Engineering Papers⌕ Search

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.

At least 1,171 records · Page 65

Constrained Block Nonlinear Neural Dynamical Models

Neural network modules conditioned by known priors can be effectively trained and combined to represent systems with nonlinear dynamics. This work explores a novel formulation for data-efficient learning of deep control-oriented nonlinear dynamical models by embedding local model structure and constraints. The proposed method consists of neural network blocks that represent input, state, and output dynamics with constraints placed on the network weights and system variables. For handling partially observable dynamical systems, we utilize a state observer neural network to estimate the states of the system's latent dynamics. We evaluate the performance of the proposed architecture and training methods on system identification tasks for three nonlinear systems: a continuous stirred tank reactor, a two tank interacting system, and an aerodynamics body. Models optimized with a few thousand system state observations accurately represent system dynamics in open loop simulation over thousands of time steps from a single set of initial conditions. Experimental results demonstrate an order of magnitude reduction in open-loop simulation mean squared error for our constrained, block-structured neural models when compared to traditional unstructured and unconstrained neural network models.

Skomski, Elliott↗

Estimation of steady-state and transcient power distributions for the RELAP analyses of the 1963 loss-of-flow and loss-of-pressure tests at BR2.

To support the safety analyses required for the conversion of the Belgian Reactor 2 (BR2) from highly-enriched uranium (HEU) to low-enriched uranium (LEU) fuel, the simulation of a number of loss-of-flow tests, with or without loss of pressure, has been undertaken. These tests were performed at BR2 in 1963 and used instrumented fuel assemblies (FAs) with thermocouples (TC) imbedded in the cladding as well as probes to measure the FAs power on the basis of their coolant temperature rise. The availability of experimental data for these tests offers an opportunity to better establish the credibility of the RELAP5-3D model and methodology used in the conversion analysis. In order to support the HEU to LEU conversion safety analyses of the BR2 reactor, RELAP simulations of a number of loss-of-flow/loss-of-pressure tests have been undertaken. Preliminary analyses showed that the conservative power distributions used historically in the BR2 RELAP model resulted in a significant overestimation of the peak cladding temperature during the transient. Therefore, it was concluded that better estimates of the steady-state and decay power distributions were needed to accurately predict the cladding temperatures measured during the tests and establish the credibility of the RELAP model and methodology. The new approach ('best estimate' methodology) uses the MCNP5, ORIGEN-2 and BERYL codes to obtain steady-state and decay power distributions for the BR2 core during the tests A/400/1, C/600/3 and F/400/1. This methodology can be easily extended to simulate any BR2 core configuration. Comparisons with measured peak cladding temperatures showed a much better agreement when power distributions obtained with the new methodology are used.

21 SPECIFIC NUCLEAR REACTORS AND ASSOCIATED PLANTS↗

MITgcm-AD v2: Open source tangent linear and adjoint modeling framework for the oceans and atmosphere enabled by the Automatic Differentiation tool Tapenade

The Massachusetts Institute of Technology General Circulation Model (MITgcm) is widely used by the climate science community to simulate planetary atmosphere and ocean circulations. A defining feature of the MITgcm is that it has been developed to be compatible with an algorithmic differentiation (AD) tool, TAF, enabling the generation of tangent-linear and adjoint models. These provide gradient information which enables dynamics-based sensitivity and attribution studies, state and parameter estimation, and rigorous uncertainty quantification. Importantly, gradient information is essential for computing comprehensive sensitivities and performing efficient large-scale data assimilation, ensuring that observations collected from satellites and in-situ measuring instruments can be effectively used to optimize a large uncertain control space. As a result, the MITgcm forms the dynamical core of a key data assimilation product employed by the physical oceanography research community: Estimating the Circulation and Climate of the Ocean (ECCO) state estimate. Although MITgcm and ECCO are used extensively within the research community, the AD tool TAF is proprietary and hence inaccessible to a large proportion of these users. The new version 2 (MITgcm-AD v2) framework introduced here is based on the source-to-source AD tool Tapenade, which has recently been open-sourced. Another feature of Tapenade is that it stores required variables by default (instead of recomputing them) which simplifies the implementation of efficient, AD-compatible code. The framework has been integrated with the MITgcm model’s main branch and is now freely available.

Adjoints↗

Orbit determination in the presence of unmodeled accelerations.

In the investigation discussed here, the unmodeled acceleration is approximated as a first-order stationary Gauss-Markov process which can be separated into a time wise correlated component and a purely random component. An extended form of the Kalman-Bucy filter is used to estimate the state and the unmodeled accelerations at each observation epoch. Applications of the estimation algorithm to simulated data and to tracking data obtained during the Apollo 10 and 11 missions indicate that a more accurate estimate of the state is obtained using the algorithm and that the estimated values of the unmodeled accelerations can be correlated with physical phenomena neglected by the mathematical model.

Tapley, B. D.↗

State-space self-tuning controllers for general multivariable stochastic systems

This paper presents a state-space approach for self-tuning control of a more general class of multivariable stochastic systems having a number of inputs equal or different from the number of outputs. The dynamic system is represented in the state-space innovation form with Luenberger's canonical structures. The model parameters and the Kalman gain are identified via either the extended least-squares algorithm or the least-squares ladder algorithm. The Kalman gain matrix and states can be estimated from the identified parameters without utilizing the standard state estimation algorithm. A long division method is introduced for finding the similarity transformation matrix.

Shieh, L. S.↗

Battery state of health monitoring by estimation of the number of cyclable Li-ions

Herein this paper introduces a method to monitor battery state of health (SOH) by estimating the number of cyclable Li-ions, a health-relevant electrochemical variable. SOH monitoring is critical to battery management in balancing the trade-off between maximizing system performance and minimizing battery degradation. The decrease of cyclable Li-ions indicates the effect on the SOH of degradation mechanisms that consume cyclable Li-ions. The unavailability of the number of cyclable Li-ions through non-invasive measurements makes its estimation necessary for in-situ SOH monitoring. In this paper, the extended Kalman filter (EKF) is used to estimate the number of cyclable Li-ions as an unknown battery parameter. The single particle model (SPM), a simplified battery electrochemical model, is used as the model in the EKF to achieve a computational complexity suitable for on-line estimation. Simulations are performed under typical electric vehicle current trajectories using an example parameter set for a hybrid-electric-vehicle battery. In the simulations, the battery is represented by the Doyle–Fuller–Newman (DFN) model, an electrochemical model with higher fidelity than the SPM. To comply with the practice, instead of using the same parameters as the DFN model in the SPM, parameterization of the SPM is performed before estimation of the number of cyclable Li-ions. The simulations show high estimation accuracy of the number of cyclable Li-ions using the EKF, even with the structural and parametric differences between the DFN model and the SPM, under both the ideal conditions and various non-ideal conditions (i.e., SOC estimation error, additional modeling error, and measurement noise).

25 ENERGY STORAGE↗

Digital flight control systems

The design of stable feedback control laws for sampled-data systems with variable rate sampling was investigated. These types of sampled-data systems arise naturally in digital flight control systems which use digital actuators where it is desirable to decrease the number of control computer output commands in order to save wear and tear of the associated equipment. The design of aircraft control systems which are optimally tolerant of sensor and actuator failures was also studied. Detection of the failed sensor or actuator must be resolved and if the estimate of the state is used in the control law, then it is also desirable to have an estimator which will give the optimal state estimate even under the failed conditions.

Caglayan, A. K.↗

Recent results of nonlinear estimators applied to hereditary systems.

An application of the extended Kalman filter to delayed systems to estimate the state and time delay is presented. Two nonlinear estimators are discussed and the results compared with those of the Kalman filter. For all the filters considered, the hereditary system was treated with the delay in the pure form and by using Pade approximations of the delay. A summary of the convergence properties of the filters studied is given. The results indicate that the linear filter applied to the delayed system performs inadequately while the nonlinear filters provide reasonable estimates of both the state and the parameters.

Schiess, J. R.↗

High-Performance Computing for Advanced Smart Grid Applications

The power grid is becoming far more complex as a result of the grid evolution meeting an information revolution. Due to the penetration of smart grid technologies, the grid is evolving as an unprecedented speed and the information infrastructure is fundamentally improved with a large number of smart meters and sensors that produce several orders of magnitude larger amounts of data. How to pull data in, perform analysis, and put information out in a real-time manner is a fundamental challenge in smart grid operation and planning. The future power grid requires high performance computing to be one of the foundational technologies in developing the algorithms and tools for the significantly increased complexity. New techniques and computational capabilities are required to meet the demands for higher reliability and better asset utilization, including advanced algorithms and computing hardware for large-scale modeling, simulation, and analysis. This chapter summarizes the computational challenges in smart grid and the need for high performance computing, and present examples of how high performance computing might be used for future smart grid operation and planning.

computational challenges↗

Enhanced Estimation of Quantum Properties with Common Randomized Measurements

We present a technique for enhancing the estimation of quantum state properties by incorporating approximate prior knowledge about the quantum state. This consists in performing randomized measurements on a quantum processor and comparing the results with those obtained from a classical computer that stores an approximation of the quantum state. We provide unbiased estimators for expectation values of multicopy observables and present performance guarantees in terms of variance bounds that depend on the prior knowledge accuracy. We demonstrate the effectiveness of our approach through experimental and numerical examples detecting mixed-state entanglement, and estimating polynomial approximations of the von Neumann entropy and state fidelities. Published by the American Physical Society 2024

Physics↗

Estimation of time- and state-dependent delays and other parameters in functional differential equations

A parameter estimation algorithm is developed which can be used to estimate unknown time- or state-dependent delays and other parameters (e.g., initial condition) appearing within a nonlinear nonautonomous functional differential equation. The original infinite dimensional differential equation is approximated using linear splines, which are allowed to move with the variable delay. The variable delays are approximated using linear splines as well. The approximation scheme produces a system of ordinary differential equations with nice computational properties. The unknown parameters are estimated within the approximating systems by minimizing a least-squares fit-to-data criterion. Convergence theorems are proved for time-dependent delays and state-dependent delays within two classes, which say essentially that fitting the data by using approximations will, in the limit, provide a fit to the data using the original system. Numerical test examples are presented which illustrate the method for all types of delay.

Murphy, K. A.↗

Estimation of time- and state-dependent delays and other parameters in functional differential equations

A parameter estimation algorithm is developed which can be used to estimate unknown time- or state-dependent delays and other parameters (e.g., initial condition) appearing within a nonlinear nonautonomous functional differential equation. The original infinite dimensional differential equation is approximated using linear splines, which are allowed to move with the variable delay. The variable delays are approximated using linear splines as well. The approximation scheme produces a system of ordinary differential equations with nice computational properties. The unknown parameters are estimated within the approximating systems by minimizing a least-squares fit-to-data criterion. Convergence theorems are proved for time-dependent delays and state-dependent delays within two classes, which say essentially that fitting the data by using approximations will, in the limit, provide a fit to the data using the original system. Numerical test examples are presented which illustrate the method for all types of delay.

Murphy, K. A.↗

Application of the Two-Step Filter to Process Ranging Measurements for Relative Navigation in an Elliptical Orbit

This estimator breaks a nonlinear estimation problem into a set of over determined 'first step' states which are linear in the observations and 'second step' states which are ultimately the states of interest. Linear estimation methods are applied to filter the observations and produce the optimal first step state estimate. The 'second step' states are obtained through iterative nonlinear parameter estimation considering the first step states as observations. It has been shown that this process exactly minimizes the least squares cost function for static problems and provides a better solution than the iterated extended Kalman filter (EKF) for dynamic problems. The two step filter is applied in this paper to process range and range rate measurements between the two spacecraft. Details of the application of the two step estimator to this problem will be given, highlighting the use of a test for ill-conditioned covariance estimates that can result from the first order covariance propagation. A comparison will be made between the performance of the two step filter and the IEKF.

Garrison, James L.↗

Estimating coccidioidomycosis endemicity while accounting for imperfect detection using spatio - temporal occupancy modeling

Coccidioidomycosis, or Valley fever, is an infectious disease caused by inhaling Coccidioides fungal spores. Incidence has risen in recent years, and it is believed the endemic region for Coccidioides is expanding in response to climate change. While Valley fever case data can help us understand trends in disease risk, using case data as a proxy for Coccidioides endemicity is not ideal because case data suffers from imperfect detection, including false positives (e.g., travel-related cases reported outside of endemic area) and false negatives (e.g., misdiagnosis or underreporting). Here we proposed a Bayesian, spatio-temporal occupancy model to relate monthly, county-level presence/absence data on Valley fever cases to latent endemicity of Coccidioides, accounting for imperfect detection. We used our model to estimate endemicity in the western United States. We estimated high probability of endemicity in southern California, Arizona, and New Mexico, but also in regions without mandated reporting, including western Texas, eastern Colorado, and southeastern Washington. We also quantified spatio-temporal variability in detectability of Valley fever, given an area is endemic to Coccidioides. We estimated an inverse relationship between lagged 3- and 9-month precipitation and case detection, and a positive association with agriculture. This work can help inform public health surveillance needs and identify areas that would benefit from mandatory case reporting.

60 APPLIED LIFE SCIENCES↗

Extended State Observer-Based Robust Model Predictive Velocity Control for Permanent Magnet Synchronous Motor

This article proposes an extended state observer based robust model predictive velocity control to decrease system prediction error under parameter uncertainties for permanent magnet synchronous motor (PMSM). We develop a new PMSM model that consists of velocity and acceleration to lump the system information and an external disturbance into a disturbance. The extended state observer (ESO) is designed to estimate the velocity, acceleration, and disturbance. By estimating the state variables and disturbance using the ESO, the model predictive control (MPC) finds the optimal control input by predicting future system behavior. Additionally, the direct current controller is designed so that the direct current converges to zero. Because the proposed method is not designed based on the cascade structure from the viewpoint of velocity control, the optimization control for the velocity and currents can be defined. Thus, the proposed method is robust against external disturbances and parameter uncertainties owing to feedback linearization, state feedback, and ESO-based MPC using the acceleration PMSM model. The proposed control algorithm was experimentally verified and it showed improved velocity tracking performance compared with ESO-based MPC using the conventional PMSM model.

42 ENGINEERING↗

UQ Methods for HPDA and Cybersecurity Models, Data, and Use Cases

This report describes our initial research to quantify uncertainties in the identification and characterization of possible attack states in a network. As a result, we should be able to estimate the current state in which the network is operating, based on a wide variety of network data, and attach a defensible measure of confidence to these state estimates. The output of this research will be new uncertainty quantification (UQ) methods to help develop a process for model development and apply UQ to characterize attacks/adversaries, create an understanding of the degree to which methods scale to "big" data, and offer methods for addressing model approaches with regard to validation and accuracy.

97 MATHEMATICS AND COMPUTING↗

Data acquisition and path selection decision making for an autonomous roving vehicle

Problems related to the guidance of an autonomous rover for unmanned planetary exploration were investigated. Topics included in these studies were: simulation on an interactive graphics computer system of the Rapid Estimation Technique for detection of discrete obstacles; incorporation of a simultaneous Bayesian estimate of states and inputs in the Rapid Estimation Scheme; development of methods for estimating actual laser rangefinder errors and their application to date provided by Jet Propulsion Laboratory; and modification of a path selection system simulation computer code for evaluation of a hazard detection system based on laser rangefinder data.

Frederick, D. K.↗