front of uncertainties and time-varying dynamics, under the effect of external perturbation of the environment and the cable. The developed control is based on a speed-gradient adaptive law with a smooth dynamic projection, suitable for a general and almost arbitrary class of parametric changes, commons to operations in Oceanic Engineering. The parasitic dynamics of the thrusters is included in the system design by means of the use of state/disturbance observers to establish the optimal input to the actuators. This modification causes a difference between the ideal thrust required by the control action and the real thrust. The result is a force error that is dealt as an endoge-nous perturbation. The error convergence in the spatial and cinematic trajectories, the boundness of the variables in the control loop and the transitory performance are analyzed in detail. The modelling of the system ship-cable-vehicle is also realized for stationary currents and waves of low and middle frequency. These results suggest that the magnitude of the cable perturbation (namely exogenous perturbation) can be controlled by means of the cable length regulation. For both types of perturbation (endogenous and exogenous), the total stability of the designed controller is proved by theorems. This stability class means that the tracking error keeps bounded around the equilibrium point of the nonpertur-bed system for bounded perturbations and bounded initial conditions. The order of error depends on the magni-tude of the perturbations. The following results are oriented to the application in planned operations of sampling and path tracking on the sea bottom, eventually reducing the naviga-tion time to a minimum over the specified trajectories. For the accomplishment of this last objective, an algorithm is designed to optimize the time used to cover the geometric trajectory reference. The algorithm is included in the scheme of the adaptive control and the convergence property is proved for the complete system.

No description available.

Although PID controllers work well on Miniature Air Vehicles (MAVs), they require tuning for each MAV. Also, they quickly lose performance in the presence of actuator failures or changes in the MAV dynamics. Adaptive control algorithms that self tune to each MAV and compensate for changes in the MAV during flight are explored. However, because the autopilots on MAVs are small, many of the adaptive control algorithms like those that employ least squares estimation may take too much code space, memory, and/or computing power. In this thesis we develop several Lyapunov-based model reference adaptive control (MRAC) schemes that are both simple and efficient with the MAV autopilot resources. Most notable are the L1 controllers that have all the benefits of traditional MRACs but have reduced high frequency content to the actuators. The schemes control both roll and pitch through aileron and elevator commands. Flight test results for the schemes are also compared.