Fully-actuated multi-rotor aerial platforms are receiving increasing research interests for the capability of six degree-of-freedom (DOF) motions such as hovering at non-horizontal attitude angles. Existing fully-actuated aerial vehicles have demonstrated such capability for a limited range of angles and limited thrust efficiencies. This thesis presents an over-actuated aerial platform that achieves maneuvering at arbitrary attitudes with uniformly high thrust efficiency over its achievable configuration space. A novel vectoring thrust force actuator by mounting a regular quad copter on a passive mechanical gimbal mechanism is proposed. The UAV platform achieves full six DOF motion with redundancies from four of these vectoring thrust actuators. We present the hierarchical controller that generates the high level virtual wrench command allocated to each gimbal actuator and the low-level actuator control to track the commanded wrench. And we demonstrate the UAV platform's 6 DOF maneuvers by both simulations and real-world experiments on a prototype we built.

Aerodynamic effects largely affect the performance of aerial vehicles, especially on over-actuated vehicles that could be subject to different airflow configurations. For those aerodynamic effects that are difficult to model but could appear repeatedly, iterative learning control (ILC) has great potential to improve the system performance. This thesis presents the applications of both model-based and data-driven ILC algorithms on the over-actuated aerial platform and shows great improvements against the aerodynamics effects. A formulation is demonstrated to convert the closed-loop dynamics of the over-actuated aerial platform to linear model with six independent control channels. Model-based and data-driven ILC are applied on one or more control channels, and by simulations and real-world experiments, the ILC algorithms are shown to have great improvement and fast convergence rate against a variety of aerodynamic effects.

Bidirectional power-transistor rectifier-inverter circuits have broad applications, ranging from uninterruptible power supplies, and electric vehicle motor drives, to renewable energy wind turbines. Sizable and expensive DC-link capacitors are commonly placed between the rectifier and inverter to decouple the circuit's nonlinear dynamics and facilitate control. Aiming to minimize the buffering capacitor and gain high performance for the nonlinear dynamics via the pulse width modulated (PWM) control of the transistor switches, this dissertation contributes to this field in three aspects. First, we developed a high-fidelity nonlinear discrete-time model of the continuous-time nonlinear system to facilitate model-based control design. This model is also more accurate than existing models published in the literature. Second, we proposed a novel framework of multi-input multi-output (MIMO) nonlinear model predictive control (NMPC), where both the non-linear model and approximate linearized model are used to strike the balance of the computational complexity and control performance. Third, we implemented the NMPC in real-time by field programmable gate array (FPGA), where the controller is modeled by a neural network trained offline by machine learning. We performed a simulation and experiment of the power converter control on a three-phase permanent magnet synchronous motor (PMSM) drive and demonstrated superior performance in comparison to the feedback linearization control (FBLC), power-balancing schemes, and conventional cascaded PI control.

Traditional quadcopters possess only four controllable inputs and are inherently underactuated, limiting their ability to interact with an environment or manipulate a payload. This work discusses a quadcopter architecture incorporating twisting and tilting joints to fully decouple the position and attitude dynamics, permitting full six-degree-of-freedom maneuvering. A dynamic model is presented which includes both the gyroscopic effects of the propellers and the configuration-dependent inertia of the articulated components. A control framework is developed first for a quarter copter model, then generalized to the full copter on a spherical joint, and finally to the full copter with unconstrained base motion on the special Euclidean group SE(3). Simulations are conducted on a Simscape multibody model, and the results are presented to demonstrate the capabilities of the craft.

This dissertation presents modeling, identification, fault detection, and control of an active magnetic bearing system, and the experimental results for applications in metal cutting and dynamic vibration suppression. A high order ARX identification method conducted within stabilized feedback loops followed by balanced truncation model reduction was efficient and effective in obtaining the open loop unstable multivariable AMBS dynamic model, which agrees well with the system's frequency response data. Also the severe flexible nature of the spindle is seen in the frequency response both while stationary and spinning, which facilitates the need for robust controller design. The experimentally obtained model is augmented to include the tip dynamics by impulse hammer testing that allows for the inclusion on cutting dynamics into the overall control system.

In cutting application there are two main concerns: the tip tracking a profile and avoidance of chatter instability. Therefore, based on the identified model, a two-step optimal repetitive controller was designed for minimal error tracking. The first controller consisted of a control input weighted linear quadratic control with integral action that ensured stability, broadband performance, and robustness. Then a plug-in multivariable repetitive controller with modifications for enhanced performance and non-harmonic compensation was designed by optimal model matching that highlights the classical trade-off between robustness and performance. A cutting force model was incorporated into the overall system, and stability lobes obtained for optimal cutting conditions.

Vibration reduction is crucial in applications of magnetic bearings such as energy storage. An adaptive feedforward control was seen to be effective in reducing housing vibrations even while the spin frequency is changing. The regressors were filtered through a multivariable filter obtained from gradient descent. The adaptive controller has a linear time invariant state space representation from which the deviations of the add-on adaptive from the nominal sensitivity and complementary sensitivity can be analyzed. Experimental results did indeed show that vibrations can be reduced even while the spindle speed is changing.

The single input single output adaptive inverse control was extended to the multivariable case and demonstrated to be effective in rejecting narrow and broadband disturbances. Two formulations that can accommodate multivariable plants were considered: the receding horizon adaptive control and the decoupling adaptive control. Both methods need a multivariable adaptive filter such as RLS or LMS; in this thesis a multivariable lattice RLS filter was chosen for its numerical stability and $O(N)$ complexity. Experimental results on the magnetic bearing system show the effectiveness of both control schemes.

Lastly, there were two main results for the standard single phase boost rectifier: 1) The formulation of a linear model in the DQ domain suitable for controller design 2) A MIMO repetitive add-on controller with peak filter modification was designed using the linear model to reduce harmonics not compensated for by a nominal controller. In the process of deriving the linear model, analysis of the equilibrium points gave the boost condition for the rectifier. Experimental results proved the effectiveness of using the linear model for controller design and the repetitive controller for reducing harmonics. The methodology can be used for designing controllers that operate at near unity power factor.

In the past few decades, computational power and speed has made it such that the Controls literature has moved away from addressing issues of finite-word-length (FWL) issues, quantization, and limited computational resources. On the other hand, the signal processing community has studied this issue extensively. In recent years, the introduction of Nano and Microelectromechanical systems (MEMS) with large bandwidth systems requires the use of high-sampling rate controllers. To satisfy such high sampling-rates, fixed-point based platforms such as Field Programmable Gate Arrays (FPGAs) and fixed-point micro-controllers are needed. This trend results in a need for high-sampling rate controllers that are more sophisticated than simple loop shaping while addressing the issues of FWL effects and limited computational resources. The aim of this dissertation is to introduce novel controllers that incorporate signal processing techniques and address these issues with controller design/realizations to control high bandwidth electro-mechanical systems.

Multirotor aerial platforms have obtained growing attentions in industry and academia, for its simplicity in mechanical structure, agility in maneuverability and ability for vertical take-off and landing (VTOL). Conventional multirotor has underactuated dynamics, and can not be fully controlled in 6 Degree-of-Freedom (DoF). In fact, only its three-dimensional position and yaw angle, called the flat outputs, can be controlled independently. However, for certain applications, such as perching on a vertical vertical wall or flying in a narrow space, the the non-flat outputs, the roll and pitch angles, are independently specified from the position requirements at some particular time. These tasks require the independent control of position and attitude at least partially for certain instants, and are generally challenging for multirotor platforms.

This dissertation addresses this issue in two aspects. Firstly, an algorithm is designed for the conventional quadcopter platforms to generate trajectories for tasks with requirements on both position and attitude. It is formulated as an optimization, and converted into a series of convex problems to solve. Constraints on dynamics, space limitations, inputs and states are explicitly included. The algorithm is verified numerically on the task of quadcopter perching at the specified location on a vertical wall.

Secondly, a fully actuated multirotor aerial platform is proposed. Commercial quadcopters and passive hinges are used to generate tiltable thrust vectors during flight. This platform has a salient feature for mechanical simplicity, as it does not require additional actuators to control the directions of thrust vectors. A controller for the proposed multirotor platform is designed to enable independent control of position and attitude.

The proposed multirotor platform has overactuation in dynamics, which renders a redundancy of 2 DoF for inputs. A new controller is proposed, under which the input allocation scheme searches within this redundancy for smaller thrust forces required to hover at different attitudes. The range of achievable attitudes is enlarged under this new scheme compared with the previously proposed controller, under the same thrust saturation limit for the platform actuators.

These controllers are validated with both simulation and experiments and demonstrated by the proposed multirotor aerial platform hovering at non-horizontal attitudes, or tracking independent trajectories for position and attitude simultaneously.

Energy consumption and efficiency continue to be an area of high interest with the diminishing supply and rising costs of fossil fuels. In the United States, the industrial and transportation sectors consume a significant portion of the resources and total energy, accounting for significant fossil-fuel-related environmental impacts such as greenhouse gas pollution and global warming. It has been estimated that between 20 to 50\% of the energy consumed is lost as waste heat in the form of hot exhaust gases, cooling water, and hot equipment surfaces. By converting the heat to a more useful energy form, waste heat recovery (WHR) technologies are capable of creating emission-free, low-cost, and sustainable energy sources. However, WHR systems, if not properly designed and operated, may not be economically and or even feasible, particularly with low-grade and highly-transient heat sources. The aim of this dissertation is to predict WHR performance and feasibility through physics-based static and dynamic Rankine Cycle (RC) modeling for low- and medium-grade transient heat sources. The dynamic models provided are low-order, control-oriented, and suitable for tractable estimation and control methods that improve operational performance and observability of key process variables. Thermoproperties for water and steam are obtained from equations of state formulations based on Helmholtz and Gibbs free energy. For general fluids and their mixtures, equations of state can be obtaned from a presented least squares property fit routine. This dissertation applies the modeling, estimation, and control methods on a WHR application for heavy-duty diesel powertrain where important RC variables are captured and monitored while satisfying specified design and operating constraints. An organic working fluid mixture of ethanol and water is considered for this organic RC (ORC) application. Estimation and control methods are simulated on a dynamic ORC model of different architectures and actuator configurations. Power load following is also examined. The simulation results indicate improved ORC operation with feedback controls over the open-loop case. Multi-input multi-output controllers that rely on state estimation feedback show better tracking performance than single-input single-output controllers. Finally, a library of static and dynamic component models are provided as a useful aid for designing general WHR systems.

Motivated by machining of thin-walled parts this dissertation investigates the modeling and control of dynamic systems, which feature finite-dimensional and distributed-parameter structures interacting with one another over time-varying contacts and time-delayed feedback. The research studies and creates methods for modeling and control of such systems and applies them to a proposed innovative use of industrial robot manipulators and end effectors to act as an active steady rest in the machining of thin-walled parts.

On modeling the combined discrete-continuous structural and machine dynamics with time varying contacts, finite element method (FEM) is employed to discretize and extract dynamic modes of the continuous structure. The time delayed feedback in the machining process is approximated by the semi-discretization method, where only the delay part of the otherwise continuous-time system is approximated by discrete-time delays and proper sampling and holding functions. To enable digital control design, sampling and zero-order hold circuit modeling is applied to the actuators in conjunction with the semi-discretization for the delayed feedback. Followed by lifting technique on the periodic-varying system and order reduction, discrete-time control oriented models for the entire machining system with the actively controlled robots and end effectors are constructed.

Linear optimal control and robust control are investigated and applied to the proposed robot assisted machining system models. In the linear quadratic state estimator feedback control, a new approach is proposed for the formulation of the cost function weighting matrices. The approach essentially places weighting on outputs and their derivatives since output has clear physical meaning in connection to the process performance. The optimal linear state estimators, aka Kalman filters, are designed taking an approach that exploits the dual properties between the linear optimal regulators and state estimators. In the robust control, parameters in the machining process, particularly the cutting stiffness that depends on the tool conditions and part material, are isolated as uncertainties in formulating the control oriented models for mu-synthesis based robust control design.

The above control design methods are applied to the turning of a thin-walled circular wheel with the actively controlled robotic steady rest. The simulation results show that the feedback controlled robotic active steady rest not only can stabilize an otherwise unstable machining system, caused by the well-known machining chatter due to the delayed feedback, but also can improve the machined forms and precision. The increased stability margins with higher speeds and larger depths of cut than possible without the proposed approach suggest the potential of improving the productivity and quality in the machining of thin-walled parts with the proposed control methods for the robotic steady rests.

Rotating machinery is commonly used in mechanical systems, including machining tools, aircraft gas turbine engines, and other industrial applications. However, the synchronous vibration due to the rotor mass unbalance is the common disturbance source encountered in rotor operations. Unbalance occurs when the rotor's inertia axis does not match its axis of geometry. Given the cost and difficulty of the perfect balancing and the current trend for high-speed operations, which is directly correlated with the greater centrifugal unbalance forces, vibration control is essential in improving achieving longer bearing, spindle, and tool life in high-speed machining. As opposed to conventional ball bearings, Active Magnetic Bearings (AMBs) provide both the contact-less support and the possibility for applying real-time force to regulate the rotor-dynamic behaviors. This additional control facility allows the rotor to spin around its principal axis of inertia if a sufficient air gap exists between the rotor and housing. Consequently, by annihilating the rotor's centrifugal force, the rotor vibration dramatically decreases and no reaction forces are transmitted to the housing. Beyond the functionality of AMB in reducing the rotor vibration, the aspects of safety and reliability have become important in all rotor operations. The AMB is a highly nonlinear and open-loop unstable system, and therefore it is common that the control system includes different controllers, tailored for different regimes of the operating conditions. However, the stability of the entire control system depends on the flawless operation of its sensors and actuators. Otherwise, the malfunctioning of any system components may disturb the stability of the supported rotor resulting in damage to the whole system. Hence, to ensure the safe operation and reliable performance of AMB, an online Fault Detection and Isolation (FDI) scheme is necessary to identify the faulty components and safeguard the system to remain in a safe region when faults are detected.

Given the inherited cost and complexity of adding redundant physical components, this dissertation presents a novel model-based FDI scheme for the AMB based on analytical redundancy by integrating two linear estimators called the Game-Theoretic Detection Filter (GTDF) and the Unknown Input Observer (UIO). Optimal inversion filters for both estimation and compensation are also introduced for the unified framework of the integrated fault detection, estimation, and Disturbance Observer-Based Control (DOBC) that safely switches between controllers when faults are detected. By extending and exploiting the Youla parameterization of stabilizing controllers, where the robust feedback control, the fault detection, and DOBC are realized by a bank of full-order state observers running in parallel, the controller in each regime always stabilizes the AMB. Lastly, a novel multi-variable Adaptive Feedforward Control (AFC) scheme featured by optimal delay-inversion-based compensation is proposed and implemented for its premium on stability and its simplicity in selecting the delay length as the single design parameter for the multi-variable controller to suppress the synchronous vibrations during varying and constant speed operations.