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.
This paper considers the suboptimal stochastic control of linear discrete-time dynamical systems with uncertain or stochastically varying parameters. The suboptimal scheme is based upon the use of the open-loop-feedback-optimal (OLFO) method. The state and parameter estimates are generated by an extended Kalman filter algorithm. Numerical results for first-order systems are presented.
The problem of designing optimal inputs in the identification of linear systems with unknown random parameters is considered using a Bayesian approach. The information matrix, which is positive definite for the class of systems analyzed, gives a measure of performance for the system inputs. The computation of the optimal closed-loop input mappings is shown to be a nontrivial exercise in adaptive control. Deterministic optimal inputs are shown to be easily computable. Numerical examples are given. A Kalman filter is used to estimate the parameters. A necessary condition for the Kalman filter not to diverge when applying linear feedback is also given.
We define a class of detection-estimation problems on matrix Lie groups in which the observation noise is multiplicative in nature. By examining the differential versions of the hypotheses, which are bilinear, we are able to derive the relevant likelihood ratio formula and the associated optimal estimation equations for the signal given the observations and the assumption that the signal is present. These estimation equations are of interest in their own right, in that they represent a finite-dimensional optimal solution to a nonlinear estimation problem and consist of a Kalman-Bucy filter along with the on-line computation of the solution of the associated Riccati equation, which is driven by the observations. The usefulness of these results is illustrated via an example concerning the detection of an actuator failure in a rigid-body rotational control system.
An identification algorithm that uses the maximum likelihood technique to identify the unknown time delays, plant parameters, and noise covariances of linear discrete stochastic systems is presented. Cases of additive white noise and colored measurement noises are considered. The likelihood function is evaluated using either a minimum-variance (Kalman) filter or a minimal-order observer. The Kalman filter is used in the identification algorithm to provide minimum-variance estimates. The minimal-order observer is a lower-dimensional and computationally simpler filter, and is advantageous especially for systems with long delays. It provides a less optimal solution to the minimum-mean-square state estimation problem. The colored-noise observer algorithm has the disadvantage of having to compute an extra error covariance matrix of lower order.
This paper presents a scheme for on-line failure detection in systems with appreciable structural dynamics. The design is suboptimal because of extensive computational requirements of optimal schemes. To accomplish failure detection, the innovations sequence of a finite order Kalman filter is examined. Because of the heavy dependence of the system on the zero-mean character of the innovations sequence of the filter much attention has been given to the design and evaluation of the Kalman filters used. The filter designs are based on modal models of the structural dynamics. Two modal models were considered, one based on an analytic finite element model and the other based on empirically derived frequency and damping. Experiments using a grid structure are presented which illustrate operation and performance of the filter designs based on these models. The general character of the results presented is that appreciable errors exist in the filter design based on the finite element model. Substantial improvement results if the design model is modified to include empirically derived frequency and damping.
A dynamic force sensing method is proposed for a robot manipulator. A 6-axis wrist force sensor which is commonly used for robot force sensing does not take the direct measurement of the force and moment applied to an end effector externally from the environment of the manipulator, but takes the resultant of the gravity, the inertial force and moment on the end effector, and the external force and moment. The separation of gravitational and inertial force and moment from the measurement of the force sensor is crucial to achieving dynamically accurate force sensing and fast manipulator motion under force feedback control. A mathematical model is derived for the force sensing process and the separation of the force and moment as a nonlinear optimal state estimation problem formulated. Then an approximate solution is given using extended Kalman filtering techniques.
The behavior of forecast error covariances in a fairly realistic setting is demonstrated via a Kalman filter algorithm. It is used to assimilate simulated data from the existing radiosonde network, from the demonstration network of 31 Doppler wind profilers in the central U.S., and from hypothetical radiometers located at five of the profiler sites. Some theoretical justification of the hypothesis advanced by Phillips (1982), and the hypothesis is used to formulate the model error covariance matrix required by the Kalman filter. The results show that assimilating the profiler wind data leads to a large reduction of forecast/analysis error in heights as well as in winds, over the profiler region and also downstream, when compared with the results of assimilating the radiosonde data alone. The forecast error covariance matrices that the Kalman filter calculates to obtain this error reduction differ considerably from those prescribed by the optimal interpolation schemes that are employed for data assimilation at operational centers.
Performance of a 30 GHz frequency division multiple access (FDMA) uplink to a processing satellite is modelled for the case where the onboard demultiplexer is implemented optically. Included in the performance model are the effects of adjacent channel interference, intersymbol interference, and spurious signals associated with the optical implementation. Demultiplexer parameters are optimized to provide the minimum bit error probability at a given bandwidth efficiency when filtered QPSK modulation is employed.
The performance seeking control (PSC) algorithm is designed to continuously optimize the performance of propulsion systems. The PSC algorithm uses a nominal propulsion system model and estimates, in flight, the engine deviation parameters (EDPs) characterizing the engine deviations with respect to nominal conditions. In practice, because of measurement biases and/or model uncertainties, the estimated EDPs may not reflect the engine's actual off-nominal condition. This factor has a direct impact on the PSC scheme exacerbated by the open-loop character of the algorithm. In this paper, the effects produced by unknown measurement biases over the estimation algorithm are evaluated. This evaluation allows for identification of the most critical measurements for application of the PSC algorithm to an F100 engine. An equivalence relation between the biases and EDPs stems from the analysis; therefore, it is undecided whether the estimated EDPs represent the actual engine deviation or whether they simply reflect the measurement biases. A new algorithm, based on the engine's (steady-state) optimization model, is proposed and tested with flight data. When compared with previous Kalman filter schemes, based on local engine dynamic models, the new algorithm is easier to design and tune and it reduces the computational burden of the onboard computer.
The performance seeking control algorithm is designed to continuously optimize the performance of propulsion systems. The performance seeking control algorithm uses a nominal model of the propulsion system and estimates, in flight, the engine deviation parameters characterizing the engine deviations with respect to nominal conditions. In practice, because of measurement biases and/or model uncertainties, the estimated engine deviation parameters may not reflect the engine's actual off-nominal condition. This factor has a necessary impact on the overall performance seeking control scheme exacerbated by the open-loop character of the algorithm. The effects produced by unknown measurement biases over the estimation algorithm are evaluated. This evaluation allows for identification of the most critical measurements for application of the performance seeking control algorithm to an F100 engine. An equivalence relation between the biases and engine deviation parameters stems from an observability study; therefore, it is undecided whether the estimated engine deviation parameters represent the actual engine deviation or whether they simply reflect the measurement biases. A new algorithm, based on the engine's (steady-state) optimization model, is proposed and tested with flight data. When compared with previous Kalman filter schemes, based on local engine dynamic models, the new algorithm is easier to design and tune and it reduces the computational burden of the onboard computer.
This paper reviews the author’s previous work on free-running timescales based on Kalman filters that act upon clock comparisons. The natural Kalman clock ensemble algorithm tends to optimize long-term timescale stability at the expense of short-term stability. By subjecting each postmeasurement error covariance matrix to a non-transparent reduction operation, one obtains corrected clocks with improved short-term stability and little sacrifice of long-term stability. A new result on covariance matrix reduction is also stated.
This paper presents the design of a multi-objective optimal controller for the 13.5 aspect ratio Common Research Model (CRM) wind tunnel model. A Pareto front analysis is performed to provide the optimal tradeoff between the competing objectives - gust load alleviation, maneuver load alleviation, and drag minimization - so that the weighting matrices for the control law are selected optimally. Also presented are two formulations of the controller: a state feedback with Kalman filter design, and an output feedback form where sensors are used directly for feedback control. The controller demonstrates a reduced peak-to-peak strain of 45.78% while also reducing drag and mean bending moment. An output feedback formulation for gust load alleviation is formulated and produces a 53% reduction in the strain response due to a gust input.
The memory wall phenomenon—where advances in processor performance significantly outpace those in memory subsystems—poses a fundamental challenge for contemporary computing systems. In memory-bound applications, memory subsystem behavior dominates performance, yet existing analysis approaches present significant limitations: detailed microarchitectural simulators require days to weeks to simulate modest workloads; hardware performance counters provide only aggregate statistics that obscure temporal and spatial access patterns; and scaled simulation approaches face challenges in capturing certain behaviors that emerge at larger scales. These limitations reflect a processor-centric design philosophy increasingly misaligned with memory-bound workloads where detailed understanding of memory access patterns, cache hierarchy interactions, and contention is critical for effective optimization. This paper presents an integrated framework for memory-centric analysis that enables effective hardware-software co-design. We describe practical trace collection techniques, including hardware-assisted processor tracing with minimal overhead and portable software-based instrumentation with statistical sampling. We present multi-perspective analysis methods that examine memory behavior from temporal, sequential, spatial, and relational viewpoints, revealing distinct optimization opportunities invisible in aggregate metrics. We detail an architectural modeling framework that uses sampled traces with temporal interpolation and confidence-based filtering to evaluate cache and memory configurations. Evaluation on representative benchmarks demonstrates that this framework achieves practical accuracy (L2 cache errors of 2.64\%, confidence-filtered L3 errors of 9.92\%, bandwidth errors of 7.33\%) while providing substantial speedup (26.8×) over cycle-accurate simulation, enabling rapid design space exploration. We demonstrate how this integrated framework enables systematic identification of both hardware optimizations (memory controller tuning, bank partitioning, NUMA configuration) and software optimizations (data layout restructuring, prefetching strategies, memory-aware scheduling). Through this comprehensive treatment of the memory-centric analysis pipeline—from trace collection through architectural modeling to co-design application—we provide researchers and practitioners with practical techniques for addressing memory bottlenecks in contemporary computing systems.
Described herein is a method and system for training nonlinear adaptive filters (or neural networks) which have embedded memory. Such memory can arise in a multi-layer finite impulse response (FIR) architecture, or an infinite impulse response (IIR) architecture. We focus on filter architectures with separate linear dynamic components and static nonlinear components. Such filters can be structured so as to restrict their degrees of computational freedom based on a priori knowledge about the dynamic operation to be emulated. The method is detailed for an FIR architecture which consists of linear FIR filters together with nonlinear generalized single layer subnets. For the IIR case, we extend the methodology to a general nonlinear architecture which uses feedback. For these dynamic architectures, we describe how one can apply optimization techniques which make updates closer to the Newton direction than those of a steepest descent method, such as backpropagation. We detail a novel adaptive modified Gauss-Newton optimization technique, which uses an adaptive learning rate to determine both the magnitude and direction of update steps. For a wide range of adaptive filtering applications, the new training algorithm converges faster and to a smaller value of cost than both steepest-descent methods such as backpropagation-through-time, and standard quasi-Newton methods. We apply the algorithm to modeling the inverse of a nonlinear dynamic tracking system 5, as well as a nonlinear amplifier 6.
NASA Lewis Research Center's Space Communications Division has been investigating high-speed digital filters that can operate at a higher speed than those in current use for a digital modulator and demodulator (modem). Using the Canonical Signed Digits (CSD) number representation for filter coefficients is a very effective way to increase the filter's speed while reducing complexity in the digital filter hardware design. This approach is a good alternative to using an expensive parallel-processing design technique or custom, application-specific integrated circuits. Such integrated circuits may not be suitable for applications that require filter speeds faster than what application-specific integrated circuits digital signal processors can offer for a dedicated channel. When a communication channel is a dedicated, multiplication process--a costly, time-consuming process--it can be greatly simplified by a replacement of the filter coefficients with CSD numbers. A computer code written with the MATLAB software package runs the program and generates CSD-represented filter coefficients that are based on minimizing minimum mean square errors. Also, the Alta Group of Cadence's Signal Processing Workstation is used to simulate and analyze the CSD filter responses. The impulse response of the root-raised cosine filter that is used as a base model is defined. From this filter, a set of coefficients is sampled and stored in a file. For the all coefficients, the optimal CSD number for each coefficient is searched on the basis of the minimum-mean-square-errors criterion. Because the distribution of CSD numbers is not uniform, quantization errors tend to be bigger for coefficients greater than 1/2. To offset errors that occur in a region of coefficients between 1/2 to 1 and to better represent fractions with CSD numbers, an extra nonzero digit is allowed for any coefficients exceeding 1/2. This will greatly improve frequency response as well as intersymbol interference at the receiver. The frequency response of a set of collected CSD-represented filter coefficients was compared with the same filter that was conventionally implemented. Analyses show CSD-implemented filters perform as well as conventional filters. Comparison of eye diagrams and bit-error-rate curves between CSD filters and traditionally implemented filters are almost indistinguishable. However, filter complexity was reduced from almost 3.5 to 1 for CSD filters. Complete computer simulation results are available. In the near future, work will focus on building actual working digital filter hardware in a field programmable gate array (FPGA).
The Navigation Campaign of the OSIRIS-REx mission consisted of three phases: Approach, Preliminary Survey and Orbital-A. These phases were designed to optimize the initial characterization of Bennu's mass, shape and spin-state to support a safe orbit insertion and a quick transition to landmark-based optical navigation tracking. The standard orbit determination filtering techniques used to navigate the spacecraft were unable to fit data from these three phases simultaneously due to numerical issues associated with the nonlinear dynamics and the long arc length. Consequently, a multi-arc filtering algorithm was implemented in order to combine the information from each of these arcs. Multi-arc solutions for Bennu's spin state and gravity field are presented here.
This document pertains to the reporting requirements of DOE contract FE-0031767. The document covers the development of a model-based estimation (MBE) library for generating digital twins of coal-fired power plants (CFPPs). The models used by the MBE are transient and physics-based. Model classes in the library that correspond to different components in CFPPs are presented with physical explanations and representative results. MBE library provides Extended Kalman Filter (EKF), and Unscented Kalman Filter (UKF) algorithms. The MBE library is generic and can be used to represent a variety of CFPPs. A digital twin of a CFPP will be created as the reduced-order model of that CFPP is embedded in the developed MBE architecture.
NIRCam is the primary near-infrared camera for JWST. It is well-suited for a wide variety of star formation and other scientific studies and will also be used for aligning the JWST primary mirror segments and telescope. NIRCam will provide high sensitivity, high spatial resolution images in a variety of filters over the 0.6 - 5 μm wavelength region including coronagraphic modes and slitless spectroscopic capabilities. NIRCam consists of two identical optical modules that have adjacent 2.2' x 2.2' fields of view on the sky. Each module is composed of two channels that observe the same field at short (< 2.4 μm) or long (> 2.4 μm) wavelengths simultaneously, with optimized spatial sampling in each channel. NIRCam optical elements include narrow (R = 100), medium (R = 10), and wide (R <= 4) filters, coronagraphic imaging spot and bar masks, and grisms that provide wide-field or time-series slitless spectroscopy. These major NIRCam features will be introduced, and some applications to galactic star formation will be briefly presented.