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.
The use of adaptive system to stabilize a launch vehicle with sloshable fuel is discussed. The adaptive system is based on a laser sensor to measure fluid displacement in the tank, a recursive-least-squares estimator to assess the natural frequency of the sloshing fuel, and a linear filter to stabilize the unstable fuel slosh. The method is found to be promising for a pitch-plane, single pendulum model.
By analyzing the motions of a thruster-controlled spacecraft, it is possible to provide on-line (1) thruster fault detection and isolation (FDI), and (2) vehicle mass- and thruster-property identification (ID). Technologies developed recently at NASA Ames have significantly improved the speed and accuracy of these ID and FDI capabilities, making them feasible for application to a broad class of spacecraft. Since these technologies use existing sensors, the improved system robustness and performance that comes with the thruster fault tolerance and system ID can be achieved through a software-only implementation. This contrasts with the added cost, mass, and hardware complexity commonly required by FDI. Originally developed in partnership with NASA - Johnson Space Center to provide thruster FDI capability for the X-38 during re-entry, these technologies are most recently being applied to the MIT SPHERES experimental spacecraft to fly on the International Space Station in 2004. The model-based FDI uses a maximum-likelihood calculation at its core, while the ID is based upon recursive least squares estimation. Flight test results from the SPHERES implementation, as flown aboard the NASA KC-1 35A 0-g simulator aircraft in November 2003 are presented.
This paper expands on the hybrid control architecture developed at the NASA Ames Research Center by addressing issues related to indirect adaptation using the recursive least squares (RLS) algorithm. Specifically, the hybrid control architecture is an adaptive flight controller that features both direct and indirect adaptation techniques. This paper will focus almost exclusively on the modifications necessary to achieve quality indirect adaptive control. Additionally this paper will present results that, using a full non -linear aircraft model, demonstrate the effectiveness of the hybrid control architecture given drastic changes in an aircraft s dynamics. Throughout the development of this topic, a thorough discussion of the RLS algorithm as a system identification technique will be provided along with results from seven well-known modifications to the popular RLS algorithm.
The purpose of this document is to explore the use of adaptive routines in battery modeling. The adaptive routines consist of real-time state estimators combined with battery parameter model components that are adjusted in real-time as battery data becomes available. Several aspects are explored. It is shown that model parameter identification is possible for simple battery models using available input/output data measurements. The online system identification used is recursive least squares. Model identification may be combined with a state observer such as the extended Kalman filter or the unscented Kalman filter to form an adaptive model combined with state estimation. However, such a combination is found to be problematic due to uncertainty, observability and stability issues. This paper is organized as follows. Section 1 introduces adaptive routines and possible roles they play in battery modeling. In Section 2 real-time parameter identification is described with results based on battery data. Section 3 reviews various state estimators and results using a simple battery model. In Section 4 parameter identification and state estimation are combined to form an adaptive routine. Finally, in Section 5 conclusions are drawn and future work is suggested.
Time-domain methods for the design of recursive digital filters using a squared error criterion are compared with a frequency-domain technique. Levy's method, which has been used to estimate transfer functions of continuous-time systems is modified to obtain design equations for digital filters. A special case of Levy's method is shown to be essentially equivalent to the time-domain methods.
Progress towards an adaptive self-tuning regulator (STR) for the cavity tone problem is discussed in this paper. Adaptive system identification algorithms were applied to an experimental cavity-flow tested as a prerequisite to control. In addition, a simple digital controller and a piezoelectric bimorph actuator were used to demonstrate multiple tone suppression. The control tests at Mach numbers of 0.275, 0.40, and 0.60 indicated approx. = 7dB tone reductions at multiple frequencies. Several different adaptive system identification algorithms were applied at a single freestream Mach number of 0.275. Adaptive finite-impulse response (FIR) filters of orders up to N = 100 were found to be unsuitable for modeling the cavity flow dynamics. Adaptive infinite-impulse response (IIR) filters of comparable order better captured the system dynamics. Two recursive algorithms, the least-mean square (LMS) and the recursive-least square (RLS), were utilized to update the adaptive filter coefficients. Given the sample-time requirements imposed by the cavity flow dynamics, the computational simplicity of the least mean squares (LMS) algorithm is advantageous for real-time control.
Generalized matrix inversion theory used in formulating recursive algorithm for least squares parameter estimation
This paper proposes a new approach of obtaining adaptive state estimation of a system in the presence of unknown system disturbances and measurement noise. In the beginning, a non-optimal Kalman filter with arbitrary initial guess for the process and measurement noises is implemented. At the same time, an adaptive transversal predictor (ATP) based on the recursive least-squares (RLS) algorithm is used to yield optimal one- to p- step-ahead output predictions using the previous input/output data. Referring to these optimal predictions the Kalman filter gain is updated and the performance of the state estimation is thus improved. If forgetting factor is implemented in the recursive least-squares algorithm, this method is also capable of dealing with the situation when the noise statistics are slowly time-varying. This feature makes this new approach especially suitable for the control of flexible structures. A numerical example demonstrates the feasibility of this real time adaptive state estimation method.
Least-squares (LS) estimation is a basic operation in many signal processing problems. Given y = Ax + v, where A is a m x n coefficient matrix, y is a m x 1 observation vector, and v is a m x 1 zero mean white noise vector, a simple least-squares solution is finding the estimated vector x which minimizes the norm of /Ax-y/. It is well known that for an ill-conditioned matrix A, solving least-squares problems by orthogonal triangular (QR) decomposition and back substitution has robust numerical properties under finite word length effect since 2-norm is preserved. Many fast algorithms have been proposed and applied to systolic arrays. Gentleman-Kung (1981) first presented the trianglular systolic array for a basic Givens reduction. McWhirter (1983) used this array structure to find the least-squares estimation errors. Then by geometric approach, several different systolic array realizations of the recursive least-squares estimation algorithms of Lee et al (1981) were derived by Kalson-Yao (1985). Basic QR decomposition algorithms are considered in this paper and it is found that under a one-row time updating situation, the Householder transformation degenerates to a simple Givens reduction. Next, an improved least-squares estimation algorithm is derived by considering a modified version of fast Givens reduction. From this approach, the basic relationship between Givens reduction and Modified-Gram-Schmidt transformation can easily be understood. This improved algorithm also has simpler computational and inter-cell connection complexities while compared with other known least-squares algorithms and is more realistic for systolic array implementation.
A square root information filter/smoother is derived using recursive least-squares arguments. The combined filter/smoother algorithm has the following attributes: (1) it has a square root structure, which enhances numerical accuracy; (2) filter and smoother mechanizations are identical in form, facilitating implementation of the smoother; and (3) storage and computation requirements are modest compared with other smoothing algorithms. Partitioning the results to separate bias parameters provides further computational economies and reduction of storage requirements.
The Dyer-McReynolds square-root information filter (SRIF) is rederived, using recursive least-square arguments. The result is applied to a system composed partly of biases. The filter sensitivity matrix, computed covariance, and consider covariance for this augmented system are reviewed. A new computationally attractive representation for the smoothed estimates, in terms of a smoothed sensitivity matrix and a smoothed computed covariance is presented.
A unified approach is presented for deriving a large class of new and previously known time- and order-recursive least-squares algorithms with systolic array architectures, suitable for high-throughput-rate and VLSI implementations of space-time filtering and system identification problems. The geometrical derivation given is unique in that no assumption is made concerning the rank of the sample data correlation matrix. This method utilizes and extends the concept of oblique projections, as used previously in the derivations of the least-squares lattice algorithms. Exponentially weighted least-squares criteria are considered for both sliding and growing memory.
Least-squares (LS) estimations and spectral decomposition algorithms constitute the heart of modern signal processing and communication problems. Implementations of recursive LS and spectral decomposition algorithms onto parallel processing architectures such as systolic arrays with efficient fault-tolerant schemes are the major concerns of this dissertation. There are four major results in this dissertation. First, we propose the systolic block Householder transformation with application to the recursive least-squares minimization. It is successfully implemented on a systolic array with a two-level pipelined implementation at the vector level as well as at the word level. Second, a real-time algorithm-based concurrent error detection scheme based on the residual method is proposed for the QRD RLS systolic array. The fault diagnosis, order degraded reconfiguration, and performance analysis are also considered. Third, the dynamic range, stability, error detection capability under finite-precision implementation, order degraded performance, and residual estimation under faulty situations for the QRD RLS systolic array are studied in details. Finally, we propose the use of multi-phase systolic algorithms for spectral decomposition based on the QR algorithm. Two systolic architectures, one based on triangular array and another based on rectangular array, are presented for the multiphase operations with fault-tolerant considerations. Eigenvectors and singular vectors can be easily obtained by using the multi-pase operations. Performance issues are also considered.
Calculation of the inertial orientation of the OSO-7 spacecraft from the times at which known stars or planets transit planes fixed in the spacecraft. Both the reference planes and the timing information are provided by the star scanner instrument aboard the spacecraft, while the star identification and the statistical estimation of a set of parameters describing the spacecraft attitude are accomplished in a ground station computer facility. A recursive least-squares determination is made of a vector of first-order differential corrections to the attitude state vector. Preliminary analysis indicates the system accuracy to be 3 arc min in each attitude Euler angle.
Approximation of noisy data in the plane by straight lines or elliptic or single-branch hyperbolic curve segments arises in pattern recognition, data compaction, and other problems. The efficient search for and approximation of data by such curves were examined. Recursive least-squares linear curve-fitting was used, and ellipses and hyperbolas are parameterized as quadratic functions in x and y. The error minimized by the algorithm is interpreted, and central processing unit (CPU) times for estimating parameters for fitting straight lines and quadratic curves were determined and compared. CPU time for data search was also determined for the case of straight line fitting. Quadratic curve fitting is shown to require about six times as much CPU time as does straight line fitting, and curves relating CPU time and fitting error were determined for straight line fitting. Results are derived on early sequential determination of whether or not the underlying curve is a straight line.
For Large Space Structures (LSS), an adaptive control system is highly desirable. The present investigation is concerned with an 'indirect' adaptive control scheme wherein the system order, mode shapes, and modal amplitudes are estimated on-line using an identification scheme based on recursive, least-squares, lattice filters. Using the identified model parameters, a modal control law based on a pole-placement scheme with the objective of vibration suppression is employed. A method is presented for closed loop adaptive control of a flexible free-free beam. The adaptive control scheme consists of a two stage identification scheme working in series and a modal pole placement control scheme. The main conclusion from the current study is that the identified parameters cannot be directly used for controller design purposes.
Identification of frequencies, damping ratios, and mode shapes of large space structures (LSSs) are examined in real time. Real time processing allows for quick updates of model processing after a reconfiguration of structural failure. Recursive lattice least squares (RLLS) was selected as the baseline algorithm for the identification. Simulation results on a one dimensional LSS demonstrated that it provides good estimates, was not ill-conditioned in the presence of under-excited modes, allowed activity by a supervisory control system which prevented damage to the LSS or excessive drift, and was capable of real-time processing for typical LSS models. A suboptimal version of RLLS, which is equivalent to simulated parallel processing, was derived. A NASTRAN model of the dual keel U.S. space station was used to demonstrate the input/identification algorithm package in a more realistic simulation. Because the first eight flexible modes were very close together, the identification was much more difficult than in the simple examples. Even so, the model was accurately identified in real time.
An averaging analysis of indirect, discrete-time, adaptive control systems is presented. The analysis results in a signal-dependent stability condition and accounts for unmodeled plant dynamics as well as exogenous disturbances. This analysis is applied to two discrete-time adaptive algorithms: an unnormalized gradient algorithm and a recursive least-squares (RLS) algorithm with resetting. Since linearization and averaging are used for the gradient analysis, a local stability result valid for small adaptation gains is found. For RLS with resetting, the assumption is that there is a long time between resets. The results for the two algorithms are virtually identical, emphasizing their similarities in adaptive control.