SEARCH · Search NASA

Results for “Linear and nonlinear operators”

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.

SourceResearch connectionFrom yearThrough year

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 379 records · Page 21

All sources NASA NTRS DOE OSTI

NASA NTRS1981

Simulation of Heavy Lift Airship dynamics over large ranges of incidence and speed

A nonlinear, multibody, six-degrees-of-freedom digital simulation has been developed to study generic Heavy Lift Airship (HLA) dynamics and control. The basic aerodynamic functions developed to model the hull, tail, and rotor loads continuously over all incidence ranges are reviewed and applied to a Quadrotor HLA with a low fineness ratio hull and a small vee-tail. Trim calculations for a test vehicle suggest control power deficiencies in crosswind stationkeeping for the unloaded vehicle. Gust responses show the importance of correctly calculating loads due to accelerated relative motion of air and hull. Numerically linearized dynamics for the test vehicle show the existence of a divergent yaw mode, and an oscillatory pitch mode whose stability characteristics are sensitive to flight speed. A considerable improvement in the vehicle's stability and response results from a simple multi-axis closed-loop control system operating on the rotors and propeller blades.

Tischler, M. B.↗

NASA NTRS1993

Space Shuttle Main Engine performance analysis

For a number of years, NASA has relied primarily upon periodically updated versions of Rocketdyne's power balance model (PBM) to provide space shuttle main engine (SSME) steady-state performance prediction. A recent computational study indicated that PBM predictions do not satisfy fundamental energy conservation principles. More recently, SSME test results provided by the Technology Test Bed (TTB) program have indicated significant discrepancies between PBM flow and temperature predictions and TTB observations. Results of these investigations have diminished confidence in the predictions provided by PBM, and motivated the development of new computational tools for supporting SSME performance analysis. A multivariate least squares regression algorithm was developed and implemented during this effort in order to efficiently characterize TTB data. This procedure, called the 'gains model,' was used to approximate the variation of SSME performance parameters such as flow rate, pressure, temperature, speed, and assorted hardware characteristics in terms of six assumed independent influences. These six influences were engine power level, mixture ratio, fuel inlet pressure and temperature, and oxidizer inlet pressure and temperature. A BFGS optimization algorithm provided the base procedure for determining regression coefficients for both linear and full quadratic approximations of parameter variation. Statistical information relative to data deviation from regression derived relations was also computed. A new strategy for integrating test data with theoretical performance prediction was also investigated. The current integration procedure employed by PBM treats test data as pristine and adjusts hardware characteristics in a heuristic manner to achieve engine balance. Within PBM, this integration procedure is called 'data reduction.' By contrast, the new data integration procedure, termed 'reconciliation,' uses mathematical optimization techniques, and requires both measurement and balance uncertainty estimates. The reconciler attempts to select operational parameters that minimize the difference between theoretical prediction and observation. Selected values are further constrained to fall within measurement uncertainty limits and to satisfy fundamental physical relations (mass conservation, energy conservation, pressure drop relations, etc.) within uncertainty estimates for all SSME subsystems. The parameter selection problem described above is a traditional nonlinear programming problem. The reconciler employs a mixed penalty method to determine optimum values of SSME operating parameters associated with this problem formulation.

Santi, L. Michael↗

NASA NTRS1980

The collective excitation of g-modes in the sun

Oscillations of the solar interior (linear g-modes) may be strongly driven by the collective influence of all the modes upon the nuclear reactions in the core. This heretofore neglected effect could couple the modes, reduce the effective amplitudes near the center, and spatially concentrate most of the oscillation energy into just a portion of the radiative interior. If operating at sufficient strength, this can reverse the conventional conclusion, drawn from single mode calculations, that almost all solar g-modes are damped. Furthermore, it would put the theory into rough harmony with three otherwise troubling observations: (1) the 'low' neutrino flux measured by Davis (1978), (2) the high correspondence found by Wolff (1976) between recurrence periods in solar activity and the rotational beat periods of g-modes, and (3) the fluctuations in the sun's diameter which imply g-mode activity at high angular harmonics (Hill and Caudell 1979). A nonlinear expression is derived for the local rate of work done on an array of oscillation modes by the nuclear reactions. Three additional tests of the model are suggested.

Wolff, C. L.↗

NASA NTRS1989

Recursive linearization of multibody dynamics equations of motion

The equations of motion of a multibody system are nonlinear in nature, and thus pose a difficult problem in linear control design. One approach is to have a first-order approximation through the numerical perturbations at a given configuration, and to design a control law based on the linearized model. Here, a linearized model is generated analytically by following the footsteps of the recursive derivation of the equations of motion. The equations of motion are first written in a Newton-Euler form, which is systematic and easy to construct; then, they are transformed into a relative coordinate representation, which is more efficient in computation. A new computational method for linearization is obtained by applying a series of first-order analytical approximations to the recursive kinematic relationships. The method has proved to be computationally more efficient because of its recursive nature. It has also turned out to be more accurate because of the fact that analytical perturbation circumvents numerical differentiation and other associated numerical operations that may accumulate computational error, thus requiring only analytical operations of matrices and vectors. The power of the proposed linearization algorithm is demonstrated, in comparison to a numerical perturbation method, with a two-link manipulator and a seven degrees of freedom robotic manipulator. Its application to control design is also demonstrated.

Lin, Tsung-Chieh↗

NASA NTRS1982

Toward a unified theory of atmospheric convective instability

A nonlinear three-dimensional truncated spectral model of shallow and moist Boussinesq convection indicates that parallel instability and thermal forcing are linked, in view of the fact that only one convective mode exists in which either or both mechanisms are operating to generate convection in the planetary boundary layer. It is also established that the wind field causes two-dimensional roll convection formation, an alignment of the convection with the wind in a preferred manner, and a propagation speed that is related to the wind component perpendicular to the roll axis. Latent heating is responsible for the decrease of the critical value of the environmental lapse rate in accordance with the slice method stability criterion. When only the upper part of the upward branch is moist and all of the downward branch is dry, latent heating also causes a finite-amplitude convective solution for Rayleigh number values lower than the critical value of linear analysis.

Shirer, H. N.↗

NASA NTRS1989

Preliminary results on noncollocated torque control of space robot actuators

In the Space Station era, more operations will be performed robotically in space in the areas of servicing, assembly, and experiment tending among others. These robots may have various sets of requirements for accuracy, speed, and force generation, but there will be design constraints such as size, mass, and power dissipation limits. For actuation, a leading motor candidate is a dc brushless type, and there are numerous potential drive trains each with its own advantages and disadvantages. This experiment uses a harmonic drive and addresses some inherent limitations, namely its backdriveability and low frequency structural resonances. These effects are controlled and diminished by instrumenting the actuator system with a torque transducer on the output shaft. This noncollocated loop is closed to ensure that the commanded torque is accurately delivered to the manipulator link. The actuator system is modelled and its essential parameters identified. The nonlinear model for simulations will include inertias, gearing, stiction, flexibility, and the effects of output load variations. A linear model is extracted and used for designing the noncollocated torque and position feedback loops. These loops are simulated with the structural frequency encountered in the testbed system. Simulation results are given for various commands in position. The use of torque feedback is demonstrated to yield superior performance in settling time and positioning accuracy. An experimental setup being finished consists of a bench mounted motor and harmonic drive actuator system. A torque transducer and two position encoders, each with sufficient resolution and bandwidth, will provide sensory information. Parameters of the physical system are being identified and matched to analytical predictions. Initial feedback control laws will be incorporated in the bench test equipment and various experiments run to validate the designs. The status of these experiments is given.

Tilley, Scott W.↗

NASA NTRS1984

Variational objective analyses for cyclone studies

The basic analysis equations, i.e., the two horizontal momentum equations, the hydrostatic equation, and the integrated continuity equation were derived for the nonlinear vertical coordinate, nondimensionalized, and expressed in finite differences on a staggered grid. Special care was taken to transform the hydrostatic equation and the pressure gradient terms of the horizontal momentum equations to nearly eliminate truncation error over steeply sloping terrain. This formulation also eliminated explicit reference to orographically induced variations in the thermodynamic variables so that the variational adjustments are on the scale of the meteorological perturbations. The analysis equations were subjected to the Euler-Lagrange operations as expressed for finite differences and an additional set of five partial differential equations was derived, bringing to nine the number of equations in Model I. Higher order terms, terms containing observed quantities, and terms containing none of the variables to be adjusted were grouped into forcing functions and the equations were solved for the zero order terms. Zero order variables were eliminated between these equations and there resulted two diagnostic equations which take the form of general linear second order partial differential equations with nonconstant coefficients.

Achtemeier, G. L.↗

NASA NTRS1997

Applications of Data Assimilation to Analysis of the Ocean on Large Scales

It is commonplace to begin talks on this topic by noting that oceanographic data are too scarce and sparse to provide complete initial and boundary conditions for large-scale ocean models. Even considering the availability of remotely-sensed data such as radar altimetry from the TOPEX and ERS-1 satellites, a glance at a map of available subsurface data should convince most observers that this is still the case. Data are still too sparse for comprehensive treatment of interannual to interdecadal climate change through the use of models, since the new data sets have not been around for very long. In view of the dearth of data, we must note that the overall picture is changing rapidly. Recently, there have been a number of large scale ocean analysis and prediction efforts, some of which now run on an operational or at least quasi-operational basis, most notably the model based analyses of the tropical oceans. These programs are modeled on numerical weather prediction. Aside from the success of the global tide models, assimilation of data in the tropics, in support of prediction and analysis of seasonal to interannual climate change, is probably the area of large scale ocean modeling and data assimilation in which the most progress has been made. Climate change is a problem which is particularly suited to advanced data assimilation methods. Linear models are useful, and the linear theory can be exploited. For the most part, the data are sufficiently sparse that implementation of advanced methods is worthwhile. As an example of a large scale data assimilation experiment with a recent extensive data set, we present results of a tropical ocean experiment in which the Kalman filter was used to assimilate three years of altimetric data from Geosat into a coarsely resolved linearized long wave shallow water model. Since nonlinear processes dominate the local dynamic signal outside the tropics, subsurface dynamical quantities cannot be reliably inferred from surface height anomalies. Because of its potential for large scale synoptic coverage of the deep ocean, acoustic travel time data should be a natural complement to satellite altimetry. Satellite data give us vertical integrals associated with thermodynamic and dynamic processes.

Miller, Robert N.↗

NASA NTRS2007

Development of a Linear Stirling Model with Varying Heat Inputs

The linear model of the Stirling system developed by NASA Glenn Research Center (GRC) has been extended to include a user-specified heat input. Previously developed linear models were limited to the Stirling convertor and electrical load. They represented the thermodynamic cycle with pressure factors that remained constant. The numerical values of the pressure factors were generated by linearizing GRC s non-linear System Dynamic Model (SDM) of the convertor at a chosen operating point. The pressure factors were fixed for that operating point, thus, the model lost accuracy if a transition to a different operating point were simulated. Although the previous linear model was used in developing controllers that manipulated current, voltage, and piston position, it could not be used in the development of control algorithms that regulated hot-end temperature. This basic model was extended to include the thermal dynamics associated with a hot-end temperature that varies over time in response to external changes as well as to changes in the Stirling cycle. The linear model described herein includes not only dynamics of the piston, displacer, gas, and electrical circuit, but also the transient effects of the heater head thermal inertia. The linear version algebraically couples two separate linear dynamic models, one model of the Stirling convertor and one model of the thermal system, through the pressure factors. The thermal system model includes heat flow of heat transfer fluid, insulation loss, and temperature drops from the heat source to the Stirling convertor expansion space. The linear model was compared to a nonlinear model, and performance was very similar. The resulting linear model can be implemented in a variety of computing environments, and is suitable for analysis with classical and state space controls analysis techniques.

Regan, Timothy F.↗

NASA NTRS1976

Guidance logic for spiral approaches

A spiral approach concept is proposed as a standard procedure for independent, commercial VTOL operations conducted in close proximity to CTOL operations. A guidance logic is developed for this VTOL application, although the results may be applicable to curved approaches by STOL or CTOL aircraft as well. The guidance concept attempts to maintain constant airspeed along a fixed radius nominal spiral. The presence of wind requires a continuous variation in bank angle and heading rate to remain on the desired path. Linear perturbation analysis is used to select satisfactory feedback gains for commanded bank angle, longitudinal acceleration and vertical speed. A wind estimator detects differences between the predicted and observed wind and uses the result to modify the nominal control. For 4-D guidance the nominal time must be computed as a function of turn angle, which requires the solution of an elliptic integral. Time-varying longitudinal accelerations are necessary for 4-D guidance when the observed wind differs from the predicted wind, or when wind shear is present. The logic and the linear feedback gains have been tested in a nonlinear simulation. Results have generally verified the performance predicted by linear analysis.

Hollister, W. M.↗

NASA NTRS2006

Calibration Technique for Polarization-Sensitive Lidars

Polarization-sensitive lidars have proven to be highly effective in discriminating between spherical and non-spherical particles in the atmosphere. These lidars use a linearly polarized laser and are equipped with a receiver that can separately measure the components of the return signal polarized parallel and perpendicular to the outgoing beam. In this work we describe a technique for calibrating polarization-sensitive lidars that was originally developed at NASA s Langley Research Center (LaRC) and has been used continually over the past fifteen years. The procedure uses a rotatable half-wave plate inserted into the optical path of the lidar receiver to introduce controlled amounts of polarization cross-talk into a sequence of atmospheric backscatter measurements. Solving the resulting system of nonlinear equations generates the system calibration constants (gain ratio, G, and offset angle, theta) required for deriving calibrated measurements of depolarization ratio from the lidar signals. In addition, this procedure also determines the mean depolarization ratio within the region of the atmosphere that is analyzed. Simulations and error propagation studies show the method to be both reliable and well behaved. Operational details of the technique are illustrated using measurements obtained as part of Langley Research Center s participation in the First ISCCP Regional Experiment (FIRE).

Alvarez, J. M.↗

NASA NTRS2024

Spacecraft Modeling, Attitude Determination, and Control: Quaternion-Based Approach

Spacecraft attitude determination and control is an important part of a spacecraft to achieve its designed mission. As of today, many spacecrafts have been successfully launched, and most of them have performed well as they were designed. Many research papers have been published to address the attitude determination and control design problems. Several text books are available for students to learn the technology and for engineers to use as references. The most popular spacecraft models for attitude determination algorithms and control design methods are the Euler angle models and the quaternion models . The Euler angle models have been proved very efficient as the linearized models are controllable, and all standard linear control system design methods are directly applicable. The drawbacks related to the Euler angle methods are (a) the designs based on linearized models may not globally stabilize the original nonlinear spacecraft, i.e., the design may not work when the attitude of the spacecraft is far away from the point where the linearization is performed; (b) the models depend on the rotational sequences, this can be error prone if several teams work on the same project and they use different rotational sequences; (c) for any rotational sequence, there is a singular point where the model is not applicable; and (d) since most attitude determination methods use quaternion to represent the spacecraft attitude, there is a need to transform quaternion into Euler angles. On the other hand, for quaternion models, people have found controllers that can globally stabilize nonlinear spacecraft systems; the models do not depend on rotational sequences and they have no singular point; and the quaternion is provided by attitude determination system and ready to use. The main problem with the quaternion model based control system design is that the linearized quaternion model is not controllable. Therefore, most published design methods heavily rely on Lyapunov functions for the nonlinear spacecraft system. But there is no systematic way to obtain a desired Lyapunov functions. Moreover, the Lyapunov function based designs focus on the closed-loop system stability but pay little attention to the closed-loop system performance. In a series of papers, the author proposed some reduced quaternion models which lead to some controllable linearized spacecraft models. Therefore, all standard linear system theory can be directly applied to analyze and design the spacecraft control systems. We showed that, in some cases, the designed control system is not only optimal for the linearized system, but also globally stabilize the original nonlinear system . Clearly, the reduced quaternion models do not depend on rotational sequences. Due to the special structure of the linearized spacecraft model, some most important design methods, such as LQR design and robust pole assignment design are very simple, enjoy the analytical solutions for some problems, have direct connection to the performance measures, such as settling time, rising time , and percentage of overshoot . All these features are attractive for high quality control system designs. The idea mentioned above is then extended to more spacecraft control problems using specific actuators such as magnetic torque bars and control momentum gyroscopes. These types of actuators may not provide exactly desired torques. Most existing methods use different conversions to get approximate solutions, meaning that these actuators may generate a torque close to but not equal to the desired one. Using the reduced quaternion models that incorporate the actuators into the system model, the control inputs are not torques but the operational parameters. The main benefit of this idea is that the control actions are not approximate but accurate. As all actuators have their operational limit, design with input constraints are also considered in this book by using recently developed interior-point optimization techniques. This book grows up from my research on the spacecraft attitude determination and control design methods in more than a decade which is focused on using reduced quaternion models because of their merits stated above. The book provides all necessary background materials on orbital dynamics, rotations and quaternion, frequently used reference frames, transformations between reference frames, space environment and disturbance torques, ephemeris astronomical vector calculations and measurement instruments, spacecraft control actuators and their models, so that the readers will get a global picture and can apply all these information into the spacecraft system modeling, attitude determination, and spacecraft control system designs, which is the main purpose of this book. This book is different from existing books in that we focus on quaternion based spacecraft control system designs and we consider only attitude control system design related problems, from spacecraft modeling, to attitude determination and estimation, to control system design method selection, to control algorithm development, and to the simulation of the control system designs. Moreover, this book addresses different attitude control tasks in the spacecraft life cycle, including spacecraft maneuver, orbit raising, attitude control, and rendezvous. Finally, this book emphasizes the state space design methods rather than the classical frequency design methods.

Attitude determination and control↗

NASA NTRS2008

An Integrated Approach to Damage Accommodation in Flight Control

In this paper we present an integrated approach to in-flight damage accommodation in flight control. The approach is based on Multiple Models, Switching and Tuning (MMST), and consists of three steps: In the first step the main objective is to acquire a realistic aircraft damage model. Modeling of in-flight damage is a highly complex problem since there is a large number of issues that need to be addressed. One of the most important one is that there is strong coupling between structural dynamics, aerodynamics, and flight control. These effects cannot be studied separately due to this coupling. Once a realistic damage model is available, in the second step a large number of models corresponding to different damage cases are generated. One possibility is to generate many linear models and interpolate between them to cover a large portion of the flight envelope. Once these models have been generated, we will implement a recently developed-Model Set Reduction (MSR) technique. The technique is based on parameterizing damage in terms of uncertain parameters, and uses concepts from robust control theory to arrive at a small number of "centered" models such that the controllers corresponding to these models assure desired stability and robustness properties over a subset in the parametric space. By devising a suitable model placement strategy, the entire parametric set is covered with a relatively small number of models and controllers. The third step consists of designing a Multiple Models, Switching and Tuning (MMST) strategy for estimating the current operating regime (damage case) of the aircraft, and switching to the corresponding controller to achieve effective damage accommodation and the desired performance. In the paper present a comprehensive approach to damage accommodation using Model Set Design,MMST, and Variable Structure compensation for coupling nonlinearities. The approach was evaluated on a model of F/A-18 aircraft dynamics under control effector damage, augmented by nonlinear cross-coupling terms and a structural dynamics model. The proposed approach achieved excellent performance under severe damage effects.

Boskovic, Jovan D.↗

NASA NTRS1989

A review of some problems in global-local stress analysis

The various types of local-global finite-element problems point out the need to develop a new generation of software. First, this new software needs to have a complete analysis capability, encompassing linear and nonlinear analysis of 1-, 2-, and 3-dimensional finite-element models, as well as mixed dimensional models. The software must be capable of treating static and dynamic (vibration and transient response) problems, including the stability effects of initial stress, and the software should be able to treat both elastic and elasto-plastic materials. The software should carry a set of optional diagnostics to assist the program user during model generation in order to help avoid obvious structural modeling errors. In addition, the program software should be well documented so the user has a complete technical reference for each type of element contained in the program library, including information on such topics as the type of numerical integration, use of underintegration, and inclusion of incompatible modes, etc. Some packaged information should also be available to assist the user in building mixed-dimensional models. An important advancement in finite-element software should be in the development of program modularity, so that the user can select from a menu various basic operations in matrix structural analysis.

Nelson, Richard B.↗

NASA NTRS2000

Solar Spicules: A Review of Recent Models and Targets for Future Observations

Since their discovery over 100 years ago, there have been many suggestions for the origin and development of solar spicules. Because the velocities of spicules are comparable to the sound and Alfven speeds of the low chromosphere, linear theory cannot fully describe them. Consequently, detailed tests of theoretical ideas had to await the development of computing power that only became available during the 1970s. This work reviews theories for spicules and spicule-like features over approximately the past 25 years, with an emphasis on the models based on nonlinear numerical simulations. These models have given us physical insight into wave propagation in the solar atmosphere, and have helped elucidate how such waves, and associated shock waves, may be capable of creating motions and structures on magnetic flux tubes in the lower solar atmosphere. So far, however, it has been difficult to reproduce the most- commonly -quoted parameters for spicules with these models, using what appears to be the most suitable input parameters. A key impediment to developing satisfactory models has been the lack of reliable observational information, which is a consequence of the small angular size and transient lifetime of spicules. I close with a list of key observational questions to be addressed with space-based satellites, such as the currently operating TRACE satellite, and especially the upcoming Solar-B mission. Answers to these questions will help determine which, if any, of the current models correctly explains spicules.

Sterling, Alphonse C.↗

NASA NTRS2000

Solar Spicules: A Review of Recent Models and Targets for Future Observations

Since their discovery over 100 years ago, there have been many suggestions for the origin and development of solar spicules. Because the velocities of spicules are comparable to the sound and Alfven speeds of the low chromosphere, linear theory cannot fully describe them. Consequently, detailed tests of theoretical ideas had to await the development of computing power that only became available during the 1970s. This work reviews theories for spicules and spicule-like features over approximately the past 25 years, with an emphasis an the models based on nonlinear numerical simulations. These models have given us physical insight into wave propagation in the solar atmosphere and have helped elucidate how such waves, and associated shock waves, may be capable of creating motions and structures on magnetic flux tubes in the lower solar atmosphere. So far, however, it has been difficult to reproduce the most commonly quoted parameters for spicules with these models, using what appears to be the most suitable input parameters. A key impediment to developing satisfactory models has been the lack of reliable observational information, which is a consequence of the small angular size and transient lifetime of spicules. I close with a list of key observational questions to be addressed with space-based satellites, such as the currently operating Transition Region and Coronal Explorer (TRACE) satellite, and especially the upcoming Solar-B mission. Answers to these questions will help determine which, if any, of the current models correctly explains spicules.

Sterling, Alphonse C.↗

NASA NTRS2016

Model Development for MODIS Thermal Band Electronic Crosstalk

MODerate-resolution Imaging Spectroradiometer (MODIS) has 36 bands. Among them, 16 thermal emissive bands covering a wavelength range from 3.8 to 14.4 m. After 16 years on-orbit operation, the electronic crosstalk of a few Terra MODIS thermal emissive bands developed substantial issues that cause biases in the EV brightness temperature measurements and surface feature contamination. The crosstalk effects on band 27 with center wavelength at 6.7 m and band 29 at 8.5 m increased significantly in recent years, affecting downstream products such as water vapor and cloud mask. The crosstalk effect is evident in the near-monthly scheduled lunar measurements, from which the crosstalk coefficients can be derived. The development of an alternative approach is very helpful for independent verification.In this work, a physical model was developed to assess the crosstalk impact on calibration as well as in Earth view brightness temperature retrieval. This model was applied to Terra MODIS band 29 empirically to correct the Earth brightness temperature measurements. In the model development, the detectors nonlinear response is considered. The impact of the electronic crosstalk is assessed in two steps. The first step consists of determining the impact on calibration using the on-board blackbody (BB). Due to the detectors nonlinear response and large background signal, both linear and nonlinear coefficients are affected by the crosstalk from sending bands. The second step is to calculate the effects on the Earth view brightness temperature retrieval. The effects include those from affected calibration coefficients and the contamination of Earth view measurements. This model links the measurement bias with crosstalk coefficients, detector non-linearity, and the ratio of Earth measurements between the sending and receiving bands. The correction of the electronic cross talk can be implemented empirically from the processed bias at different brightness temperature. The implementation can be done through two approaches. As routine calibration assessment for thermal infrared bands, the trending over select Earth scenes is processed for all the detectors in a band and the band averaged bias is derived at a certain time. In this case, the correction of an affected band can be made using the regression of the model with band averaged bias and then corrections of detector differences are applied. The second approach requires the trending for individual detectors and the bias for each detector is used for regression with the model. A test using the first approach was made for Terra MODIS band 29 with the biases derived from long-term trending of brightness temperature over ocean and Dome-C.

Chang, Tiejun↗

NASA NTRS2007

Identification and Control of Aircrafts using Multiple Models and Adaptive Critics

We compared two possible implementations of local linear models for control: one approach is based on a self-organizing map (SOM) to cluster the dynamics followed by a set of linear models operating at each cluster. Therefore the gating function is hard (a single local model will represent the regional dynamics). This simplifies the controller design since there is a one to one mapping between controllers and local models. The second approach uses a soft gate using a probabilistic framework based on a Gaussian Mixture Model (also called a dynamic mixture of experts). In this approach several models may be active at a given time, we can expect a smaller number of models, but the controller design is more involved, with potentially better noise rejection characteristics. Our experiments showed that the SOM provides overall best performance in high SNRs, but the performance degrades faster than with the GMM for the same noise conditions. The SOM approach required about an order of magnitude more models than the GMM, so in terms of implementation cost, the GMM is preferable. The design of the SOM is straight forward, while the design of the GMM controllers, although still reasonable, is more involved and needs more care in the selection of the parameters. Either one of these locally linear approaches outperform global nonlinear controllers based on neural networks, such as the time delay neural network (TDNN). Therefore, in essence the local model approach warrants practical implementations. In order to call the attention of the control community for this design methodology we extended successfully the multiple model approach to PID controllers (still today the most widely used control scheme in the industry), and wrote a paper on this subject. The echo state network (ESN) is a recurrent neural network with the special characteristics that only the output parameters are trained. The recurrent connections are preset according to the problem domain and are fixed. In a nutshell, the states of the reservoir of recurrent processing elements implement a projection space, where the desired response is optimally projected. This architecture trades training efficiency by a large increase in the dimension of the recurrent layer. However, the power of the recurrent neural networks can be brought to bear on practical difficult problems. Our goal was to implement an adaptive critic architecture implementing Bellman s approach to optimal control. However, we could only characterize the ESN performance as a critic in value function evaluation, which is just one of the pieces of the overall adaptive critic controller. The results were very convincing, and the simplicity of the implementation was unparalleled.

Principe, Jose C.↗