NASA Logo

NTRS

NTRS - NASA Technical Reports Server

Collections About News Help Login
Back to Results
A Comparison of Three Random Number Generators for Aircraft Dynamic Modeling ApplicationsThree random number generators, which produce Gaussian white noise sequences, were compared to assess their suitability in aircraft dynamic modeling applications. The first generator considered was the MATLAB (registered) implementation of the Mersenne-Twister algorithm. The second generator was a website called Random.org, which processes atmospheric noise measured using radios to create the random numbers. The third generator was based on synthesis of the Fourier series, where the random number sequences are constructed from prescribed amplitude and phase spectra. A total of 200 sequences, each having 601 random numbers, for each generator were collected and analyzed in terms of the mean, variance, normality, autocorrelation, and power spectral density. These sequences were then applied to two problems in aircraft dynamic modeling, namely estimating stability and control derivatives from simulated onboard sensor data, and simulating flight in atmospheric turbulence. In general, each random number generator had good performance and is well-suited for aircraft dynamic modeling applications. Specific strengths and weaknesses of each generator are discussed. For Monte Carlo simulation, the Fourier synthesis method is recommended because it most accurately and consistently approximated Gaussian white noise and can be implemented with reasonable computational effort.
Document ID
20170005202
Acquisition Source
Langley Research Center
Document Type
Technical Memorandum (TM)
Authors
Grauer, Jared A. (NASA Langley Research Center Hampton, VA, United States)
Date Acquired
June 6, 2017
Publication Date
May 1, 2017
Subject Category
Aircraft Stability And Control
Statistics And Probability
Report/Patent Number
L-20807
NASA/TM-2017-219612
NF1676L-26863
Funding Number(s)
WBS: WBS 109492.02.07.02.10
Distribution Limits
Public
Copyright
Work of the US Gov. Public Use Permitted.

Related Records

There are no records associated with this record.
No Preview Available