Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Kalman filtering”

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.

At least 379 records · Page 21
All sourcesNASA NTRSDOE OSTI
NASA NTRS1991

A fast algorithm for control and estimation using a polynomial state-space structure

One of the major problems associated with the control of flexible structures is the estimation of system states. Since the parameters of the structures are not constant under varying loads and conditions, conventional fixed parameter state estimators can not be used to effectively estimate the states of the system. One alternative is to use a state estimator which adapts to the condition of the system. One such estimator is the Kalman filter. This filter is a time varying recursive digital filter which is based upon a model of the system being measured. This filter adapts the model according to the output of the system. Previously, the Kalman filter has only been used in an off-line capacity due to the computational time required for implementation. With recent advances in computer technology, it is becoming a viable tool for use in the on-line environment. A distributed Kalman filter implementation is described for fast estimation of the state of a flexible arm. A key issue, is the sensor structure and initial work on a distributed sensor that could be used with the Kalman filter is presented.

Shults, James R.↗
NASA NTRS2005

An Optimal Orthogonal Decomposition Method for Kalman Filter-Based Turbofan Engine Thrust Estimation

A new linear point design technique is presented for the determination of tuning parameters that enable the optimal estimation of unmeasured engine outputs such as thrust. The engine s performance is affected by its level of degradation, generally described in terms of unmeasurable health parameters related to each major engine component. Accurate thrust reconstruction depends upon knowledge of these health parameters, but there are usually too few sensors to be able to estimate their values. In this new technique, a set of tuning parameters is determined which accounts for degradation by representing the overall effect of the larger set of health parameters as closely as possible in a least squares sense. The technique takes advantage of the properties of the singular value decomposition of a matrix to generate a tuning parameter vector of low enough dimension that it can be estimated by a Kalman filter. A concise design procedure to generate a tuning vector that specifically takes into account the variables of interest is presented. An example demonstrates the tuning parameters ability to facilitate matching of both measured and unmeasured engine outputs, as well as state variables. Additional properties of the formulation are shown to lend themselves well to diagnostics.

Litt, Jonathan S.↗
NASA NTRS2007

An Optimal Orthogonal Decomposition Method for Kalman Filter-Based Turbofan Engine Thrust Estimation

A new linear point design technique is presented for the determination of tuning parameters that enable the optimal estimation of unmeasured engine outputs, such as thrust. The engine's performance is affected by its level of degradation, generally described in terms of unmeasurable health parameters related to each major engine component. Accurate thrust reconstruction depends on knowledge of these health parameters, but there are usually too few sensors to be able to estimate their values. In this new technique, a set of tuning parameters is determined that accounts for degradation by representing the overall effect of the larger set of health parameters as closely as possible in a least squares sense. The technique takes advantage of the properties of the singular value decomposition of a matrix to generate a tuning parameter vector of low enough dimension that it can be estimated by a Kalman filter. A concise design procedure to generate a tuning vector that specifically takes into account the variables of interest is presented. An example demonstrates the tuning parameters ability to facilitate matching of both measured and unmeasured engine outputs, as well as state variables. Additional properties of the formulation are shown to lend themselves well to diagnostics.

Litt, Jonathan S.↗
NASA NTRS2007

An Optimal Orthogonal Decomposition Method for Kalman Filter-Based Turbofan Engine Thrust Estimation

A new linear point design technique is presented for the determination of tuning parameters that enable the optimal estimation of unmeasured engine outputs, such as thrust. The engine s performance is affected by its level of degradation, generally described in terms of unmeasurable health parameters related to each major engine component. Accurate thrust reconstruction depends on knowledge of these health parameters, but there are usually too few sensors to be able to estimate their values. In this new technique, a set of tuning parameters is determined that accounts for degradation by representing the overall effect of the larger set of health parameters as closely as possible in a least-squares sense. The technique takes advantage of the properties of the singular value decomposition of a matrix to generate a tuning parameter vector of low enough dimension that it can be estimated by a Kalman filter. A concise design procedure to generate a tuning vector that specifically takes into account the variables of interest is presented. An example demonstrates the tuning parameters ability to facilitate matching of both measured and unmeasured engine outputs, as well as state variables. Additional properties of the formulation are shown to lend themselves well to diagnostics.

Litt, Jonathan S.↗
NASA NTRS1998

Comparison of Five System Identification Algorithms for Rotorcraft Higher Harmonic Control

This report presents an analysis and performance comparison of five system identification algorithms. The methods are presented in the context of identifying a frequency-domain transfer matrix for the higher harmonic control (HHC) of helicopter vibration. The five system identification algorithms include three previously proposed methods: (1) the weighted-least- squares-error approach (in moving-block format), (2) the Kalman filter method, and (3) the least-mean-squares (LMS) filter method. In addition there are two new ones: (4) a generalized Kalman filter method and (5) a generalized LMS filter method. The generalized Kalman filter method and the generalized LMS filter method were derived as extensions of the classic methods to permit identification by using more than one measurement per identification cycle. Simulation results are presented for conditions ranging from the ideal case of a stationary transfer matrix and no measurement noise to the more complex cases involving both measurement noise and transfer-matrix variation. Both open-loop identification and closed- loop identification were simulated. Closed-loop mode identification was more challenging than open-loop identification because of the decreasing signal-to-noise ratio as the vibration became reduced. The closed-loop simulation considered both local-model identification, with measured vibration feedback and global-model identification with feedback of the identified uncontrolled vibration. The algorithms were evaluated in terms of their accuracy, stability, convergence properties, computation speeds, and relative ease of implementation.

Jacklin, Stephen A.↗
NASA NTRS1991

Geodesy by radio interferometry - Water vapor radiometry for estimation of the wet delay

An important source of error in VLBI estimates of baseline length is unmodeled variations of the refractivity of the neutral atmosphere along the propagation path of the radio signals. This paper presents and discusses the method of using data from a water vapor radiomete (WVR) to correct for the propagation delay caused by atmospheric water vapor, the major cause of these variations. Data from different WVRs are compared with estimated propagation delays obtained by Kalman filtering of the VLBI data themselves. The consequences of using either WVR data or Kalman filtering to correct for atmospheric propagation delay at the Onsala VLBI site are investigated by studying the repeatability of estimated baseline lengths from Onsala to several other sites. The repeatability obtained for baseline length estimates shows that the methods of water vapor radiometry and Kalman filtering offer comparable accuracies when applied to VLBI observations obtained in the climate of the Swedish west coast. For the most frequently measured baseline in this study, the use of WVR data yielded a 13 percent smaller weighted-root-mean-square (WRMS) scatter of the baseline length estimates compared to the use of a Kalman filter. It is also clear that the 'best' minimum elevationi angle for VLBI observations depends on the accuracy of the determinations of the total propagation delay to be used, since the error in this delay increases with increasing air mass.

Elgered, G.↗
DOE OSTI2021

Open Source Fault-tolerant Grid Frequency Measurement for Solar Inverters

The Discrete Fourier transform (DFT) based measurement algorithms are one of the most common measurement algorithms for grid parameter estimation such as rms, phase angle, frequency. Over the past few years, many DFT based algorithms have been developed to enhance its measurement accuracy under steady-state and/or dynamic grid conditions. For example, an adaptive band-pass filter utilizing exponential modulation filter has been proposed to reduce measurement errors at the presence of large frequency deviations. Measurement accuracy of different algorithms including FIR filter, extended Kalman filtering (EKF), and enhanced DFT method have been compared in detail under different grid conditions. Two artificial signals that have 90-degree phase difference were constructed by the Clarke transformation to address the frequency spectrum leakage of DFT. A multi-module approach was developed to enhance both steady-state and dynamic measurement accuracies, in which each module was developed to eliminate some specific errors. Besides DFT-based measurement algorithms, some signal model-based algorithms have been developed to further improve the accuracy under dynamic conditions. However, a key drawback of the state-of-the-art algorithms is that they cannot perform measurements accurately during system transient faults. In the Blue Cut Fire event, there was a phase angle jump of about 26 degrees in the voltage waveform during the transient fault. The phase angle jump fault will cause waveform discontinuity, and these algorithms will fail to provide reliable measurements during this period because they typically assume the waveform to be measured is continuous, no matter what method (DFT, PLL, EKF, FIR, or Taylor WLS) is used for estimation. In fact, the measurement errors during the system transient faults like phase-jump is not required in the IEEE Standard. As a result, although a measurement instrument can pass the strict IEEE Standard, it could still be the source of the problem in the future if we have similar system transient faults, which could happen again. Therefore, developing the fault-tolerant measurement technology is the key to solve the problem.

14 SOLAR ENERGY↗
NASA NTRS2017

Comparison of EKF and UKF for Robotic Missions to Mars

The batched Kalman lter works well for a wide variety of orbit determination scenarios but requires linearized approximations to the underlying dynamical system. Furthermore, implementation depends on the computation of partial derivatives for all measurements and parameters used in the lter, a non-trivial task even when analytical derivations are possible. To address these limitations, many extensions and modi - cations to the linear Kalman lter have been proposed over the years since Kalman's initial publication. Our investigation examines two popular variants of the Kalman lter, namely the Extended Kalman Filter (EKF) and Unscented Kalman Filter (UKF), within the context of spacecraft missions to Mars. The EKF relies on the same mathematical basis as the classic version of the lter, however it sequentially updates the linearization as the ltering process is conducted. In contrast, the UKF is built upon the unscented transform which samples the nearby solution space and numerically propagates perturbed trajectories as part of the update / estimation process. We assess the performance of the ltering algorithms under a variety of mission scenarios with the goal of establishing decision criteria for which implementation to use under di ering circumstances.

Ely, Todd↗
NASA NTRS2021

Detection and Tracking of Aircraft in the Far-Field from Small Unmanned Aerial Systems

Onboard far-field aircraft detection is needed for safe non-cooperative traffic mitigation in autonomous small Unmanned Aerial System (sUAS) operations. Machine vision systems, based on standard optics and visible light detectors, possess the ideal size, weight, and power (SWaP) requirements for sUAS. This work presents the design and analysis of a novel aircraft detection and tracking pipeline based on optical sensing alone. Key contributions of the work include a refined range inequality model based on sensing and detection with FAA well-clear separation assurance distances between aircraft in mind, a detector fusion method to maximize the benefit of two image detectors, and a comparative analysis of Linear Kalman-filtering and Extended Kalman-filtering to seek optimal tracking performance. The pipeline is evaluated offline against multiple intruder platforms, using two types of flight encounters: multirotor sUAS vs. fixed-wing sUAS and multirotor sUAS vs. general aviation(GA)plane. Analysis is restricted to the rate-limiting head-on and departing collision volume cases vertically separated for safety. Results indicate that it is feasible to use the proposed optical spatial-temporal tracking algorithm to provide adequate alerting time to prevent penetration of well-clear separation volumes for both sUAS and GA aircraft.

Unmanned Aerial System↗
NASA NTRS2022

Exploration of an Adaptive Routine for Battery Modeling

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.

Adaptive↗
NASA NTRS1994

A real-time algorithm for integrating differential satellite and inertial navigation information during helicopter approach

A real-time, high-rate precision navigation Kalman filter algorithm is developed and analyzed. This Navigation algorithm blends various navigation data collected during terminal area approach of an instrumented helicopter. Navigation data collected include helicopter position and velocity from a global position system in differential mode (DGPS) as well as helicopter velocity and attitude from an inertial navigation system (INS). The goal of the Navigation algorithm is to increase the DGPS accuracy while producing navigational data at the 64 Hertz INS update rate. It is important to note that while the data was post flight processed, the Navigation algorithm was designed for real-time analysis. The design of the Navigation algorithm resulted in a nine-state Kalman filter. The Kalman filter's state matrix contains position, velocity, and velocity bias components. The filter updates positional readings with DGPS position, INS velocity, and velocity bias information. In addition, the filter incorporates a sporadic data rejection scheme. This relatively simple model met and exceeded the ten meter absolute positional requirement. The Navigation algorithm results were compared with truth data derived from a laser tracker. The helicopter flight profile included terminal glideslope angles of 3, 6, and 9 degrees. Two flight segments extracted during each terminal approach were used to evaluate the Navigation algorithm. The first segment recorded small dynamic maneuver in the lateral plane while motion in the vertical plane was recorded by the second segment. The longitudinal, lateral, and vertical averaged positional accuracies for all three glideslope approaches are as follows (mean plus or minus two standard deviations in meters): longitudinal (-0.03 plus or minus 1.41), lateral (-1.29 plus or minus 2.36), and vertical (-0.76 plus or minus 2.05).

Hoang, TY↗
NASA NTRS1994

A Real-Time Algorithm for Integrating Differential Satellite and Inertial Navigation Information During Helicopter Approach

A real-time, high-rate precision navigation Kalman filter algorithm is developed and analyzed. This navigation algorithm blends various navigation data collected during terminal area approach of an instrumented helicopter. Navigation data collected include helicopter position and velocity from a global position system in differential mode (DGPS) as well as helicopter velocity and attitude from an inertial navigation system (INS). The goal of the navigation algorithm is to increase the DGPS accuracy while producing navigational data at the 64 Hertz INS update rate. It is important to note that while the data was post flight processed, the navigation algorithm was designed for real-time analysis. The design of the navigation algorithm resulted in a nine-state Kalman filter. The Kalman filter's state matrix contains position, velocity, and velocity bias components. The filter updates positional readings with DGPS position, INS velocity, and velocity bias information. In addition, the filter incorporates a sporadic data rejection scheme. This relatively simple model met and exceeded the ten meter absolute positional requirement. The navigation algorithm results were compared with truth data derived from a laser tracker. The helicopter flight profile included terminal glideslope angles of 3, 6, and 9 degrees. Two flight segments extracted during each terminal approach were used to evaluate the navigation algorithm. The first segment recorded small dynamic maneuver in the lateral plane while motion in file vertical plane was recorded by the second segment. The longitudinal, lateral, and vertical averaged positional accuracies for all three glideslope approaches are as follows (mean +/- two standard deviations in meters): longitudinal (-0.03 +/- 1.41), lateral (-1.29 +/- 2.36), and vertical (-0.76 +/- 2.05).

Hoang, Ty↗
NASA NTRS1995

An automated method of tuning an attitude estimator

Attitude determination is a major element of the operation and maintenance of a spacecraft. There are several existing methods of determining the attitude of a spacecraft. One of the most commonly used methods utilizes the Kalman filter to estimate the attitude of the spacecraft. Given an accurate model of a system and adequate observations, a Kalman filter can produce accurate estimates of the attitude. If the system model, filter parameters, or observations are inaccurate, the attitude estimates may be degraded. Therefore, it is advantageous to develop a method of automatically tuning the Kalman filter to produce the accurate estimates. In this paper, a three-axis attitude determination Kalman filter, which uses only magnetometer measurements, is developed and tested using real data. The appropriate filter parameters are found via the Process Noise Covariance Estimator (PNCE). The PNCE provides an optimal criterion for determining the best filter parameters.

Mason, Paul A. C.↗
NASA NTRS1981

Sensor failure detection system

Advanced concepts for detecting, isolating, and accommodating sensor failures were studied to determine their applicability to the gas turbine control problem. Five concepts were formulated based upon such techniques as Kalman filters and a screening process led to the selection of one advanced concept for further evaluation. The selected advanced concept uses a Kalman filter to generate residuals, a weighted sum square residuals technique to detect soft failures, likelihood ratio testing of a bank of Kalman filters for isolation, and reconfiguring of the normal mode Kalman filter by eliminating the failed input to accommodate the failure. The advanced concept was compared to a baseline parameter synthesis technique. The advanced concept was shown to be a viable concept for detecting, isolating, and accommodating sensor failures for the gas turbine applications.

Beattie, E. C.↗
NASA NTRS1981

Geomagnetic modeling by optimal recursive filtering

The results of a preliminary study to determine the feasibility of using Kalman filter techniques for geomagnetic field modeling are given. Specifically, five separate field models were computed using observatory annual means, satellite, survey and airborne data for the years 1950 to 1976. Each of the individual field models used approximately five years of data. These five models were combined using a recursive information filter (a Kalman filter written in terms of information matrices rather than covariance matrices.) The resulting estimate of the geomagnetic field and its secular variation was propogated four years past the data to the time of the MAGSAT data. The accuracy with which this field model matched the MAGSAT data was evaluated by comparisons with predictions from other pre-MAGSAT field models. The field estimate obtained by recursive estimation was found to be superior to all other models.

Gibbs, B. P.↗
NASA NTRS2004

CDGPS-Based Relative Navigation for Multiple Spacecraft

This thesis investigates the use of Carrier-phase Differential GPS (CDGPS) in relative navigation filters for formation flying spacecraft. This work analyzes the relationship between the Extended Kalman Filter (EKF) design parameters and the resulting estimation accuracies, and in particular, the effect of the process and measurement noises on the semimajor axis error. This analysis clearly demonstrates that CDGPS-based relative navigation Kalman filters yield good estimation performance without satisfying the strong correlation property that previous work had associated with "good" navigation filters. Several examples are presented to show that the Kalman filter can be forced to create solutions with stronger correlations, but these always result in larger semimajor axis errors. These linear and nonlinear simulations also demonstrated the crucial role of the process noise in determining the semimajor axis knowledge. More sophisticated nonlinear models were included to reduce the propagation error in the estimator, but for long time steps and large separations, the EKF, which only uses a linearized covariance propagation, yielded very poor performance. In contrast, the CDGPS-based Unscented Kalman relative navigation Filter (UKF) handled the dynamic and measurement nonlinearities much better and yielded far superior performance than the EKF. The UKF produced good estimates for scenarios with long baselines and time steps for which the EKF would diverge rapidly. A hardware-in-the-loop testbed that is compatible with the Spirent Simulator at NASA GSFC was developed to provide a very flexible and robust capability for demonstrating CDGPS technologies in closed-loop. This extended previous work to implement the decentralized relative navigation algorithms in real time.

Mitchell, Megan Leigh↗
NASA NTRS1972

A biased filter for linear discrete dynamic systems.

A recursive estimator, the ridge filter, was developed for the linear discrete dynamic estimation problem. Theorems were established to show that the ridge filter can be, on the average, closer to the expected value of the system state than the Kalman filter. On the other hand, Kalman filter, on the average, is closer to the instantaneous system state than the ridge filter. The ridge filter has been formulated in such a way that the computational features of the Kalman filter are preserved.

Chang, J. W.↗
NASA NTRS1994

Advanced data assimilation in strongly nonlinear dynamical systems

Advanced data assimilation methods are applied to simple but highly nonlinear problems. The dynamical systems studied here are the stochastically forced double well and the Lorenz model. In both systems, linear approximation of the dynamics about the critical points near which regime transitions occur is not always sufficient to track their occurrence or nonoccurrence. Straightforward application of the extended Kalman filter yields mixed results. The ability of the extended Kalman filter to track transitions of the double-well system from one stable critical point to the other depends on the frequency and accuracy of the observations relative to the mean-square amplitude of the stochastic forcing. The ability of the filter to track the chaotic trajectories of the Lorenz model is limited to short times, as is the ability of strong-constraint variational methods. Examples are given to illustrate the difficulties involved, and qualitative explanations for these difficulties are provided. Three generalizations of the extended Kalman filter are described. The first is based on inspection of the innovation sequence, that is, the successive differences between observations and forecasts; it works very well for the double-well problem. The second, an extension to fourth-order moments, yields excellent results for the Lorenz model but will be unwieldy when applied to models with high-dimensional state spaces. A third, more practical method--based on an empirical statistical model derived from a Monte Carlo simulation--is formulated, and shown to work very well. Weak-constraint methods can be made to perform satisfactorily in the context of these simple models, but such methods do not seem to generalize easily to practical models of the atmosphere and ocean. In particular, it is shown that the equations derived in the weak variational formulation are difficult to solve conveniently for large systems.

Miller, Robert N.↗