NASA Logo

NTRS

NTRS - NASA Technical Reports Server

Collections About News Help Login
Back to Results
A two dimensional power spectral estimate for some nonstationary processesA two dimensional estimate for the power spectral density of a nonstationary process is being developed. The estimate will be applied to helicopter noise data which is clearly nonstationary. The acoustic pressure from the isolated main rotor and isolated tail rotor is known to be periodically correlated (PC) and the combined noise from the main and tail rotors is assumed to be correlation autoregressive (CAR). The results of this nonstationary analysis will be compared with the current method of assuming that the data is stationary and analyzing it as such. Another method of analysis is to introduce a random phase shift into the data as shown by Papoulis to produce a time history which can then be accurately modeled as stationary. This method will also be investigated for the helicopter data. A method used to determine the period of a PC process when the period is not know is discussed. The period of a PC process must be known in order to produce an accurate spectral representation for the process. The spectral estimate is developed. The bias and variability of the estimate are also discussed. Finally, the current method for analyzing nonstationary data is compared to that of using a two dimensional spectral representation. In addition, the method of phase shifting the data is examined.
Document ID
19900005527
Acquisition Source
Legacy CDMS
Document Type
Thesis/Dissertation
Authors
Smith, Gregory L. (Hampton Univ. VA, United States)
Date Acquired
September 6, 2013
Publication Date
January 1, 1989
Subject Category
Numerical Analysis
Report/Patent Number
NAS 1.26:186100
NASA-CR-186100
Accession Number
90N14843
Funding Number(s)
CONTRACT_GRANT: NAG1-768
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