Stochastic modeling of human learning behavior
Stochastic modeling of human learning behavior in manual control task
SEARCH · Search NASA
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.
Stochastic modeling of human learning behavior in manual control task
Stochastic model for computerized information handling facility with finite capacity
A stochastic model of turbulent mixing was developed for a reactor in which mixing is represented by n-body fluid particle interactions. The model was used to justify the assumption (made in previous investigations of the role of turbulent mixing on burner generated thermal nitric oxide and carbon monoxide emissions) that for a simple plug flow reactor, composition nonuniformities can be described by a Gaussian distribution function in the local fuel:air equivalence ratio. Recent extensions of this stochastic model to include the combined effects of turbulent mixing and secondary air entrainment on thermal generation of nitric oxide in gas turbine combustors are discussed. Finally, rate limited upper and lower bounds of the nitric oxide produced by thermal fixation of molecular nitrogen and oxidation of organically bound fuel nitrogen are estimated on the basis of the stochastic model for a plug flow burner; these are compared with experimental measurements obtained using a laboratory burner operated over a wide range of test conditions; good agreement is obtained.
A stochastic model of particle dispersion by turbulence, proposed by Gosman and Ioannides (1981), is evaluated. The method employs a k-epsilon model to estimate turbulence properties. Dispersion is determined by computing particle motion, with random sampling to obtain instantaneous flow properties for a statistically significant number of particle trajectories. The stochastic model yields good results particularly when eddy lifetimes are evaluated. The method allows the effects of large relative velocities between the particles and the flow, drag properties at Reynolds numbers greater than the Stokes flow regime, and the variations of local turbulence properties to be readily handled, at least for boundary layer flows.
No abstract prepared.
In this study, we analyze three-dimensional particle-laden, isotropic turbulence to develop an understanding of inertial particle dynamics from a kinetic energy perspective. Data trends implying inhomogeneous sampling of the flow by particles are identified and used to support a proposed particle behavior: particles appear to accumulate in regions of low flow kinetic energy over time because they lose kinetic energy and slow down in such regions, ultimately causing them to spend more time there. To elucidate this behavior, we derive a particle kinetic energy equation from the particle momentum equation, which incorporates inertial effects through the Schiller–Naumann drag correlation. Upon extracting fundamental physics from this equation, hypotheses regarding the role of the Stokes number in the temporal change of particle kinetic energy and the previously proposed particle behavior are evaluated using simulation data considering three Stokes numbers. Finally, a Fokker–Planck equation is used to derive the steady-state probability density function of the particle kinetic energy. The model fits the simulation data well and provides a tool for further investigation into understanding preferential concentration, as well as a reduced order model for predicting particle kinetic energy in turbulent flows.
We propose a general stochastic model for the UT1-LOD system and derive the corresponding Kalman filter model.
A computational methodology is developed to address the solution of high-dimensional stochastic problems. It utilizes high-dimensional model representation (HDMR) technique in the stochastic space to represent the model output as a finite hierarchical correlated function expansion in terms of the stochastic inputs starting from lower-order to higher-order component functions. HDMR is efficient at capturing the high-dimensional input-output relationship such that the behavior for many physical systems can be modeled to good accuracy only by the first few lower-order terms. An adaptive version of HDMR is also developed to automatically detect the important dimensions and construct higher-order terms using only the important dimensions. The newly developed adaptive sparse grid collocation (ASGC) method is incorporated into HDMR to solve the resulting sub-problems. By integrating HDMR and ASGC, it is computationally possible to construct a low-dimensional stochastic reduced-order model of the high-dimensional stochastic problem and easily perform various statistic analysis on the output. Several numerical examples involving elementary mathematical functions and fluid mechanics problems are considered to illustrate the proposed method. The cases examined show that the method provides accurate results for stochastic dimensionality as high as 500 even with large-input variability. The efficiency of the proposed method is examined by comparing with Monte Carlo (MC) simulation.
Nearly all proglacial water discharge from the Greenland Ice Sheet is routed englacially, from the surface to the bed, via moulins. Identification of moulins in high-resolution imagery is a frequent topic of study, but the processes controlling how and where moulins form remain poorly understood. We seek to leverage information gained from the development of a physical model of moulin formation, remotely sensed ice-sheet data products, and an analytic model of ice-flow perturbations to develop a predictive stochastic model of moulin distribution across Greenland. Here we present initial results from the physical model of moulin formation and characterize the sensitivity of moulin geometry to a range of model parameters. This parameterization of moulin formation is the first step in developing a stochastic model that will be a predictive, computationally efficient representation of the englacial hydrologic system.
A linear stochastic model of the human operator is developed and applied to the problem of piloted control of an aircraft. The pilot and aircraft are modeled as linear time-invariant systems containing both process and measurement noise. The loop closure by the pilot is determined by formulating the problem as an optimal stochastic control problem. The solution to the optimal control problem yields not only the pilot's optimal control output which he uses to control the vehicle, but also the optimal combination of his observations of the vehicle states upon which the pilot bases his control. A method is presented so that, using experimental pilot vehicle data, the cost functional which is minimized in the optimal control problem will be numerically equal to the pilot rating that the pilot would associate with the given vehicle and task.
Applying extreme temperature events for future conditions is not straightforward for infrastructure resilience analyses. This work introduces a stochastic model that fills this gap. The model uses at least 50 years of daily extreme temperature records, climate normals with 10%-90% confidence intervals, and shifts/offsets for increased frequency and intensity of heat wave events. Intensity and frequency are shifted based on surface temperature anomaly from 1850-1900 for 32 models from CMIP6. A case study for Worcester, Massachusetts passed 85% of cases using the two-sided Kolmogorov-Smirnov -value test with 95% confidence for both temperature and duration. Future shifts for several climate scenarios to 2020, 2040, 2060, and 2080 had acceptable errors between the shifted model and 10- and 50-year extreme temperature event thresholds with the largest error being 2.67 degrees C. The model is likely to be flexible enough for other patterns of extreme weather such as extreme precipitation and hurricanes.
Existing stochastic models of the galactic magnetic field are considered and found to suffer certain defects. A new model is proposed which overcomes faults of the previous theories while retaining their strengths.
Existing stochastic models of the galactic magnetic field are considered and found to suffer certain defects. A new model is proposed which overcomes faults of the previous theories while retaining their strenghts.
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.
This paper describes a methodology for making cost effective fatigue design decisions. The methodology is based on a probabilistic model for the stochastic process of fatigue crack growth with time. The development of a particular model for the stochastic process is also discussed in the paper. The model is based on the assumption of continuous time and discrete space of crack lengths. Statistical decision theory and the developed probabilistic model are used to develop the procedure for making fatigue design decisions on the basis of minimum expected cost or risk function and reliability bounds. Selections of initial flaw size distribution, NDT, repair threshold crack lengths, and inspection intervals are discussed.
A simplified stochastic model is proposed for crack initiation and short-crack growth under creep and creep-fatigue conditions. Material inhomogeneity provides the random nature of crack initiation and early growth. In the model, the influence of microstructure is introduced by the variability of: (1) damage accumulation along grain boundaries, (2) critical damage required for crack initiation or growth, and (3) the grain-boundary length. The probabilities of crack initiation and growth are derived by using convolution integrals. The model is calibrated and used to predict the crack density and crack-growth rate of short cracks of 304 stainless steel under creep and creep-fatigue conditions. The mean-crack initiation lives are predicted to be within an average deviation of about 10 percent from the experimental results. The predicted cumulative distributions of crack-growth rate follow the experimental data closely. The applicability of the simplified stochastic model is discussed and the future research direction is outlined.
A simplified stochastic model is proposed for crack initiation and short-crack growth under creep and creep-fatigue conditions. Material inhomogeneity provides the random nature of crack initiation and early growth. In the model, the influence of microstructure is introduced by the variability of: (1) damage accumulation along grain boundaries, (2) critical damage required for crack initiation or growth, and (3) the grain-boundary length. The probabilities of crack initiation and growth are derived by using convolution integrals. The model is calibrated and used to predict the crack density and crack-growth rate of short cracks of 304 stainless steel under creep and creep-fatigue conditions. The mean-crack initiation lives are predicted to be within an average deviation of about 10 percent from the experimental results. The predicted comulative distributions of crack-growth rate follow the experimental data closely. The applicability of the simplified stochastic model is discussed and the future research direction is outlined.
We collected and cleaned the synthetic grid data produced by NREL for the Texas and New York synthetic grids. We developed a high dimensional joint stochastic model for load at the zone level, and solar and wind power productions at the asset level, capturing the spatial and temporal dependencies between all the variables, and demonstrated how such a model could be fitted to historical data. We designed and implemented a simulation engine which can produce Monte Carlo scenarios for the hourly day-ahead values of load, and solar and wind power productions at the spatial and temporal resolutions of the historical data used to fit the model. Finally we developed an open-source Python package which can, from an input grid model, efficiently use forecasts and large numbers of Monte Carlo scenarios to provide unit commitment and economic dispatch for each of these scenarios. The high dimensional stochastic model and the subsequent Monte Carlo simulation engine were implemented in the package PGscen and the corresponding UC and ED optimization programs in the package Vatic.