By Topic

Date 21-23 June 1989

Filter Results

Displaying Results 1 - 25 of 528
  • Proceedings of the 1989 American Control Conference

    Page(s): 1
    Save to Project icon | PDF file iconPDF (25 KB)  
    Freely Available from IEEE
  • Copyright page

    Page(s): 2
    Save to Project icon | PDF file iconPDF (22 KB)  
    Freely Available from IEEE
  • Foreword

    Page(s): 3
    Save to Project icon | Request Permissions | PDF file iconPDF (110 KB)  
    Freely Available from IEEE
  • AACC Awards

    Page(s): 4 - 6
    Save to Project icon | PDF file iconPDF (462 KB)  
    Freely Available from IEEE
  • Operating Committee

    Page(s): 7
    Save to Project icon | PDF file iconPDF (55 KB)  
    Freely Available from IEEE
  • Program Committee

    Page(s): 7
    Save to Project icon | PDF file iconPDF (55 KB)  
    Freely Available from IEEE
  • Technical program

    Page(s): 8 - 34
    Save to Project icon | PDF file iconPDF (1654 KB)  
    Freely Available from IEEE
  • Indirect Adaptive Feed-Forward Tracking Controllers

    Page(s): 1 - 6
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (482 KB)  

    This paper presents a general indirect adaptive feed-forward control scheme for the purpose of tracking time varying signals. An approach which allows independent design of the adaptive identifier and the feed-forward tracking controller is given and so is the characterization of its tracking performance. When disturbances and unmodeled dynamics exist, methods such as signal filtering, dead zone and constrained parameter estimation are important for obtaining robust adaptive identifies, without which the standard approach could not be successful. Application to a hydraulic servo system, the dynamics of which are nonlinear, time varying and of infinite order, demonstrates the usefulness of the proposed adaptive scheme. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • On Adaptive Tuning for Linear Distributed Systems using Compensator Prior Knowledge

    Page(s): 7 - 12
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (469 KB)  

    In this paper we address the problem of adaptive controller tuning for continuous, linear time invariant (LTI) distributed systems which are stabilizable with finite dimensional LTI compensators. We use the word tuning to underscore the fact that the main task of the adaptive controller is not stabilization, but to adjust its parameters to improve performance. Our main concern is the incorporation on the adaptive controller design of frequency domain prior knowledge on LTI stabilizing parametrizations. This information is used to define a new filtered error that drives the parameter update law. First, we show that, for each stabilizing parametrization, we can always define suitable filters that yield a globally convergent adaptive algorithm. Then, for each ideal filter we derive a frequency condition that the estimator filters must satisfy to insure global convergence. No restrictions, besides linearity on the adjustable parameters, are imposed on the controller structure (or order). The results are illustrated with a simulated example of a system with unknown input delay. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Expert Hierarchical Adaptive Control

    Page(s): 13 - 18
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (524 KB)  

    The goal of this work is to develop an expert, hierarchical control scheme that adapts its structure in accordance with its knowledge about the plant. When the process is poorly defined, a coarse control action is provided by a rule-based controller. The next stage of the controller is a multiple model adaptive control procedure that is chosen if the system identifies an approximate model of the process. The most precise control is determined by a model reference adaptive controller that is used if the sufficiency conditions ensuring its stability are satisfied. The heuristics used to determine switching between different controllers are orchestrated by an expert system that bases its decisions on plant state and controller parameter values. The controller structure is presented through an example of a regulator in a drug delivery system. Details regarding the sufficiency conditions for the advancement of the controller through the different stages are given. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Adaptive Time-Optimal Control of Flexible Structures

    Page(s): 19 - 24
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (484 KB)  

    The Extended Proximate Time-Optimal Servo (XPTOS) developed in [8] is analyzed for stability using new techniques. The XPTOS system operates in closed loop, and blends in its structure the characteristics of a time-optimal control law and the fine tracking properties of a properly tuned linear regulator. This study is addressed to the (idealized) case of flexible structures that contain a single or dominant structural mode. Simulation results demonstrate the performance of the XPTOS, and delineate its range of applicability. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • A Comparison of Feedback Linearization and Singular Perturbation Techniques for the Control of Flexible Joint Robots

    Page(s): 25 - 30
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (459 KB)  

    Analytical and experimental results comparing feedback linearization and singular perturbation control techniques for flexible joint robots are presented. The feedback linearization approach results in a globally decoupled linear system which can be controlled using techniques of linear system theory. The singular perturbation approach exploits a natural time-scale separation that results from large joint stiffness to compute a composite control law to damp the joint oscillations while exploiting the desirable properties of rigid robot dynamics. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Robustness Investigation of Adaptive Control Under Reference Model Switching for a Cylindrical Robot

    Page(s): 31 - 38
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (467 KB)  

    Space robots are expected to play a significant role in future space missions. This requires a thorough investigation into the capabilities of such robots. A typical task that will need to be performed by the space robot is the capture and manipulation of large objects. This could increase the mass of the robot by as much as 100%. Additionally, the object capture task will require the robot to handle high dynamic disturbance. The control system of the robot must maintain stability under these harsh conditions. In this paper, the authors have conducted an investigation into the robustness properties of a new robot control strategy based on the Direct Model Reference Adaptive Control (MRAC) Algorithm under switching of the reference model. Experimental results of the investigation are presented for a three degree-of-freedom cylindrical robot. The new control scheme is demonstrated to be a powerful approach to the difficult problems that will be encountered in space robotics control. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Robust Pole Assignment for Computed Torque Robotic Manipulators Control

    Page(s): 37 - 41
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (287 KB)  

    Robotic manipulator control designers often assume accurate knowledge of the system parameters. In practice, such as assumption will often lead to unacceptable deterioration in controller performance and robust approaches for the design of manipulator controllers are required. In this paper, the computed torque method is used to reduce the manipulator controller design to a linear problem. A robust pole assignment approach is used to select a suitable linear state feedback for the nominal computed torque model. The effect of modeling errors is accounted for by a state-dependent acceleration disturbance vector. Stability bounds for this acceleration vector are obtained using the Lyapunov approach and robustness measures to assess its effect are suggested. A 2-D.O.F. manipulator design example is given to demonstrate the design approach. Simulation results suggest that state feedback that recouples the state variables decoupled in the computed torque approach may offer some advantages over decoupled designs. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Design of Nonlinear Adaptive Tracking Controllers for Industrial Robots

    Page(s): 42 - 47
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (337 KB)  

    This paper proposes a controller for a class of industrial robots which fully handles system nonlinearity and adaptively estimates plant parameters. Using a nonlinear input transformation, we obtain a system which is linear in the transformed inputs. The resulting system is then controlled with a continious-time PD tracking controller. The critical part of this scheme is estimating the parameters of the nonlinear input transformation. We estimate them using a least-squares with a forgeting factor based on aperiodic samples of inputs and outputs. Simulation results are presented for a nonlinear three-degree-of-freedom manipulator. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Elastically Coupled Precision Pointing by Slew-Induced Deformation Shaping

    Page(s): 48 - 53
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (461 KB)  

    Computed torques for pointing and tracking require compensation for slew-induced structural, forebody/aftbody, or optical train alignment deformations. Thus even if only line-of-sight variables are to be commanded yet full state feedback is needed, with consequent high bandwidth control requirements. The solution investigated here is to decouple the unwanted deformation state by feedforward of the line-of-sight slew dynamics into the deformation control forces or moments, for an apparatus consisting of a mirror mounted on an optical bench, that is itself mounted on a rotating table. Adjustable elastic interfaces are used to model slew-induced deformations as angular differences between mirror mounting, optical bench and rate table. Low bandwidth control with fewer actuators than degrees of freedom is shown to be possible, by correcting the computed torques so as to force the interstage angular differences (standing for slew-induced deformations) to evolve in a "slow" integral manifold wherein they are modeled as functions of the mirror pointing angle. Simultaneous mirror pointing and independent rate table pointing is also shown to be possible, which represents the situation of pointing an instrument elastically mounted on a maneuvering platform. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • A Multivariable Self-Tuning Controller for Injection Molding Machines

    Page(s): 54 - 61
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (796 KB)  

    The stochastic dynamic characteristics of an injection molding machine are discussed and the mainly controlled parameters of the process are determined. A multivariable self-tuning controller with explicit method for this dynamic system is then proposed. The effectiveness of the controller is proved by the corresponding computer simulation. On the basis of on-line adjusting function for the weighting matrices of the cost functions found in the simulation, an artificially intelligent self-learning level is recommended as an updated addition to the controller for supervision and coordination. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Neural Networks in GTA Weld Modeling and Control

    Page(s): 62 - 67
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (476 KB)  

    Solutions to modeling the Gas Tungsten Arc(GTA) Welding process using a non-conventional technique is presented here. This approach is a non-linear modeling technique employing neural networks which has exhibited the potential to learn to model the time response of a non-linear, multivariable system. This paper examines the feasibility of this approach an alternative to existing techniques Potential problems with this approach are also discussed. A control architecture using a second neural network is also suggested. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Estimation of Time Varying Parameters in Discrete Time Dynamic Systems: A Tool Wear Estimation Example

    Page(s): 68 - 74
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (557 KB)  

    The recursive least squares, Kalman filter, and basis function methods for the estimation of time varying parameters are described and compared for a particular example problem. A generalization of these methods for estimation of time varying parameters is presented, based on an adaptive Kalman filter algorithm. The adaptive Kalman filter (or adaptive observer) utilizes a state model with unknown coefficients, of the time varying parameters. All the other estimation methods presented for time varying parameters can be obtained as special cases of the proposed method. The method proposed shows excellent performance on the simple example problem considered, but can be difficult to apply. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Stability of the Optimal Locus System

    Page(s): 75 - 80
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (426 KB)  

    The optimal locus is a new real-time control method for processes subject to constraints that vary as the process progresses. This paper analyze of the stability conditions of an optimal-locus based system. A general stability criterion is found for optimizing a quadratic cost function of a process subject to linear varying constraints. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Feedrate Optimization for Machine Tool Control Subject to Contour Error Constraints

    Page(s): 81 - 86
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (394 KB)  

    Optimization techniques are applied to feedrate and voltage control in order to improve performance of biaxial contouring systems. This is done by maintaining contour error within a prescribed tolerance while tracking the trajectory in minimum time. In addition to the error constraint, the current and voltage constraints are required to keep the motors from overloading. The design of the controllers requires two steps. First, an LQR-type innerloop controller for the armature voltage is designed for each axis to assure stability using a performance index that penalizes current, armature voltage, and axial error. Next, nonlinear programming techniques are used to find the outerloop control or feedrate profile (which determines the tracking time) and the weighting matrices (which determine the innerloop controller gains) subject to voltage, current, and two-dimensional contour error constraints for a given trajectory. Although the innerloop controllers are originally designed to minimize a weighted sum of axial error, current, and voltage, the outerloop optimization of the feedrate takes into account the two-dimensional contour error explicitly. In this manner, it is possible to tune the controllers to a particular trajectory to achieve the best performance. The method is tested via simulation using a model of a representative biaxial DC motor contouring system for an elliptical trajectory. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Trajectory Planning for High Speed Multiple Axis Contouring Systems

    Page(s): 87 - 94
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (624 KB)  

    In the high speed coordinated motion of multiple axis linear systems two objectives must be achieved: 1) specified position accuracy and 2) well behaved actuator torque output. Objective (1) can be achieved through the design of an appropriate feedforward controller and (2) is achieved both through feedforward controller design and trajectory planning. Since actuator saturation will result in loss of position accuracy, failure to achieve (2) will also result in the violation of (1). This paper considers the generation of minimum time tracking velocity profiles for multiple axis contouring systems along prespecified spatial curves while subject to hard actuator torque limits. The result is applicable to systems with a general number of axes which may be dynamically coupled. It is also demonstrated that the trajectory generation and controller design problems are interdependent, and a method of selecting a trajectory generation algorithm for a given system under discrete time feedforward/feedback control is described. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Dicode - An Expert System for Designing Distillation Column Controls

    Page(s): 95 - 100
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (339 KB)  

    An expert system for designing control systems for distillation columns, called DICODE (DItillation COlumn Design Expert), has been developed. This expert system was the result of a joint project between researchers at the University of Maryland and E.I. Dupont & Co. The expert system contains approximately 300 rules and it has gone through extensive prototyping. This paper focuses on the distillation control aspects of the expert system. Initially in developing DICODE the design methodology of Page Buckley was used and then that of Greg Shinskey was added. Examples are presented that show that the two design methodologies produce results that are diametrically opposed to one another in a number of cases. An attempt is made to explain how such differences can arise between the results produced by known experts. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • A Case-Study of Multivariable Control for a Multicomponent Distillation Unit on a Pilot Plant Scale

    Page(s): 101 - 106
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (499 KB)  

    Robust multivariable control strategies based on H¿-minimization are developed for a pilot plant distillation unit consisting of two coupled distillation columns for the separation of the mixture Methanol/Ethanol/n-Propanol. A comparison is made between several control strategies in nonlinear simulation and experiment. Controller design is based on transfer function matrices describing column dynamics in the vicinty of the operating point. The controllers are designed to keep product specifications by plate temperature control. The problem of choosing appropriate locations for temperature measurements is also addressed. Constraints on the manipulated variables have to be considered in the design. Qualitative physical intuition is used for selecting appropriate weighing functions. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • The Design of Squaring Compensators for the Feedback Control of Nonsquare Processes

    Page(s): 107 - 113
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (663 KB)  

    An improved method to design control systems for nonsquare processes, processes in which the number of inputs and outputs are not equal, is presented. The control system utilizes squaring compensators that, when combined with the nonsquare process, form a square system that is decentrally controlled. The compensators are designed to reduce row and column interaction of the square system over a particular frequency range. The parameters of the squaring compensators can be determined analytically or by optimization. The design procedure is applied to the control of two distillation columns. The control systems are analyzed using modern analysis tools and tested by computer simulation. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.