Nonlinear Time-Domain Models of Human Controllers

Taylor Jr., Lawrence W. (1968) Nonlinear Time-Domain Models of Human Controllers. Technical Report NASA TM X-60996, Research Engineering, NASA Dryden Flight Research Center.

Full text available as: PDF - Requires Adobe Acrobat Reader or other PDF viewer.

Abstract

Analysis in the time domain is more advantageous than analysis in the frequency domain because (1) fewer values are needed for a characterization (this is especially true if the proposed method expansion for the weighting function is used) and (2) greater accuracy is achieved, especially when the input disturbance is not large compared with the remnant. The fit error over the data used to determine the model decreases with increased memory time and order of nonlinearity. This result can be misleading, however, as the average variance of the weighting terms increases. The fit error over new or independent data should be used to assess which memory time and what order of nonlinearity should be used.

EPrint Type:	NASA Technical Memorandum
Keywords:	Linear prediction, Mathematical models, Nonlinearity, Operator performance, Pilot performance
Subjects:	(51 - 55) Life Sciences: Other, Life Sciences
ID Code:	834
Deposited On:	05 April 2006
Additional Information:	19 pages.