This thesis describes the design and implementation of an induction motor field-oriented controller. Simulation studies are carried out on important aspects of the drive system, based on which a high-performance induction motor drive control is designed and implemented using a PC computer and interface hardware. The simulation software is based on a mathematical model that uses tensor techniques and includes the pulse width modulator and the flux, current, speed and position calculations.

The main work of this thesis is concerned with the comparison of conjugate gradient with hybrid conjugate gradient methods when they are applied to optimal control problems. Descriptions of the conjugate gradient. and hybrid conjugate gradient methods, for general optimisation, in finite and infinite dimensions are also given. The numerical methods for solving the differential equations and the line searches required in the optimisation are discussed next.

No description available.

No description available.

Artificial neural networks (ANNs) are highly-capable alternatives to traditional problem solving schemes due to their ability to solve non-linear systems with a nonalgorithmic approach. The applications of ANNs range from process control to pattern recognition and, with increasing importance, robotics. This paper demonstrates continuous control of a robot using the deep deterministic policy gradients (DDPG) algorithm, an actor-critic reinforcement learning strategy, originally conceived by Google DeepMind. After training, the robot performs controlled locomotion within an enclosed area. The paper also details the robot design process and explores the challenges of implementation in a real-time system.

Scientists have been working on making robots act like human beings for decades. Therefore, how to imitate human motion has became a popular academic topic in recent years. Nevertheless, there are infinite trajectories between two points in three-dimensional space. As a result, imitation learning, which is an algorithm of teaching from demonstrations, is utilized for learning human motion. Dynamic Movement Primitives (DMPs) is a framework for learning trajectories from demonstrations. Likewise, DMPs can also learn orientations given rotational movement's data. Also, the simulation is implemented on Robot Baxter which has seven degrees of freedom (DOF) and the Inverse Kinematic (IK) solver has been pre-programmed in the robot, which means that it is able to control a robot system as long as both translational and rotational data are provided. Taking advantage of DMPs, complex motor movements can achieve task-oriented regeneration without parametric adjustment and consideration of instability. In this work, discrete DMPs is utilized as the framework of the whole system. The sample task is to move the objects into the target area using Robot Baxter which is a robotic arm-hand system. For more effective learning, a weighted learning algorithm called Local Weighted Regression (LWR) is implemented. To achieve the goal, the weights of basis functions are firstly trained from the demonstration using DMPs framework as well as LWR. Then, regard the weights as learning parameters and substitute the weights, desired initial state, desired goal state as well as time-correlated parameters into a DMPs framework. Ultimately, the translational and rotational data for a new task-specific trajectory is generated. The visualized results are simulated and shown in Virtual Robot Experimentation Platform (VREP). For accomplishing the tasks better, independent DMP is used for each translation or rotation axis. With relatively low computational cost, motions with relatively high complexity can also be achieved. Moreover, the task-oriented movements can always be successfully stabilized even though there are some spatial scaling and transformation as well as time scaling. Twelve videos are included in supplementary materials of this thesis. The videos mainly describe the simulation results of Robot Baxter shown on Virtual Robot Experimentation Platform (VREP). The specific information can be found in the appendix.

The primary aims of this thesis, is to provide a body of knowledge on variable structure system theory and to apply the developed design concepts to control practical systems. It introduces the concept of a structure. The main aim in designing variable structure controllers, is to synthesize a variable structure system from two or more single structure systems, in such a way that the ensuing system out-performs its component structures. When a sliding mode is defined, the ensuing closed loop behaviour of the system is invariant to plant parameter changes and external disturbances. A variable structure controller was designed for a servo motor and successfully applied to the system. In practice, the phase plane representative point does not slide at infinite frequency with infinitesimal amplitude along the switching surface(s). Thus, the concept of a quasi-sliding regime was introduced. For high performance system specifications, the phase plane representative point could cycle about the origin. In some instances, sliding could be lost. For high speed applications, a novel design modification ensured that the system did not lose sliding. In addition, the controller could track a rapidly changing set point. Successful results support the developed theory.

Thesis: M.S., Massachusetts Institute of Technology, Department of Aeronautics and Astronautics, 1977 / Includes bibliographical references. / M.S. / M.S. Massachusetts Institute of Technology, Department of Aeronautics and Astronautics

This dissertation addresses the design of controllers for multi variable finite-dimensional, linear, autonomous dynamical systems with distinct sets of slow and fast dynamics, which thus display multiple time scale behavior. It seeks, specifically, to compose overall controllers from lower-order dynamic output compensators, which are designed separately for slow and fast approximating models of the given plant. Reduction of the dynamic order of the design problem and the avoidance of numerically ill-conditioned interaction between modes of disparate orders of magnitude are among the patent advantages which pertain to such a design. As is well known, the explicit singularly perturbed systems, as a class, possess the multiple time scale property, while the broader class of implicit singularly perturbed systems and the multiple time scale systems are partially overlapping system classes. A composite state feedback controller scheme for the explicit singularly perturbed system has long been known. In connection with dynamic output controllers, however, only the case of the explicit singularly perturbed system with, restrictively, open-loop-stable fast dynamics has so far received attention in the literature. The dissertation, in providing a composite dynamic controller design suitable as well to the implicit singularly perturbed multiple time scale system, which furthermore is permitted to exhibit fast (or "parasitic") as well as slow (or "normal") open-loop instabilities, thus presents a more comprehensive dynamic controller strategy for this system than so far reported in the literature. Working with multivariable transfer functions, the dissertation applies certain fractional representation techniques of modem Algebraic System Theory to the frequency domain study of the multiple time scale system. Following the work of D. W. Luse and H. K. Khalil, we replace the transfer matrix of the multiple time scale system with two or more lower-order transfer functions, each of which has validity, in its own respective frequency range, as an approximation to the first. Following the work. of M. Vidyasagar, we write the rational transfer function of each of these approximating lower-order subsystems as a "fraction" over the Ring of proper and stable rational matrices. Parametrizations, in terms of "free" matrices belonging to this Ring, of the sets of stabilizing controllers for the lower-order subsystems :md the corresponding achievable stable closed-loop behaviors then enable the relevant design syntheses to be achieved. In this development we exploit, specifically, a Theorem proved by Luse and Khalil concerning the relation of the closed-loop poles in a feedback configuration of multiple time scale systems to the poles in corresponding lower-order closed-loop systems. The dissertation's novel contribution thus resides in (i) interpreting the Theorem of Luse and Khalil as the outline for a possible separate and-composite dynamic output controller strategy, and in (ii) adapting algebraic techniques derivative from Vidyasagar for actually realizing the putative strategy as a set of concretely implementable design procedures. Three specific design procedures are developed and formalized in the dissertation: the first for achieving mere stabilization of the multiple time scale system, the second for the placement of slow and fast poles within specified subregions of the Complex Plane, and the third for achieving entirely arbitrary pole placement. Since these procedures derive, methodologically, from Vidyasagar's fractional representations, they are intrinsically multivariable in character. Since the procedures are validated by the Theorem of Luse and Khalil, they are applicable, in principle, to the very broadest class of linear autonomous multiple time scale systems. The dissertation presents its three design procedures in high-Ievel-algorized form. For application to the explicit singularly perturbed system, the three procedures entail no further matrix operations than addition, multiplication, inversion and the determination of linear constant controller and observer gains, the most basic of operations in any available Control software. In connection with the implicitly singularly perturbed multiple time scale system, their concrete application requires some further computational development pertaining to the attainment of coprime factorizations of general rational matrices, a topic of independent active interest in the current literature on Control. Elsewhere in the literature, singularly perturbed discrete time, distributed, multidimensional, time-varying and nonlinear systems have been studied. Such systems have been studied, furthermore, in several contexts involving optimal and stochastic control, but nearly always from the time domain point of view. The dissertation tackles only the problems of robust stabilization and pole placement in the finite-dimensional, linear, autonomous case. Future work will attempt to extend its results on stabilization and pole-placement, appropriately, to some of the other general multiple time scale system classes. Further frequency domain investigations, related to the dissertation, may as well explore other problems pertaining to the multiple time scale systems which so far have received treatment only in the available time domain literature.

Note: