Search NASA⌕ Search

SEARCH · Search NASA

Results for “prediction algorithms”

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 541 records · Page 30

Analytical Algorithms to Quantify the Uncertainty in Remaining Useful Life Prediction

This paper investigates the use of analytical algorithms to quantify the uncertainty in the remaining useful life (RUL) estimate of components used in aerospace applications. The prediction of RUL is affected by several sources of uncertainty and it is important to systematically quantify their combined effect by computing the uncertainty in the RUL prediction in order to aid risk assessment, risk mitigation, and decisionmaking. While sampling-based algorithms have been conventionally used for quantifying the uncertainty in RUL, analytical algorithms are computationally cheaper and sometimes, are better suited for online decision-making. While exact analytical algorithms are available only for certain special cases (for e.g., linear models with Gaussian variables), effective approximations can be made using the the first-order second moment method (FOSM), the first-order reliability method (FORM), and the inverse first-order reliability method (Inverse FORM). These methods can be used not only to calculate the entire probability distribution of RUL but also to obtain probability bounds on RUL. This paper explains these three methods in detail and illustrates them using the state-space model of a lithium-ion battery.

Sankararaman, Shankar↗

Automated electrical demand peak leveling in a manufacturing facility with short term energy storage for smart grid participation

Demand side management of energy is a vital function in the smart grid that allows for greater integration of renewable energy resources, and is facilitated with economic incentives in energy and demand pricing schedules. Manufacturing systems can take advantage of these incentives to reduce their energy costs through active energy management. This paper is a simulation analysis on the implementation of a novel algorithm that levelizes the maximum power load of an industrial bakery that operates under an on-peak demand price structure. The algorithm takes advantage of an untapped thermal energy storage resource in the facility, a chilled glycol buffer tank, to level the short intra-day power oscillations that the facility experiences. One of the key advantages of the algorithm is that it does not require precise demand forecasting or complex control algorithms, such as model predictive control. Even with substantial error in peak power estimates (5%), the algorithm is expected to result in at least a 2% peak reduction. Under ideal prediction, the algorithm reduces the on-peak facility power maximum by 7.3% of the possible 9.0% reduction the storage can provide. Further, the algorithm uses very few facility specific inputs, and can readily be adapted for any facility with short intra-day oscillatory power profiles and untapped storage capacity. The key insight of the paper is the employment of a novel algorithm to leverage untapped energy storage in manufacturing facilities to transform them into smart grid participants with no major capital investment.

24 POWER TRANSMISSION AND DISTRIBUTION↗

Assessment of Envelope- and Machine Learning-Based Electrical Fault Type Detection Algorithms for Electrical Distribution Grids

This study introduces envelope- and machine learning (ML)-based electrical fault type detection algorithms for electrical distribution grids, advancing beyond traditional logic-based methods. The proposed detection model involves three stages: anomaly area detection, ML-based fault presence detection, and ML-based fault type detection. Initially, an envelope-based detector identifying the anomaly region was improved to handle noisier power grid signals from meters. The second stage acts as a switch, detecting the presence of a fault among four classes: normal, motor, switching, and fault. Finally, if a fault is detected, the third stage identifies specific fault types. This study explored various feature extraction methods and evaluated different ML algorithms to maximize prediction accuracy. The performance of the proposed algorithms is tested in an emulated software–hardware electrical grid testbed using different sample rate meters/relays, such as SEL735, SEL421, SEL734, SEL700GT, and SEL351S near and far from an inverter-based photovoltaic array farm. The performance outcomes demonstrate the proposed model’s robustness and accuracy under realistic conditions.

24 POWER TRANSMISSION AND DISTRIBUTION↗

MO-FG-202-01: A Fast Yet Sensitive EPID-Based Real-Time Treatment Verification System

Purpose: To create a real-time EPID-based treatment verification system which robustly detects treatment delivery and patient attenuation variations. Methods: Treatment plan DICOM files sent to the record-and-verify system are captured and utilized to predict EPID images for each planned control point using a modified GPU-based digitally reconstructed radiograph algorithm which accounts for the patient attenuation, source energy fluence, source size effects, and MLC attenuation. The DICOM and predicted images are utilized by our C++ treatment verification software which compares EPID acquired 1024×768 resolution frames acquired at ∼8.5hz from Varian Truebeam™ system. To maximize detection sensitivity, image comparisons determine (1) if radiation exists outside of the desired treatment field; (2) if radiation is lacking inside the treatment field; (3) if translations, rotations, and magnifications of the image are within tolerance. Acquisition was tested with known test fields and prior patient fields. Error detection was tested in real-time and utilizing images acquired during treatment with another system. Results: The computational time of the prediction algorithms, for a patient plan with 350 control points and 60×60×42cm^3 CT volume, is 2–3minutes on CPU and <27 seconds on GPU for 1024×768 images. The verification software requires a maximum of ∼9ms and ∼19ms for 512×384 and 1024×768 resolution images, respectively, to perform image analysis and dosimetric validations. Typical variations in geometric parameters between reference and the measured images are 0.32°for gantry rotation, 1.006 for scaling factor, and 0.67mm for translation. For excess out-of-field/missing in-field fluence, with masks extending 1mm (at isocenter) from the detected aperture edge, the average total in-field area missing EPID fluence was 1.5mm2 the out-of-field excess EPID fluence was 8mm^2, both below error tolerances. Conclusion: A real-time verification software, with EPID images prediction algorithm, was developed. The system is capable of performing verifications between frames acquisitions and identifying source(s) of any out-of-tolerance variations. This work was supported in part by Varian Medical Systems.

60 APPLIED LIFE SCIENCES↗

An Efficient Deterministic Approach to Model-based Prediction Uncertainty Estimation

Prognostics deals with the prediction of the end of life (EOL) of a system. EOL is a random variable, due to the presence of process noise and uncertainty in the future inputs to the system. Prognostics algorithm must account for this inherent uncertainty. In addition, these algorithms never know exactly the state of the system at the desired time of prediction, or the exact model describing the future evolution of the system, accumulating additional uncertainty into the predicted EOL. Prediction algorithms that do not account for these sources of uncertainty are misrepresenting the EOL and can lead to poor decisions based on their results. In this paper, we explore the impact of uncertainty in the prediction problem. We develop a general model-based prediction algorithm that incorporates these sources of uncertainty, and propose a novel approach to efficiently handle uncertainty in the future input trajectories of a system by using the unscented transformation. Using this approach, we are not only able to reduce the computational load but also estimate the bounds of uncertainty in a deterministic manner, which can be useful to consider during decision-making. Using a lithium-ion battery as a case study, we perform several simulation-based experiments to explore these issues, and validate the overall approach using experimental data from a battery testbed.

Daigle, Matthew J.↗

Using an external surrogate for predictor model training in real-time motion management of lung tumors

Purpose: Precise prediction of respiratory motion is a prerequisite for real-time motion compensation techniques such as beam, dynamic couch, or dynamic multileaf collimator tracking. Collection of tumor motion data to train the prediction model is required for most algorithms. To avoid exposure of patients to additional dose from imaging during this procedure, the feasibility of training a linear respiratory motion prediction model with an external surrogate signal is investigated and its performance benchmarked against training the model with tumor positions directly. Methods: The authors implement a lung tumor motion prediction algorithm based on linear ridge regression that is suitable to overcome system latencies up to about 300 ms. Its performance is investigated on a data set of 91 patient breathing trajectories recorded from fiducial marker tracking during radiotherapy delivery to the lung of ten patients. The expected 3D geometric error is quantified as a function of predictor lookahead time, signal sampling frequency and history vector length. Additionally, adaptive model retraining is evaluated, i.e., repeatedly updating the prediction model after initial training. Training length for this is gradually increased with incoming (internal) data availability. To assess practical feasibility model calculation times as well as various minimum data lengths for retraining are evaluated. Relative performance of model training with external surrogate motion data versus tumor motion data is evaluated. However, an internal–external motion correlation model is not utilized, i.e., prediction is solely driven by internal motion in both cases. Results: Similar prediction performance was achieved for training the model with external surrogate data versus internal (tumor motion) data. Adaptive model retraining can substantially boost performance in the case of external surrogate training while it has little impact for training with internal motion data. A minimum adaptive retraining data length of 8 s and history vector length of 3 s achieve maximal performance. Sampling frequency appears to have little impact on performance confirming previously published work. By using the linear predictor, a relative geometric 3D error reduction of about 50% was achieved (using adaptive retraining, a history vector length of 3 s and with results averaged over all investigated lookahead times and signal sampling frequencies). The absolute mean error could be reduced from (2.0 ± 1.6) mm when using no prediction at all to (0.9 ± 0.8) mm and (1.0 ± 0.9) mm when using the predictor trained with internal tumor motion training data and external surrogate motion training data, respectively (for a typical lookahead time of 250 ms and sampling frequency of 15 Hz). Conclusions: A linear prediction model can reduce latency induced tracking errors by an average of about 50% in real-time image guided radiotherapy systems with system latencies of up to 300 ms. Training a linear model for lung tumor motion prediction with an external surrogate signal alone is feasible and results in similar performance as training with (internal) tumor motion. Particularly for scenarios where motion data are extracted from fluoroscopic imaging with ionizing radiation, this may alleviate the need for additional imaging dose during the collection of model training data.

60 APPLIED LIFE SCIENCES↗

Data-Enabled Predictive Control for Building HVAC Systems

Model predictive control is widely used as a control technology for the computation of optimal control inputs of building heating, ventilating, and air conditioning (HVAC) systems. However, both the benefits and widespread adoption of model predictive control (MPC) are hindered by the effort of model creation, calibration, and accuracy of the predictions. In this paper, we apply the data-enabled predictive control (DeePC) algorithm for designing controls for building HVAC systems. The algorithm solely depends on input/output data from the system to predict future state trajectories without the need for system identification. The algorithm relies on the idea that a vector space of all input–output trajectories of a discrete-time linear time-invariant (LTI) system is spanned by time-shifts of a single measured trajectory, given the input signal is persistently exciting. Closed-loop simulations using EnergyPlus are performed to demonstrate the approach. The simulated building modeled in EnergyPlus is a modified commercial large office prototype building served by an air handling unit-variable air volume HVAC system. Temperature setpoints of zones are used as control variables to minimize the HVAC energy cost of the building considering a time-of-use electricity rate structure. Furthermore, sensitivity analysis is conducted to gain insights into the effect of parameter tuning on DeePC performance. Simulation results are used to illustrate the performance of the algorithm and compare the algorithm with model-based MPC and occupancy-based setpoint controller. Overall, DeePC achieves similar performance compared to MPC for lower engineering effort.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗

Stochastic Model Predictive Control With Gaussian Wind Direction Preview for Wake Steering

This article addresses the problem of wake steering control for wind farms that explicitly consider the tradeoff between farm-level power generation and yaw duty cycle under variable and uncertain wind conditions. A novel stochastic model predictive control (MPC) algorithm is presented, which utilizes a stochastic model of the freestream wind field components in a receding horizon framework to compute optimal yaw set points that maximize the expected value of the farm power while constraining the yaw actuation. Different configurations of the algorithm are evaluated using a steady-state wind farm simulator. The proposed stochastic MPC algorithm can plan control actions over a future prediction horizon based on probabilistic estimates of the incoming wind magnitude and direction.

17 WIND ENERGY↗

A numerical solution of Duffing's equations including the prediction of jump phenomena

Numerical methodology for the solution of Duffing's differential equation is presented. Algorithms for the prediction of multiple equilibrium solutions and jump phenomena are developed. In addition, a filtering algorithm for producing steady state solutions is presented. The problem of a rigidly clamped circular plate subjected to cosinusoidal pressure loading is solved using the developed algorithms (the plate is assumed to be in the geometrically nonlinear range). The results accurately predict regions of solution multiplicity and jump phenomena.

Moyer, E. T., Jr.↗

Diffusion algorithms and data reduction routine for onsite launch predictions for the transport of Titan 3 C exhaust effluents

The NASA/MSFC multilayer diffusion algorithms have been specialized for the prediction of the surface impact for the dispersive transport of the exhaust effluents from the launch of a Titan 3 vehicle. This specialization permits these transport predictions to be made at the launch range in real time so that the effluent monitoring teams can optimize their monitoring grids. Basically, the data reduction routine requires just the meteorology profiles for the thermodynamics and kinematics of the atmosphere as an input. These profiles are graphed along with the resulting exhaust cloud rise history, the center line concentrations and dosages, and the hydrogen chloride isopleths.

Stephens, J. B.↗

Machine Learning for Prediction of Thermodynamic Descriptors

Our objective is to apply machine learning (ML) algorithms for the prediction of molecular catalysis descriptors from geometric properties derived from experimental crystallographic databases. Catalysis is often considered a “low-data” discipline that is poorly suited for ML methods. An exception is the extensive structural information that is available for molecular catalysts through the Cambridge Structural Database (CSD), which contains atomically precise molecular structures from X-ray diffraction analysis for >600K metal complexes. As a proof-of-principle, we targeted the prediction of hydricity, a thermodynamic property that provides understanding and control of catalytic hydride transfer. We built a training set composed of ~100 molecular complexes with a known hydricity and structural information from the CSD. This data set was converted into a machine-readable format using the smooth overlap of atomic positions (SOAP) representation and further labeled with simple electronic descriptors for the metal centers. Multiple different neural networks were trained on this data set, and the accuracy of the hydricity predictions ranged from < 2 kcal/mol to 20 kcal/mol. The accuracy of each model was highly sensitive to which compounds were in the train versus test set, underscoring the challenges associated with small and chemically diverse data sets. Finally, to further augment the data set, we attempted to experimentally measure several new hydricity values, however these experiments were unsuccessful due to undesired chemical reactivity of the selected complexes.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Concept and simulation study of a novel localization method for robotic endoscopic capsules using multiple positron emission markers

Purpose: Over the last decade, wireless capsule endoscope has been the tool of choice for noninvasive inspection of the gastrointestinal tract, especially in the small intestine. However, the latest clinical products have not been equipped with a sufficiently accurate localization system which makes it difficult to determine the location of intestinal abnormalities, and to apply follow-up interventions such as biopsy or drug delivery. In this paper, the authors present a novel localization method based on tracking three positron emission markers embedded inside an endoscopic capsule. Methods: Three spherical {sup 22}Na markers with diameters of less than 1 mm are embedded in the cover of the capsule. Gamma ray detectors are arranged around a patient body to detect coincidence gamma rays emitted from the three markers. The position of each marker can then be estimated using the collected data by the authors’ tracking algorithm which consists of four consecutive steps: a method to remove corrupted data, an initialization method, a clustering method based on the Fuzzy C-means clustering algorithm, and a failure prediction method. Results: The tracking algorithm has been implemented inMATLAB utilizing simulation data generated from the Geant4 Application for Emission Tomography toolkit. The results show that this localization method can achieve real-time tracking with an average position error of less than 0.4 mm and an average orientation error of less than 2°. Conclusions: The authors conclude that this study has proven the feasibility and potential of the proposed technique in effectively determining the position and orientation of a robotic endoscopic capsule.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

PACOSS program

The objectives of the PACOSS program were to demonstrate the respective roles of passive and active control for structures that represented future Large Space Structures (LSS), to develop means to introduce passive vibration control, and to experimentally verify the damping predictions and the control algorithms. In order to meet the objectives, the program was divided into an analytical simulation phase to establish the respective roles of passive and active damping on a LSS-type structure, and a design, analysis, and test phase to validate the passive damping and the control algorithm performance for a structure. Predictable amounts of damping can be designed into a LSS structure, the best control strategy uses a combination of passive damping and active controls, and a more optimum system can be achieved by an early interaction between the structural designer, controls engineer, and the damping designer.

Richards, K. E., Jr.↗

Hardware Implementation of Lossless Adaptive Compression of Data From a Hyperspectral Imager

Efficient onboard data compression can reduce the data volume from hyperspectral imagers on NASA and DoD spacecraft in order to return as much imagery as possible through constrained downlink channels. Lossless compression is important for signature extraction, object recognition, and feature classification capabilities. To provide onboard data compression, a hardware implementation of a lossless hyperspectral compression algorithm was developed using a field programmable gate array (FPGA). The underlying algorithm is the Fast Lossless (FL) compression algorithm reported in Fast Lossless Compression of Multispectral- Image Data (NPO-42517), NASA Tech Briefs, Vol. 30, No. 8 (August 2006), p. 26 with the modification reported in Lossless, Multi-Spectral Data Comressor for Improved Compression for Pushbroom-Type Instruments (NPO-45473), NASA Tech Briefs, Vol. 32, No. 7 (July 2008) p. 63, which provides improved compression performance for data from pushbroom-type imagers. An FPGA implementation of the unmodified FL algorithm was previously developed and reported in Fast and Adaptive Lossless Onboard Hyperspectral Data Compression System (NPO-46867), NASA Tech Briefs, Vol. 36, No. 5 (May 2012) p. 42. The essence of the FL algorithm is adaptive linear predictive compression using the sign algorithm for filter adaption. The FL compressor achieves a combination of low complexity and compression effectiveness that exceeds that of stateof- the-art techniques currently in use. The modification changes the predictor structure to tolerate differences in sensitivity of different detector elements, as occurs in pushbroom-type imagers, which are suitable for spacecraft use. The FPGA implementation offers a low-cost, flexible solution compared to traditional ASIC (application specific integrated circuit) and can be integrated as an intellectual property (IP) for part of, e.g., a design that manages the instrument interface. The FPGA implementation was benchmarked on the Xilinx Virtex IV LX25 device, and ported to a Xilinx prototype board. The current implementation has a critical path of 29.5 ns, which dictated a clock speed of 33 MHz. The critical path delay is end-to-end measurement between the uncompressed input data and the output compression data stream. The implementation compresses one sample every clock cycle, which results in a speed of 33 Msample/s. The implementation has a rather low device use of the Xilinx Virtex IV LX25, making the total power consumption of the implementation about 1.27 W.

Keymeulen, Didlier↗

Accurate Fault Prediction of BlueGene/P RAS Logs Via Geometric Reduction

The authors are building two algorithms for fault prediction using raw system-log data. This work is preliminary, and has only been applied to a limited dataset, however the results seem promising. The conclusions are that: (1) obtaining useful data from RAS-logs is challenging; (2) extracting concentrated information improves efficiency and accuracy; and (3) function evaluation algorithms are fast and lend well to scaling.

97 MATHEMATICS AND COMPUTING↗

Prediction of Aerodynamic Coefficient using Genetic Algorithm Optimized Neural Network for Sparse Data

Wind tunnels use scale models to characterize aerodynamic coefficients, Wind tunnel testing can be slow and costly due to high personnel overhead and intensive power utilization. Although manual curve fitting can be done, it is highly efficient to use a neural network to define the complex relationship between variables. Numerical simulation of complex vehicles on the wide range of conditions required for flight simulation requires static and dynamic data. Static data at low Mach numbers and angles of attack may be obtained with simpler Euler codes. Static data of stalled vehicles where zones of flow separation are usually present at higher angles of attack require Navier-Stokes simulations which are costly due to the large processing time required to attain convergence. Preliminary dynamic data may be obtained with simpler methods based on correlations and vortex methods; however, accurate prediction of the dynamic coefficients requires complex and costly numerical simulations. A reliable and fast method of predicting complex aerodynamic coefficients for flight simulation I'S presented using a neural network. The training data for the neural network are derived from numerical simulations and wind-tunnel experiments. The aerodynamic coefficients are modeled as functions of the flow characteristics and the control surfaces of the vehicle. The basic coefficients of lift, drag and pitching moment are expressed as functions of angles of attack and Mach number. The modeled and training aerodynamic coefficients show good agreement. This method shows excellent potential for rapid development of aerodynamic models for flight simulation. Genetic Algorithms (GA) are used to optimize a previously built Artificial Neural Network (ANN) that reliably predicts aerodynamic coefficients. Results indicate that the GA provided an efficient method of optimizing the ANN model to predict aerodynamic coefficients. The reliability of the ANN using the GA includes prediction of aerodynamic coefficients to an accuracy of 110% . In our problem, we would like to get an optimized neural network architecture and minimum data set. This has been accomplished within 500 training cycles of a neural network. After removing training pairs (outliers), the GA has produced much better results. The neural network constructed is a feed forward neural network with a back propagation learning mechanism. The main goal has been to free the network design process from constraints of human biases, and to discover better forms of neural network architectures. The automation of the network architecture search by genetic algorithms seems to have been the best way to achieve this goal.

Rajkumar, T.↗