Showing papers presented at "American Control Conference in 1984"

Impedance Control: An Approach to Manipulation

Abstract: Manipulation fundamentally requires a manipulator to be mechanically coupled to the object being manipulated. A consideration of the physical constraints imposed by dynamic interaction shows that control of a vector quantity such as position or force is inadequate and that control of the manipulator impedance is also necessary. Techniques for control of manipulator behaviour are presented which result in a unified approach to kinematically constrained motion, dynamic interaction, target acquisition and obstacle avoidance.

Distributed Asynchronous Deterministic and Stochastic Gradient Optimization Algorithms

Abstract: We present a model for asynchronous distributed computation and then proceed to analyze the convergence of natural asynchronous distributed versions of a large class of deterministic and stochastic gradient-like algorithms. We show that such algorithms retain the desirable convergence properties of their centralized counterparts, provided that the time between consecutive communications between processors plus communication delays are not too large.

Robust Adaptive Control

Abstract: The paper addresses an open problem concerned with the boundedness of signals in an adaptive loop when external perturbations are present. A complete solution is provided for the case of a first order plant with an unknown parameter by analyzing a nonlinear differential equation in R2. It is shown that if the persistent excitation of the reference input is larger than the perturbation in some sense, the solutions will be globally bounded. The same methodology appears to be applicable to the general adaptive control problem.

Enhancement of Robot Accuracy using Endpoint Feedback and a Macro-Micro Manipulator System

Abstract: The problem of robot positioning accuracy is caused primarily by unmeasured deflections of the structure and poor actuator servo resolution. This is particulary troublesome when large manipulators and payloads are considered. This paper presents an approach to this problem based on the use of a direct enpoint position and orientation measurement. While this insures accurate measurement of endpoint errors, it is still difficult for most manipulators to accurately make fine movements. To address this problem the concept of a micromanipulator mounted on the end of the large manipulator is forwarded. A full scale prototype micromanipulator was fabricated and tested to verify the feasibility of such a device. To make the control of such a device tractable it is proposed to design the control assuming no dynamic coupling between the macro and micro manipulator. This approach was invetigated by simulation and experimentation and it is found to work well provided the bandwidth of the micromanipulator is kept below the fundamental frequency of vibration of the basic robot structure.

How Autopilot Requirements Constrain the Aerodynamic Design of Homing Missiles

Abstract: When an aerodynamically controlled missile is used in a homing application, the transfer function of the vehicle becomes part of an overall homing and attitude control feedback loop. Therefore, the missile must be designed so that its aerodynamics meet the constraints required to accomplish homing successfully. For radar homing, these constraints are stringent enough to require an autopilot that controls the aerodynamic transfer function using body instruments and internal feedback loops. Even with an autopilot, there are limits on the aerodynamic moment parameters such as M ? and M ? that must be considered at the very beginning of the missile system design. This paper explores the nature and sources of these limits and presents numerical examples to illustrate what kinds of numbers are required.

A Feedback Theory of Two-Degree-of-Freedom Optimal Wiener-Hopf Design

Abstract: The design of linear two-degree-of-freedom stabilizing controllers in a quadratic-cost setting is treated The class of all such controllers which give finite cost is established and the trade-off possible between optimum performance, tracking-cost sensitivity, and stability margins is discussed

Simplified Grasping and Manipulation with Dextrous Robot Hands

Abstract: A method is presented for stably grasping 2 dimensional polygonal objects with a dextrous hand when object models are not available. Basic constraints on object vertex angles are found for feasible grasping with two fingers. Local tactile information can be used to determine the finger motion that will reach feasible grasping locations. With an appropriate choice of finger stiffnesses, a hand can automatically grasp these objects with two fingers. The bounded slip of a part in a hand is shown to be valuable for adapting the fingers and object to a stable situation. Examples are given to show the ability of this grasping method to accommodate disturbance forces and to perform simple part reorientations and regrasping operations.

Concepts and Algorithms for Terminal-Area Traffic Management

Abstract: The nation's air-traffic-control system is the subject of an extensive modernization program, including the planned introduction of advanced automation techniques. This paper gives an overview of a concept for automating terminal-area traffic management. Four-dimensional (4D) guidance techniques, which play an essential role in the automated system, are reviewed. One technique, intended for on-board computer implementation, is based on application of optimal control theory. The second technique is a simplified approach to 4D guidance intended for ground computer implementation. It generates advisory messages to help the controller maintain scheduled landing times of aircraft not equipped with on-board 4D guidance systems. An operational system for the second technique, recently evaluated in a simulation, is also described.

Applications of a Nonlinear Controller Design Approach based on Quasilinear System Models

Abstract: In this paper, we report on recent progress in developing nonlinear control system design techniques based on sinusoidal-input describing function (SIDF) methods. Primarily, this involves illustrating a fundamental difference between SIDF and random-input describing function (RIDF) models of nonlinear systems, developing the nonlinear controller design method more fully, and demonstrating it by applying it to a significant nonlinear control design problem in robotics. Based on these results, the use of this nonlinear controller design method should be substantially better understood.

Dynamic Simulation of a Flexible Robot Arm and Controller

Abstract: High performance requirements in robotics have led to the consideration of structural flexibility in robot arms. This paper employs an assumed modes method to model the flexible motion of the last link of a spherical coordinate robot arm. This model is used to investigate relationships between the arm structural flexibility and a linear controller for the rigid body motion. This simple controller is used to simulate the controllers currently used in industrial robots. The simulation results illustrate the differences between leadscrew driven and unconstrained axes of the robot; they indicate the trade-off between speed and accuracy; and show potential instability mechanisms due to the interaction between the controller and the robot structural flexibility.

Robustness of Luenberger Observers: Linear Systems Stabilized via Nonlinear Control

Abstract: Given a dynamical system whose state equations include time-varying uncertain parameters, it is often desirable to design a state feedback controller leading to uniform asymptotic stability of a given equilibrium point If, however, the controller operates on some estimate of the state, instead of the true state itself, it is of interest to know whether the desired stability will be preserved; eg, suppose that the measured output is processed by a Luenberger observer This paper concentrates on the scenario above and in addition, our analysis permits the controller to be nonlinear As a first step, inequalities are developed which have implications on the system's robustness; that is, when the uncertain parameters satisfy these inequalities, it becomes possible to separately design controller and observer This amounts to an extension of the classical separation theorem to the case when the controller is nonlinear It is also of interest to note that the approach given here enables us to guarantee stability for some nonzero range of admissible parameter variations This is achieved by introducing a certain "tuning parameter" into the Lyapunov function which is used to assure the stability of the cambined plant-observer-controller system

A Fault Detection and Isolation Methodology Theory and Application

Abstract: The paper presents the theoretical development and practical applications of a fault detection and isolation (PDI) methodology which exploits all available sources of redundant information. By concurrent checking of relative consistencies among all redundant measurements, the most consistent and inconsistent subsets are identified for measurement estimation and fault isolation, respectively. The FDI algorithm is particularly suitable for real-time applications using commercially available microcomputers and has been tested on-line in operating nuclear reactors.

Exact spacecraft detumbling and reorientation maneuvers with gimbaled thrusters and reaction wheels

Abstract: The equations of rotational motion for a spacecraft equipped with external jets and internal reaction wheels are shown to be feedback-equivalent to those of a linear system in attitude parameter space. Reorientation maneuvers are thereby formulated as linear optimal control problems with least mean square acceleration in attitude parameter space, solved in closed form and implementable either with internal or external torque commands, the choice depending on power and throttling requirements. For prior detumbling, an alternative solution with least mean square torque by angular momentum feedback is also given, that is implementable with gimbaled pairs of thrusters at constant throttle. Such a detumbling maneuver may then be followed by an acceleration-commanded rest-to-rest maneuver by means of the reaction wheels.

Optimal Control of a Robot with Obstacles

Abstract: The Maximum principle is used to design an optimal control for a robot based on minimizing the time for a specified motion. With the use of an optimization technique, the optimal control problem is solved numerically using the complete, highly non-linear dynamic model of the manipulator. Robot geometrical and actuator limitations, as well as obstacles inside the work space, are considered in the optimal control solution.

Life with Pattern Adaptation

Abstract: This paper presents a human case history of the evolution-to-practice of a distinctive approach to adaptive control against a background of other methods dating back to 1950 This concept, based on pattern recognition, provides direct performance feedback in a natural and intuitive form and guarantees adaptive convergence to satisfactory loop behavior Recent field experience has shown a new design following this concept to be extremely successful as a trouble-free, general-purpose adaptive controller Despite the apparent advantages of this concept, it was as slow to come to commercialization as were competing concepts and for many of the same reasons, notably a lack of agreement on realistic benefits of adaptation

Adaptive Detection Threshold Optimization for Tracking in Clutter

Abstract: Adaptive detection threshold optimization for improved downstream tracker performance with a Probabilistic Data Association Filter (PDAF) is investigated. Prior and posterior optimization algorithms which minimize the mean-square estimation error over the signal processor's detection thresholds and which depend on observations up to the previous and current iteration, respectively, are given. These algorithms are suitable for real-time implementation. Simulation results are presented.

Multirate and composite control of two-time-scale discrete-time systems

Abstract: The design of stabilizing feedback control of singularly perturbed diserete-time systems is decomposed into the design of slow and fast controllers which are combined to form the composite control. Composite control strategies are developed for the case of single rate measurements (all variables are measured at the same rate) as well as for the case of multirate measurements (slow variables are measured at a rate slower than that of fast variables).

A Novel Opto-Fluidic Interface

Abstract: A novel opto-fluidic interface which responds to DC signals was constructed and tested. The interface was a laminar proportional amplifier (LPA) modified to include an optical coupling region in the power nozzle. The opto-fluidic gain of the interface was measured to be 15.50 kPa per watt of input optical power. A pneumatic cylinder actuator was optically activated by the opto-fluidic interface connected to a fluidic gain block. In this configuration, an output pressure of 55.2 kPa was produced using an input optical pover of 10 mW. The bandwidth of the opto-fluidic interface was 140 Hz.

Robust Fault Detection: The Effect of Model Error

Abstract: This paper examines the issues involved in robust fault detection in the presence of system model error. The underlying theory presented is analogous to concepts in robust control design, e.g., [1]-[5], which require the introduction of a metric or norm measure of signals defined on an appropriate function space.

Implications of Internal Model Control for PID Controllers

Abstract: For a large number of single input-single output (SISO) typically used in the process industries the Internal Model Control (IMC) design procedure is shown to lead to PID controllers, occasionally augmented with a first order lag. These PID controllers have as their only tuning parameter the closed loop time constant or equivalently, the closed loop bandwidth. On-line adjustments are therefore much simpler than for general PID controllers. As a special case, PID tuning rules for systems modeled by a first order lag with deadtime are derived analytically. The superiority of these rules in terms of both closed loop performance and robustness is demonstrated.