Estimation of state with acceptable accuracy constraints Interim report
Minimum variance method for estimating state vector with acceptable accuracy constraints in orbit calculation
VARIANCE↗
SEARCH · Engineering Papers
Results for “State estimation”
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.
On-Orbit Multi-Field Wavefront Control with a Kalman Filter
A document describes a multi-field wavefront control (WFC) procedure for the James Webb Space Telescope (JWST) on-orbit optical telescope element (OTE) fine-phasing using wavefront measurements at the NIRCam pupil. The control is applied to JWST primary mirror (PM) segments and secondary mirror (SM) simultaneously with a carefully selected ordering. Through computer simulations, the multi-field WFC procedure shows that it can reduce the initial system wavefront error (WFE), as caused by random initial system misalignments within the JWST fine-phasing error budget, from a few dozen micrometers to below 50 nm across the entire NIRCam Field of View, and the WFC procedure is also computationally stable as the Monte-Carlo simulations indicate. With the incorporation of a Kalman Filter (KF) as an optical state estimator into the WFC process, the robustness of the JWST OTE alignment process can be further improved. In the presence of some large optical misalignments, the Kalman state estimator can provide a reasonable estimate of the optical state, especially for those degrees of freedom that have a significant impact on the system WFE. The state estimate allows for a few corrections to the optical state to push the system towards its nominal state, and the result is that a large part of the WFE can be eliminated in this step. When the multi-field WFC procedure is applied after Kalman state estimate and correction, the stability of fine-phasing control is much more certain. Kalman Filter has been successfully applied to diverse applications as a robust and optimal state estimator. In the context of space-based optical system alignment based on wavefront measurements, a KF state estimator can combine all available wavefront measurements, past and present, as well as measurement and actuation error statistics to generate a Maximum-Likelihood optimal state estimator. The strength and flexibility of the KF algorithm make it attractive for use in real-time optical system alignment when WFC alone cannot effectively align the system.
Lou, John↗
Alternative smoothing algorithms for on-line estimation problems
In recent years, much emphasis has been placed on the Kalman filter to solve on-line estimation problems. Smoothers, which determine state estimates at a certain time t using measurements between 0 and T, have been used primarily in off-line estimation schemes. This paper presents alternative forms of smoothers that can be used for on-line estimation problems. The alternative smoothing algorithms provide state estimates which are more refined than Kalman filtered state estimates. The alternative smoothing algorithms are illustrated in two examples. The first example considers a linear discrete-time system while the second example considers a nonlinear continuous-time system with discrete-time measurements.
Baumgartner, Eric T.↗
Deep Interacting Multiple Model Filtering
In this paper, a deep learning-based multiple model estimation framework is presented for the state estimation of hybrid dynamical systems from high dimensional observations such as camera images. A low dimensional vector which represents the measurement of the latent dynamical system and its corresponding variance are learned using a deep encoder neural network. An Interacting Multiple Model (IMM) filter is used to generate the latent state estimates and covariances using multiple dynamical models, which can be learned using backpropagation through time. The state estimates of the dynamical system and the corresponding covariance matrix are generated from the latent state estimates and covariance using a deep decoder neural network. The whole network is trained in an end-to-end manner using a loss function which minimizes the negative log-likelihood of the neural network parameters. Simulation results are presented using a 2D bouncing ball example and estimation error statistics are computed which demonstrates the accuracy and consistency of the estimation.
Ghananeel Rotithor↗
Towards Resilient Autonomous Navigation of Drones
Robots and particularly drones are especially useful in exploring extreme environments that pose hazards to humans. To ensure safe operations in these situations, usually perceptually degraded and without good GNSS, it is critical to have a reliable and robust state estimation solution. The main body of literature in robot state estimation focuses on developing complex algorithms favoring accuracy. Typically, these approaches rely on a strong underlying assumption: the main estimation engine will not fail during operation. In contrast, we propose an architecture that pursues robustness in state estimation by considering redundancy and heterogeneity in both sensing and estimation algorithms. The architecture is designed to expect and detect failures and adapt the behavior of the system to ensure safety. To this end, we present HeRO (Heterogeneous Redundant Odometry): a stack of estimation algorithms running in parallel supervised by a resiliency logic. This logic carries out three main functions: a) perform confidence tests both in data quality and algorithm health; b) re-initialize those algorithms that might be malfunctioning; c) generate a smooth state estimate by multiplexing the inputs based on their quality. The state and quality estimates are used by the guidance and control modules to adapt the mobility behaviors of the system. The validation and utility of the approach are shown with real experiments on a ying robot for the use case of autonomous exploration of subterranean environments, with particular results from the STIX event of the DARPA Subterranean Challenge.
Agha-mohammadi, Ali-akbar↗
Velocimeter LIDAR Based Bulk Velocity Estimation for Terrain Relative Navigation Applications
A novel batch state estimation approach to estimate the translational and angular velocity states of a vehicle using a state-of-the-art velocimeter Light Detection and Ranging (LIDAR) sensor for use in Terrain and Hazard Relative Navigation (TRN/HRN) applications is presented in this paper. The velocimeter LIDAR is capable of measuring three dimensional position and line-of-sight (LOS) velocity associated with every pixel in the field-of-view (FOV). This new batch state estimation methodology is shown to provide accurate and statistically consistent velocity estimates with no a priori information. In contrast to traditional computer vision approaches for TRN, the proposed technique is not dependent on image features. This feature alleviates the need for accurate feature detection and correspondence to a predefined map making it suitable for unknown operating environments. Following a detailed development of the mathematical details associated with the batch state estimation methodology, the efficacy and utility of the proposed algorithms are evaluated through emulation robotics experiments performed at Texas A&M’s Land, Air, and Space Robotics (LASR) laboratory.
Skulsky, Eli↗
Autonomous Component Health Management with Failed Component Detection, Identification, and Avoidance
This paper details a novel scheme for autonomous component health management (ACHM) with failed actuator detection and failed sensor detection, identification, and avoidance. This new scheme has features that far exceed the performance of systems with triple-redundant sensing and voting, yet requires fewer sensors and could be applied to any system with redundant sensing. Relevant background to the ACHM scheme is provided, and the simulation results for the application of that scheme to a single-axis spacecraft attitude control system with a 3rd order plant and dual-redundant measurement of system states are presented. ACHM fulfills key functions needed by an integrated vehicle health monitoring (IVHM) system. It is: autonomous; adaptive; works in realtime; provides optimal state estimation; identifies failed components; avoids failed components; reconfigures for multiple failures; reconfigures for intermittent failures; works for hard-over, soft, and zero-output failures; and works for both open- and closed-loop systems. The ACHM scheme combines a prefilter that generates preliminary state estimates, detects and identifies failed sensors and actuators, and avoids the use of failed sensors in state estimation with a fixed-gain Kalman filter that generates optimal state estimates and provides model-based state estimates that comprise an integral part of the failure detection logic. The results show that ACHM successfully isolates multiple persistent and intermittent hard-over, soft, and zero-output failures. It is now ready to be tested on a computer model of an actual system.
Davis, Robert N.↗
Adaptive Filtering Using Recurrent Neural Networks
A method for adaptive (or, optionally, nonadaptive) filtering has been developed for estimating the states of complex process systems (e.g., chemical plants, factories, or manufacturing processes at some level of abstraction) from time series of measurements of system inputs and outputs. The method is based partly on the fundamental principles of the Kalman filter and partly on the use of recurrent neural networks. The standard Kalman filter involves an assumption of linearity of the mathematical model used to describe a process system. The extended Kalman filter accommodates a nonlinear process model but still requires linearization about the state estimate. Both the standard and extended Kalman filters involve the often unrealistic assumption that process and measurement noise are zero-mean, Gaussian, and white. In contrast, the present method does not involve any assumptions of linearity of process models or of the nature of process noise; on the contrary, few (if any) assumptions are made about process models, noise models, or the parameters of such models. In this regard, the method can be characterized as one of nonlinear, nonparametric filtering. The method exploits the unique ability of neural networks to approximate nonlinear functions. In a given case, the process model is limited mainly by limitations of the approximation ability of the neural networks chosen for that case. Moreover, despite the lack of assumptions regarding process noise, the method yields minimum- variance filters. In that they do not require statistical models of noise, the neural- network-based state filters of this method are comparable to conventional nonlinear least-squares estimators.
Parlos, Alexander G.↗
On-line parameter estimation using a high sensitivity estimator
An on-line parameter identification method is presented. The method is based on a recursive formulation of the maximum likelihood method, with a significant modification on the gains of the state estimator. In the conventional maximum likelihood method, the Kalman gains are used in the state estimator. This produces unbiased, minimum variance parameter estimates in the presence of process noise and measurement noise, but it also slows the convergence rate when the identification is done on-line. Here we suggest choosing the gains to maximize a measure of the sensitivities of the state estimates to parameter variations. One such criterion is to minimize the trace of the inverse information matrix. This increases the convergence rate significantly. After one or two time constants, the gains are switched to the Kalman values to assure unbiased, minimum-variance estimates. The state estimate will initially be nonoptimal, and may not be adequate for control purposes. In this case, a parallel Kalman filter which uses the identifier's parameter estimates can be used. This method is applied here for the identification of a simple first-order system, and for the identification of short-period stability derivatives of an F-8 aircraft from simulated data.
Mishne, D.↗
Estimating aircraft states
Computer program provides weighted least-squares estimates of aircraft states from measurements recorded during routine flight tests. Program contains standard six-degree-of-freedom kinematic equations.
Wingrove, R. C.↗
Identification of rotating assembly inertial and bearing parameters.
The problem of estimating the state of a turbopump rotating assembly and identifying unknown products of inertia and unknown bearing parameters is considered. A linearized, extended Kalman filtering approach has been used and found to be successful for both state estimation and parameter identification in an inherently nonlinear problem. The input data were simulated Bently test data obtained from a verified 12-dimensional state-space model and were compared with real Bently test data obtained from NASA. The results prove the feasibility of using this model and this type of test data to obtain the hidden parameters of a typical turbopump rotating assembly.
Alspach, D. L.↗
A program to form a multidisciplinary data base and analysis for dynamic systems
Diverse sets of experimental data and analysis programs have been assembled for the purpose of facilitating research in systems identification, parameter estimation and state estimation techniques. The data base analysis programs are organized to make it easy to compare alternative approaches. Additional data and alternative forms of analysis will be included as they become available.
Taylor, L. W.↗
Self-Tuning Adaptive-Controller Using Online Frequency Identification
A real time adaptive controller was designed and tested successfully on a fourth order laboratory dynamic system which features very low structural damping and a noncolocated actuator sensor pair. The controller, implemented in a digital minicomputer, consists of a state estimator, a set of state feedback gains, and a frequency locked loop (FLL) for real time parameter identification. The FLL can detect the closed loop natural frequency of the system being controlled, calculate the mismatch between a plant parameter and its counterpart in the state estimator, and correct the estimator parameter in real time. The adaptation algorithm can correct the controller error and stabilize the system for more than 50% variation in the plant natural frequency, compared with a 10% stability margin in frequency variation for a fixed gain controller having the same performance at the nominal plant condition. After it has locked to the correct plant frequency, the adaptive controller works as well as the fixed gain controller does when there is no parameter mismatch. The very rapid convergence of this adaptive system is demonstrated experimentally, and can also be proven with simple root locus methods.
Chiang, W. W.↗
Kalman filtering, smoothing, and recursive robot arm forward and inverse dynamics
The recursive difference equations of Kalman filtering and Bryson-Frazier fixed time-interval smoothing, arising in the state estimation theory for linear state space systems, are used here to solve problems of serial manipulator inverse and forward dynamics. The configuration analyzed is that of a joint connected N-link serial manipulator attached to an immobile base. The equivalence between the filtering and smoothing techniques from state estimation theory and recursive robot dynamics methods is demonstrated. Several areas for future research are suggested.
Rodriguez, Guillermo↗
Improving Ocean Angular Momentum Estimates Using a Model Constrained by Data
Ocean angular momentum (OAM) calculations using forward model runs without any data constraints have, recently revealed the effects of OAM variability on the Earth's rotation. Here we use an ocean model and its adjoint to estimate OAM values by constraining the model to available oceanic data. The optimization procedure yields substantial changes in OAM, related to adjustments in both motion and mass fields, as well as in the wind stress torques acting on the ocean. Constrained and unconstrained OAM values are discussed in the context of closing the planet's angular momentum budget. The estimation procedure, yields noticeable improvements in the agreement with the observed Earth rotation parameters, particularly at the seasonal timescale. The comparison with Earth rotation measurements provides an independent consistency check on the estimated ocean state and underlines the importance of ocean state estimation for quantitative. studies of the variable large-scale oceanic mass and circulation fields, including studies of OAM.
Ponte, Rui M.↗
An Empirical State Error Covariance Matrix for the Weighted Least Squares Estimation Method
State estimation techniques effectively provide mean state estimates. However, the theoretical state error covariance matrices provided as part of these techniques often suffer from a lack of confidence in their ability to describe the uncertainty in the estimated states. By a reinterpretation of the equations involved in the weighted least squares algorithm, it is possible to directly arrive at an empirical state error covariance matrix. This proposed empirical state error covariance matrix will contain the effect of all error sources, known or not. Results based on the proposed technique will be presented for a simple, two observer, measurement error only problem.
Joseph Henry Frisbee↗
An Empirical State Error Covariance Matrix for the Weighted Least Squares Estimation Method
State estimation techniques effectively provide mean state estimates. However, the theoretical state error covariance matrices provided as part of these techniques often suffer from a lack of confidence in their ability to describe the un-certainty in the estimated states. By a reinterpretation of the equations involved in the weighted least squares algorithm, it is possible to directly arrive at an empirical state error covariance matrix. This proposed empirical state error covariance matrix will contain the effect of all error sources, known or not. Results based on the proposed technique will be presented for a simple, two observer, measurement error only problem.
Frisbee, Joseph H., Jr.↗
Optimal post-experiment estimation of poorly modeled dynamic systems
Recently, a novel strategy for post-experiment state estimation of discretely-measured dynamic systems has been developed. The method accounts for errors in the system dynamic model equations in a more general and rigorous manner than do filter-smoother algorithms. The dynamic model error terms do not require the usual process noise assumptions of zero-mean, symmetrically distributed random disturbances. Instead, the model error terms require no prior assumptions other than piecewise continuity. The resulting state estimates are more accurate than filters for applications in which the dynamic model error clearly violates the typical process noise assumptions, and the available measurements are sparse and/or noisy. Estimates of the dynamic model error, in addition to the states, are obtained as part of the solution of a two-point boundary value problem, and may be exploited for numerous reasons. In this paper, the basic technique is explained, and several example applications are given. Included among the examples are both state estimation and exploitation of the model error estimates.
Mook, D. Joseph↗