京都大学 / 新制・課程博士 / 博士(工学) / 甲第23158号 / 工博第4802号 / 新制||工||1751(附属図書館) / 京都大学大学院工学研究科電気工学専攻 / (主査)教授 引原 隆士, 教授 大村 善治, 特定講師 木村 真之 / 学位規則第4条第1項該当 / Doctor of Philosophy (Engineering) / Kyoto University / DFAM

The rising levels of autonomy exhibited by complex cyber-physical systems have brought questions related to safety and adaptation to the forefront of the minds of controls and robotics engineers. Often, such autonomous systems are deemed to be safety-critical in the sense that failures during operation could significantly harm the system itself, other autonomous systems, or, in the worst-case, humans interacting with such a system. Complicating the design of control and decision-making algorithms for safety-critical systems is that they must cope with various degrees of uncertainty as they are deployed autonomously in increasingly real-world environments. These challenges motivate the use of learning-based techniques that can adapt to such uncertainties while adhering to safety-critical constraints. The main objective of this dissertation is to present a unified framework for the design of controllers that learn from data online with formal guarantees of safety. Rather than using a controller trained on an a priori dataset collected offline that is then statically deployed on a system, we are interested in using real-time data to continuously update the control policy online and cope with uncertainties that are challenging to characterize until deployment. We approach the problem of designing such learning-based control algorithms for safety-critical systems through the use of certificate functions, such as Control Lyapunov Functions (CLFs) and Control Barrier Functions (CBFs), from nonlinear control theory. To this end, we first discuss how modern data-driven techniques can be integrated into traditional adaptive control frameworks to develop classes of CLFs and CBFs that facilitate the design of both controllers and learning algorithms that guarantee, respectively, stability and safety by construction. Next, we shift from the problem of safe adaptive control to safe reinforcement learning where we demonstrate how similar ideas from adaptive control can be extended to safely learn the value functions of optimal control problems online using data from a single trajectory. Finally, we discuss an extension of the aforementioned approaches to richer control specifications given in the form of temporal logic formulas, which provide a formal way to express complex control objectives beyond that of stability and safety. / 2025-05-30T00:00:00Z

This thesis develops Lyapunov-based control techniques for nonlinear process systems subject to input constraints and stochastic uncertainty. The problems considered include those which focus on the null-controllable region (NCR) for unstable systems. The NCR is the set of states in the state-space from where controllability to desired equilibrium point is possible. For unstable systems, the presence of input constraints induces bounds on the NCR and thereby limits the ability of any controller to steer the system at will. Common approaches for applying control to such systems utilize Control Lyapunov Functions (CLFs) . Such functions can be used for both designing controllers and also preforming closed--loop stability analysis. Existing CLF-based controllers result in closed--loop stability regions that are subsets of the NCR and do not guarantee closed--loop stability from the entire NCR. In effort to mitigate this shortcoming, we introduce a special type of CLF known as a Constrained Control Lyapunov Function (CCLF) which accounts for the presence of input constraints in its definition. CCLFs result in closed--loop stability regions which correspond to the NCR. We demonstrate how CCLFs can be constructed using a function defined by the NCR boundary trajectories for varying values of the available control capacity. We first consider linear systems and utilize available explicit characterization of the NCR to construct CCLFs. We then develop a Model Predictive Control (MPC) design which utilizes this CCLF to achieve stability from the entire NCR for linear anti-stable systems. We then consider the problem of nonlinear systems where explicit characterizations of the NCR boundary are not available. To do so, the problem of boundary construction is considered and an algorithm which is computationally tractable is developed and results in the construction of the boundary trajectories. This algorithm utilizes properties of the boundary pertaining to control equilibrium points to initialize the controllability minimum principle. We then turn to the problem of closed--loop stabilization from the entire NCR for nonlinear systems. Following a similar development as the CCLF construction for linear systems, we establish the validity of the use of the NCR as a CCLF for nonlinear systems. This development involves relaxing the conditions which define a classical CLF and results in CCLF-based control achieving stability to an to an equilibrium manifold. To achieve stabilization from the entire NCR, the CCLF-based control design is coupled with a classical CLF-based controller in a hybrid control framework. In the final part of this thesis, we consider nonlinear systems subject to stochastic uncertainty. Here we design a Lyapunov-based model predictive controller (LMPC) which provides an explicitly characterized region from where stability can be probabilistically obtained. The design exploits the constraint-handling ability of model predictive controllers in order to inherent the stabilization in probability characterization of a Lyapunov-based feedback controller. All the proposed control designs along with the NCR boundary computation are illustrated using simulation results. / Thesis / Doctor of Philosophy (PhD)

<p> The topic of study within includes the development and application of nonlinear control technologies on sampled systems. Discrete control structures are introduced that expand on existing differential geometric and predictive control methods. The differential geometric techniques are described from the error trajectory context, which are typically only derived for continuous application. The discrete error trajectory controllers introduced have one of two configurations. The first configuration requires satisfaction of the error trajectory objective at the next sampling interval through prediction of system behaviour over the controller sampling interval. This objective found limited success and it is observed that satisfaction of the error trajectory objective at discrete intervals does not generally result in the intended response. The second configuration minimizes the integrated distance from the error manifold defined by the error trajectory objective over the entire controller sampling interval. It is observed that this integrated error trajectory controller best emulates the intent of the continuous controller in the discrete domain. Techniques borrowed from predictive control are incorporated into the integrated error trajectory controller such as input move suppression and constraints to produce an optimal error trajectory controller, further improving performance.</p> <p> The predictive control method introduced utilizes a transformation of the input space. The differentiating property of input transformation predictive control (ITPC) from other methods is the prediction technique that is capable of estimating the future behaviour of nonlinear systems through elementary matrix operations similar to the dynamic matrix control (DMC) prediction technique. This is achieved by separation of the steady state and dynamic system properties and the introduction of an intermediate state prediction layer. This allows for the nonlinear prediction of system behaviour without the need to numerically integrate the system model.</p> <p> Two example systems are used to demonstrate application of the discrete error trajectory and ITPC on nonlinear controllers. Performance for these control structures is compared to technologies accepted within the control community for a broad range for characteristics including, computation efficiency, design effort and other nonlinear performance criteria, with favourable results.</p> / Thesis / Master of Engineering (MEngr)

The stability and safety of robotic systems are heavily impacted by delays and parametric uncertainties due to external disturbances, modeling inaccuracies, reaction forces, and variations in dynamics. This work addresses the effects of parametric uncertainties in the application of payload transportation by robotic systems that involve time delays and state constraints. The problem is split into two research questions: control of a quadrotor UAV in the presence of delays and control of robotic systems with state constraints. The first two papers explore the approaches for remotely operated quadrotors in the presence of delays and uncertainties. Specifically, the first paper surveys the existing methods for controlling a payload-carrying UAV and further presents a class of control techniques in theory that focus on time-delayed systems. The second paper proposes an adaptive control solution for the tracking control of a quadrotor UAV to transport various unknown payloads in the presence of unknown time-varying delays. The proposed controller is robust to modeling uncertainties and does not require knowledge of the uncertainties' bounds. The performance of the controller is verified on a MATLAB-SIMULINK simulated environment. The final three papers deal with enforcing state constraints on tracking control to ensure the safety of the robots in the presence of parametric uncertainties. The third paper exploits state constraints in the post-grasping scenario of the space debris disposal application. This work proposes a robust control for a space robot to follow the desired trajectory without any violation to safely grasp, carry, and release unknown payloads in their respective regions. The controller is tested in a MATLAB-SIMULINK environment with the dynamics of a planar space robot. The fourth paper introduces an adaptive control technique without any a priori knowledge of the system dynamics or the bounds of uncertainties to impose state constraints in control. The proposed controller is designed for a generic Euler-Lagrangian system in the presence of parametric uncertainties, where the state-dependent nature of the uncertainties introduces unboundedness in the overall uncertainty. The controller is validated in simulation using a robotic manipulator in a pick-and-place operation. The final paper proposes an adaptive controller for the tracking control of an experimental planar space robot. The proposed controller enforces constraints on the robot's states and their derivatives on the tracking control for transporting different payloads without any knowledge of the dynamics of the robot or the bounds of the uncertainties. The controller is validated on the experimental space robot. The stability of the proposed controllers is studied analytically using the Lyapunov theory. The results are presented with various plots and numerically analyzed on the metrics of root mean squared errors and peak errors.

This thesis involves the modeling, validation, and control studies of a Control-Oriented, Reconfigurable, and Acausal Floating Turbine Simulator (CRAFTS), that is currently under development. CRAFTS uses Modelica®, an object-oriented, declarative, multi-domain modeling language for physical system modeling in the Dymola environment. The CRAFTS simulator facilitates rapid dynamic simulation of wind turbines with various model variants and enables control co-design. A major emphasis of this thesis is in the validation of the CRAFTS simulator for a 15-MW land-based wind turbine through several test cases. These test cases were collaboratively developed in conjunction with other participating research entities. CRAFTS has undergone rigorous testing, with a particular emphasis on comparison against the industry standard OpenFAST platform (developed by the National Renewable Energy Lab (NREL)) as well as experimental data. Open loop testing scenarios scrutinize the wind turbine dynamic conditions such as varying rotor speed and pitching angle maneuvers. Diverse combinations of ramp and step commands have been employed to modulate rotor speeds and pitching angles. Validation results indicate very good agreement between CRAFTS and baseline results. CRAFTS was also tested under various types of closed-loop control scenarios, such as different types of wind profiles and various wind velocities. Wind types encompass stepped winds, wind gusts, steady winds, and sinusoidal wind patterns. In closed loop testing, firstly an industry standard controller ROSCO (also developed by NREL) was used. Thereafter, a nonlinear controller developed in our prior research was implemented and investigated. The closed loop performance of the CRAFTS model was compared with OpenFAST. The tests confirmed the validity of the CRAFTS model under closed-loop and also validated the nonlinear controller. The work was a critical element in the development of the CRAFTS simulator. Validation tests provided valuable insight into the accuracy of the underlying physics and often provided valuable feedback that led to model improvements. The work has laid the foundations for more advanced research, especially in the area of multivariable control design for floating offshore wind turbines.

No description available.

No description available.

No description available.

No description available.