Engineering topics

Hyland, David C.

Publications and source records attributed to Hyland, David C..

12 records

NASA NTRS1995

Multiprocessor Adaptive Control Of A Dynamic System

Architecture for fully autonomous digital electronic control system developed for use in identification and adaptive control of dynamic system. Architecture modular and hierarchical. Combines relatively simple, standardized processing units into complex parallel-processing subsystems. Although architecture based on neural-network concept, processing units themselves not neural networks; processing units implemented by programming of currently available microprocessors.

Juang, Jer-Nan↗

NASA NTRS1994

Streamlined design and self reliant hardware for active control of precision space structures

Precision space structures may require active vibration control to satisfy critical performance requirements relating to line-of-sight pointing accuracy and the maintenance of precise, internal alignments. In order for vibration control concepts to become operational, it is necessary that their benefits be practically demonstrated in large scale ground-based experiments. A unique opportunity to carry out such demonstrations on a wide variety of experimental testbeds was provided by the NASA Control-Structure Integration (CSI) Guest Investigator (GI) Program. This report surveys the experimental results achieved by the Harris Corporation GI team on both Phases 1 and 2 of the program and provides a detailed description of Phase 2 activities. The Phase 1 results illustrated the effectiveness of active vibration control for space structures and demonstrated a systematic methodology for control design, implementation test. In Phase 2, this methodology was significantly streamlined to yield an on-site, single session design/test capability. Moreover, the Phase 2 research on adaptive neural control techniques made significant progress toward fully automated, self-reliant space structure control systems. As a further thrust toward productized, self-contained vibration control systems, the Harris Phase II activity concluded with experimental demonstration of new vibration isolation hardware suitable for a wide range of space-flight and ground-based commercial applications.The CSI GI Program Phase 1 activity was conducted under contract NASA1-18872, and the Phase 2 activity was conducted under NASA1-19372.

Hyland, David C.↗

NASA NTRS1993

On neural networks in identification and control of dynamic systems

This paper presents a discussion of the applicability of neural networks in the identification and control of dynamic systems. Emphasis is placed on the understanding of how the neural networks handle linear systems and how the new approach is related to conventional system identification and control methods. Extensions of the approach to nonlinear systems are then made. The paper explains the fundamental concepts of neural networks in their simplest terms. Among the topics discussed are feed forward and recurrent networks in relation to the standard state-space and observer models, linear and nonlinear auto-regressive models, linear, predictors, one-step ahead control, and model reference adaptive control for linear and nonlinear systems. Numerical examples are presented to illustrate the application of these important concepts.

Phan, Minh↗

NASA NTRS1992

High performance, accelerometer-based control of the Mini-MAST structure

Many large space system concepts will require active vibration control to satisfy critical performance requirements such as line of sight pointing accuracy and constraints on rms surface roughness. In order for these concepts to become operational, it is imperative that the benefits of active vibration control be shown to be practical in ground based experiments. The results of an experiment shows the successful application of the Maximum Entropy/Optical Projection control design methodology to active vibration control for a flexible structure. The testbed is the Mini-Mast structure at NASA-Langley and has features dynamically traceable to future space systems. To maximize traceability to real flight systems, the controllers were designed and implemented using sensors (four accelerometers and one rate gyro) that are actually mounted to the structure. Ground mounted displacement sensors that could greatly ease the control design task were available but were used only for performance evaluation. The use of the accelerometers increased the potential of destabilizing the system due to spillover effects and motivated the use of precompensation strategy to achieve sufficient compensator roll-off.

Collins, Emmanuel G., Jr.↗

NASA NTRS1991

High performance, accelerometer-based control of the Mini-MAST structure at Langley Research Center

Many large space system concepts will require active vibration control to satisfy critical performance requirements such as line of sight pointing accuracy and constraints on rms surface roughness. In order for these concepts to become operational, it is imperative that the benefits of active vibration control be shown to be practical in ground based experiments. The results of an experiment shows the successful application of the Maximum Entropy/Optimal Projection control design methodology to active vibration control for a flexible structure. The testbed is the Mini-Mast structure at NASA-Langley and has features dynamically traceable to future space systems. To maximize traceability to real flight systems, the controllers were designed and implemented using sensors (four accelerometers and one rate gyro) that are actually mounted to the structure. Ground mounted displacement sensors that could greatly ease the control design task were available but were used only for performance evaluation. The use of the accelerometers increased the potential of destabilizing the system due to spillover effects and motivated the use of precompensation strategy to achieve sufficient compensator roll-off.

Collins, Emmanuel G., Jr.↗

NASA NTRS1991

Robust decentralized control laws for the ACES structure

Control system design for the Active Control Technique Evaluation for Spacecraft (ACES) structure at NASA Marshall Space Flight Center is discussed. The primary objective of this experiment is to design controllers that provide substantial reduction of the line-of-sight pointing errors. Satisfaction of this objective requires the controllers to attenuate beam vibration significantly. The primary method chosen for control design is the optimal projection approach for uncertain systems (OPUS). The OPUS design process allows the simultaneous tradeoff of five fundamental issues in control design: actuator sizing, sensor accuracy, controller order, robustness, and system performance. A brief description of the basic ACES configuration is given. The development of the models used for control design and control design for eight system loops that were selected by analysis of test data collected from the structure are discussed. Experimental results showing that very significant performance improvement is achieved when all eight feedback loops are closed are presented.

Collins, Emmanuel G., Jr.↗

NASA NTRS1991

Active and passive vibration suppression for space structures

The relative benefits of passive and active vibration suppression for large space structures (LSS) are discussed. The intent is to sketch the true ranges of applicability of these approaches using previously published technical results. It was found that the distinction between active and passive vibration suppression approaches is not as sharp as might be thought at first. The relative simplicity, reliability, and cost effectiveness touted for passive measures are vitiated by 'hidden costs' bound up with detailed engineering implementation issues and inherent performance limitations. At the same time, reliability and robustness issues are often cited against active control. It is argued that a continuum of vibration suppression measures offering mutually supporting capabilities is needed. The challenge is to properly orchestrate a spectrum of methods to reap the synergistic benefits of combined advanced materials, passive damping, and active control.

Hyland, David C.↗

NASA NTRS1991

Demonstrations of LSS active vibration control technology on representative ground-based testbeds

This paper describes two experiments which successfully demonstrate control of flexible structures. The first experiment involved control design and implementation for the ACES structure at NASA Marshall Space Flight Center, while the second experiment was conducted using the Multi-Hex Prototype structure. The paper concludes with some remarks on the lessons learned from conducting these experiments.

Hyland, David C.↗

NASA NTRS1990

Design and implementation of robust decentralized control laws for the ACES structure at Marshall Space Flight Center

Many large space system concepts will require active vibration control to satisfy critical performance requirements such as line-of-sight accuracy. In order for these concepts to become operational it is imperative that the benefits of active vibration control be practically demonstrated in ground based experiments. The results of the experiment successfully demonstrate active vibration control for a flexible structure. The testbed is the Active Control Technique Evaluation for Spacecraft (ACES) structure at NASA Marshall Space Flight Center. The ACES structure is dynamically traceable to future space systems and especially allows the study of line-of-sight control issues.

Collins, Emmanuel G., Jr.↗

NASA NTRS1990

Design and implementation of robust decentralized control laws for the ACES structure at Marshall Space Flight Center

An experiment was conducted to design controllers that would provide substantial reduction of line-of-sight control errors. The satisfaction of this objective required the controllers to attenuate the beam vibration significantly. Particular emphasis was placed on controller simplicity (i.e., reduced-order and decentralized controller architectures). Complexity reduction in control law implementation is of paramount interest due to stringent limitations on throughput of even state-of-the-art space qualified processors. The results of this experiment successfully demonstrate active vibrator control for a flexible structure. The testbed is the ACES structure at the NASA Marshall Space Flight Center. The ACES structure is dynamically traceable to future space systems and especially allows the study of line-of-sight control issues.

Collins, Emmanuel G., Jr.↗

NASA NTRS1990

Experimental demonstration of active vibration control for flexible structures

Active vibration control of flexible structures for future space missions is addressed. Three experiments that successfully demonstrate control of flexible structures are described. The first is the pendulum experiment. The structure is a 5-m compound pendulum and was designed as an end-to-end test bed for a linear proof mass actuator and its supporting electronics. Experimental results are shown for a maximum-entropy/optimal-projection controller designed to achieve 5 percent damping in the first two pendulum modes. The second experiment was based upon the Harris Multi-Hex prototype experiment (MHPE) apparatus. This is a large optical reflector structure comprising a seven-panel array and supporting truss which typifies a number of generic characteristics of large space systems. The third experiment involved control design and implementation for the ACES structure at NASA Marshall Space Flight Center. The authors conclude with some remarks on the lessons learned from conducting these experiments.

Phillips, Douglas J.↗

NASA NTRS1989

Integrated control-system design via generalized LQG (GLQG) theory

Thirty years of control systems research has produced an enormous body of theoretical results in feedback synthesis. Yet such results see relatively little practical application, and there remains an unsettling gap between classical single-loop techniques (Nyquist, Bode, root locus, pole placement) and modern multivariable approaches (LQG and H infinity theory). Large scale, complex systems, such as high performance aircraft and flexible space structures, now demand efficient, reliable design of multivariable feedback controllers which optimally tradeoff performance against modeling accuracy, bandwidth, sensor noise, actuator power, and control law complexity. A methodology is described which encompasses numerous practical design constraints within a single unified formulation. The approach, which is based upon coupled systems or modified Riccati and Lyapunov equations, encompasses time-domain linear-quadratic-Gaussian theory and frequency-domain H theory, as well as classical objectives such as gain and phase margin via the Nyquist circle criterion. In addition, this approach encompasses the optimal projection approach to reduced-order controller design. The current status of the overall theory will be reviewed including both continuous-time and discrete-time (sampled-data) formulations.

Bernstein, Dennis S.↗