SEARCH · Engineering Papers
Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.
Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.
Learn-To-Fly merges real-time modeling, learning adaptive control, and real-time guidance to achieve: (1) autonomous self-learning aircraft to disrupt our current vehicle development paradigm, (2) novel configuration development with no ground testing, and (3) robust autonomous flight of novel aircraft.
Explore the source record for details and available documents.
Explore the source record for details and available documents.
This work assesses an adaptive approach to fault recovery in autonomous robotic space operations, which uses indicators of opportunity, such as physiological state measurements and observations of past human assistant performance, to inform future selections. We validated our reinforcement learning approach using data we collected from humans executing simulated mission scenarios. We present a method of structuring human-factors experiments that permits collection of relevant indicator of opportunity and assigned assistance task performance data, as well as evaluation of our adaptive approach, without requiring large numbers of test subjects. Application of our reinforcement learning algorithm to our experimental data shows that our adaptive assistant selection approach can achieve lower cumulative regret compared to existing non-adaptive baseline approaches when using real human data. Our work has applications beyond space robotics to any application where autonomy failures may occur that require external intervention.
This paper presents a novel, safe tracking control design method that learns the parameters of an uncertain Euler-Lagrange (EL) system online using adaptive learning laws. A barrier function (BF) is first used to transform the full-state constrained EL-dynamics into an equivalent unconstrained dynamics. An adaptive tracking controller is then developed along with the parameter update law in the transformed state space such that the states remain bounded for all time within a prescribed bound. A stability analysis is developed that considers the EL-dynamics’ uncertainty, yielding a semi-globally uniformly ultimately bounded (SGUUB) tracking error and the parameter estimation error. The controller design is validated in simulations using a two-link planar manipulator. The results show the proposed method’s ability to track the reference trajectory while remaining inside each of the predefined state bounds.
The introduction of high-level automated systems in the aircraft cockpit has provided several benefits, e.g., new capabilities, enhanced operational efficiency, and reduced crew workload. At the same time, conventional 'static' automation has sometimes degraded human operator monitoring performance, increased workload, and reduced situation awareness. Adaptive automation represents an alternative to static automation. In this approach, task allocation between human operators and computer systems is flexible and context-dependent rather than static. Adaptive automation, or adaptive task allocation, is thought to provide for regulation of operator workload and performance, while preserving the benefits of static automation. In previous research we have reported beneficial effects of adaptive automation on the performance of both pilots and non-pilots of flight-related tasks. For adaptive systems to be viable, however, such benefits need to be examined jointly in the context of a single set of tasks. The studies carried out under this project evaluated a systematic method for combining different forms of adaptive automation. A model for effective combination of different forms of adaptive automation, based on matching adaptation to operator workload was proposed and tested. The model was evaluated in studies using IFR-rated pilots flying a general-aviation simulator. Performance, subjective, and physiological (heart rate variability, eye scan-paths) measures of workload were recorded. The studies compared workload-based adaptation to to non-adaptive control conditions and found evidence for systematic benefits of adaptive automation. The research provides an empirical basis for evaluating the effectiveness of adaptive automation in the cockpit. The results contribute to the development of design principles and guidelines for the implementation of adaptive automation in the cockpit, particularly in general aviation, and in other human-machine systems. Project goals were met or exceeded. The results of the research extended knowledge of automation-related performance decrements in pilots and demonstrated the positive effects of adaptive task allocation. In addition, several practical implications for cockpit automation design were drawn from the research conducted. A total of 12 articles deriving from the project were published.
The first results of an axisymmetric magnetohydrodynamic (MHD) model of the interaction of an expanding cometary atmosphere with the solar wind are presented. The model assumes that far upstream the plasma flow lines are parallel to the magnetic field vector. The effects of mass loading and ion-neutral friction are taken into account by the governing equations, whcih are solved on an adaptively refined unstructured grid using a Monotone Upstream Centered Schemes for Conservative Laws (MUSCL)-type numerical technique. The combination of the adaptive refinement with the MUSCL-scheme allows the entire cometary atmosphere to be modeled, while still resolving both the shock and the near nucleus of the comet. The main findingsare the following: (1) A shock is formed approximately = 0.45 Mkm upstream of the comet (its location is controlled by the sonic and Alfvenic Mach numbers of the ambient solar wind flow and by the cometary mass addition rate). (2) A contact surface is formed approximately = 5,600 km upstream of the nucleus separating an outward expanding cometary ionosphere from the nearly stagnating solar wind flow. The location of the contact surface is controlled by the upstream flow conditions, the mass loading rate and the ion-neutral drag. The contact surface is also the boundary of the diamagnetic cavity. (3) A closed inner shock terminates the supersonic expansion of the cometary ionosphere. This inner shock is closer to the nucleus on dayside than on the nightside.
An initial wind tunnel test was made to validate a new wall adaptation method for 3-D models in test sections with two adaptive walls. First part of the adaptation strategy is an on-line assessment of wall interference at the model position. The wall induced blockage was very small at all test conditions. Lift interference occurred at higher angles of attack with the walls set aerodynamically straight. The adaptation of the top and bottom tunnel walls is aimed at achieving a correctable flow condition. The blockage was virtually zero throughout the wing planform after the wall adjustment. The lift curve measured with the walls adapted agreed very well with interference free data for Mach 0.7, regardless of the vertical position of the wing in the test section. The 2-D wall adaptation can significantly improve the correctability of 3-D model data. Nevertheless, residual spanwise variations of wall interference are inevitable.
The physiological adaptation to weightless exposure includes cardiovascular deconditioning arising in part from a loss of total circulating blood volume and resulting in a reduction of orthostatic tolerance. The crew of the Shuttle orbiter are less tolerant to acceleration forces in the head-to-foot direction during the reentry phase of the flight at a time they must function at a high level of performance. The factors that contribute to orthostatic intolerance during and following reentry and to predict the likelihood of impaired crew performance are evaluated. A computer simulation approach employing a mathematical model of the cardiovascular system is employed. It is shown that depending on the severity of blood volume loss, the reentry acceleration stress may be detrimental to physiologic function and may place the physiologic status of the crew near the borderline of some type of impairment. They are in agreement with conclusions from early ground-based experiments and from observations of early Shuttle flights.
The functional model of a two dimensional adaptive truss structure which can purposefully change its geometrical configuration is introduced. The details of design and fabrication such as kinematic analysis, dynamic characteristics analysis and some test results are presented for the demonstration of this two dimensional truss concept.
Adaptive wall research at the University of Southampton has been directed towards the development of testing techniques for use in nonporous test sections where two flexible walls are profiled in single curvature. This paper highlights the recent advances that have been made in the testing of 2D airfoils through the speed of sound and the testing of 3D models at high subsonic speeds. Techniques have been developed to accommodate the variety of flow regimes encountered in near sonic airfoil tests. The experimental evidence to date suggests that the new techniques coupled with established procedures allow airfoil data, free from top and bottom wall interference, to be gathered from adaptive flexible walled test sections throughout the entire subsonic, transonic and supersonic speed ranges. Techniques applicable to the testing of 3D models have evolved primarily from experience gained by testing sidewall mounted half-wings. Emphasis has been placed upon models with planforms similar to those of current transport wings. Techniques for high subsonic speeds have now been developed to the point where the residual levels of interference are low.
Estimating aircraft models from test data requires several simplifying assumptions that introduce biases into the parameter estimates. In this paper, these biases are defined and a method for estimating the biases is discussed. Having an estimate of the bias allows the parameters estimated from test to be integrated into a high-order model. The Integrated Adaptive Wing Technology Maturation (IAWTM) wind tunnel model is discussed and the bias is demonstrated for one of the testing configurations. The methodology was able to estimate these biases and apply corrections to high-order aeroelastic models to improve the fit to test data. The consideration of the biases allows more meaningful comparisons and avoids the erroneous differences between pretest predictions and the fitted model.
Estimating aircraft models from test data requires several simplifying assumptions that introduce biases into the parameter estimates. In this paper, these biases are defined and a method for estimating the biases is discussed. Having an estimate of the bias allows the parameters estimated from test to be integrated into a high-order model. The Integrated Adaptive Wing Technology Maturation (IAWTM) wind tunnel model is discussed and the bias is demonstrated for one of the testing configurations. The methodology was able to estimate these biases and apply corrections to high-order aeroelastic models to improve the fit to test data. The consideration of the biases allows more meaningful comparisons and avoids the erroneous differences between pretest predictions and the fitted model.
The Grodins' respiratory control model was investigated and it was determined that the following modifications were necessary before the model would be adaptable for current research efforts: (1) the controller equation must be modified to allow for integration of the respiratory system model with other physiological systems; (2) the system must be more closely correlated to the salient physiological functionings; (3) the respiratory frequency and the heart rate should be expanded to illustrate other physiological relationships and dependencies; and (4) the model should be adapted to particular individuals through a better defined set of initial parameter values in addition to relating these parameter values to the desired environmental conditions. Several of Milhorn's respiratory control models were also investigated in hopes of using some of their features as modifications for Grodins' model.
Assembly and operation of large space structures (LSS) in orbit will require robot-assisted docking and berthing of partially-assembled structures. These operations require new solutions to the problems of controls. This is true because of large transient and persistent disturbances, controller-structure interaction with unmodeled modes, poorly known structure parameters, slow actuator/sensor dynamical behavior, and excitation of nonlinear structure vibrations during control and assembly. For on-orbit assembly, controllers must start with finite element models of LSS and adapt on line to the best operating points, without compromising stability. This is not easy to do, since there are often unmodeled dynamic interactions between the controller and the structure. The indirect adaptive controllers are based on parameter estimation. Due to the large number of modes in LSS, this approach leads to very high-order control schemes with consequent poor stability and performance. In contrast, direct model reference adaptive controllers operate to force the LSS to track the desirable behavior of a chosen model. These schemes produce simple control algorithms which are easy to implement on line. One problem with their use for LSS has been that the model must be the same dimension as the LSS - i.e., quite large. A control theory based on the command generator tracker (CGT) ideas of Sobel, Mabins, Kaufman and Wen, Balas to obtain very low-order models based on adaptive algorithms was developed. Closed-loop stability for both finite element models and distributed parameter models of LSS was proved. In addition, successful numerical simulations on several LSS databases were obtained. An adaptive controller based on our theory was also implemented on a flexible robotic manipulator at Martin Marietta Astronautics. Computation schemes for controller-structure interaction with unmodeled modes, the residual mode filters or RMF, were developed. The RMF theory was modified to compensate slow actuator/sensor dynamics. These new ideas are being applied to LSS simulations to demonstrate the ease with which one can incorporate slow actuator/sensor effects into our design. It was also shown that residual mode filter compensation can be modified for small nonlinearities to produce exponentially stable closed-loop control. A theory for disturbance accommodating controllers based on reduced order models of structures was developed, and stability results for these controllers in closed-loop with large-scale finite element models of structures were obtained.
A flexible fuzzy set preference model first requires approximate methodologies for implementation. Fuzzy sets must be defined for each individual consumer using computer software, requiring a minimum of time and expertise on the part of the consumer. The amount of information needed in defining sets must also be established. The model itself must adapt fully to the subject's choice of attributes (vague or precise), attribute levels, and importance weights. The resulting individual-level model should be fully adapted to each consumer. The methodologies needed to develop this model will be equally useful in a new generation of intelligent systems which interact with ordinary consumers, controlling electronic devices through fuzzy expert systems or making recommendations based on a variety of inputs. The power of personal computers and their acceptance by consumers has yet to be fully utilized to create interactive knowledge systems that fully adapt their function to the user. Understanding individual consumer preferences is critical to the design of new products and the estimation of demand (market share) for existing products, which in turn is an input to management systems concerned with production and distribution. The question of what to make, for whom to make it and how much to make requires an understanding of the customer's preferences and the trade-offs that exist between alternatives. Conjoint analysis is a widely used methodology which de-composes an overall preference for an object into a combination of preferences for its constituent parts (attributes such as taste and price), which are combined using an appropriate combination function. Preferences are often expressed using linguistic terms which cannot be represented in conjoint models. Current models are also not implemented an individual level, making it difficult to reach meaningful conclusions about the cause of an individual's behavior from an aggregate model. The combination of complex aggregate models and vague linguistic preferences has greatly limited the usefulness and predictive validity of existing preference models. A fuzzy set preference model that uses linguistic variables and a fully interactive implementation should be able to simultaneously address these issues and substantially improve the accuracy of demand estimates. The parallel implementation of crisp and fuzzy conjoint models using identical data not only validates the fuzzy set model but also provides an opportunity to assess the impact of fuzzy set definitions and individual attribute choices implemented in the interactive methodology developed in this research. The generalized experimental tools needed for conjoint models can also be applied to many other types of intelligent systems.
We habituated the dominant time constant of the horizontal vestibuloocular reflex (VOR) of rhesus and cynomolgus monkeys by repeated testing with steps of velocity about a vertical axis and adapted the gain of the VOR by altering visual input with magnifying and reducing lenses. After baseline values were established, the nodulus and ventral uvula of the vestibulocerebellum were ablated in two monkeys, and the effects of nodulouvulectomy and flocculectomy on VOR gain adaptation and habituation were compared. The VOR time constant decreased with repeated testing, rapidly at first and more slowly thereafter. The gain of the VOR was unaffected. Massed trials were more effective than distributed trials in producing habituation. Regardless of the schedule of testing, the VOR time constant never fell below the time constant of the semicircular canals (approximately 5 s). This finding indicates that only the slow component of the vestibular response, the component produced by velocity storage, was habituated. In agreement with this, the time constant of optokinetic after-nystagmus (OKAN) was habituated concurrently with the VOR. Average values for VOR habituation were obtained on a per session basis for six animals. The VOR gain was adapted by natural head movements in partially habituated monkeys while they wore x 2.2 magnifying or x 0.5 reducing lenses. Adaptation occurred rapidly and reached about +/- 30%, similar to values obtained using forced rotation. VOR gain adaptation did not cause additional habituation of the time constant. When the VOR gain was reduced in animals with a long VOR time constant, there were overshoots in eye velocity that peaked at about 6-8 s after the onset or end of constant-velocity rotation. These overshoots occurred at times when the velocity storage integrator would have been maximally activated by semicircular canal input. Since the activity generated in the canals is not altered by visual adaptation, this finding indicates that the gain element that controls rapid changes in eye velocity in the VOR is separate from that which couples afferent input to velocity storage. Nodulouvulectomy caused a prompt and permanent loss of habituation, returning VOR time constants to initial values. VOR gain adaptation, which is lost after flocculectomy, was unaffected by nodulouvulectomy. Flocculectomy did not alter habituation of the VOR or of OKAN. Using a simplified model of the VOR, the decrease in the duration of vestibular nystagmus due to habituation was related to a decrement in the dominant time constant of the velocity storage integrator (1/h0).(ABSTRACT TRUNCATED AT 400 WORDS).
Performance Seeking Control attempts to find the operating condition that will generate optimal performance and control the plant at that operating condition. In this paper a nonlinear multivariable Adaptive Performance Seeking Control (APSC) methodology will be developed and it will be demonstrated on a nonlinear system. The APSC is comprised of the Positive Gradient Control (PGC) and the Fuzzy Model Reference Learning Control (FMRLC). The PGC computes the positive gradients of the desired performance function with respect to the control inputs in order to drive the plant set points to the operating point that will produce optimal performance. The PGC approach will be derived in this paper. The feedback control of the plant is performed by the FMRLC. For the FMRLC, the conventional fuzzy model reference learning control methodology is utilized, with guidelines generated here for the effective tuning of the FMRLC controller.