Stochastic Systems

Estimation of jerk-wise processes with incomplete knowledge

M. V. Grishin, A. V. Dobrovidov

Moscow

Abstract: A method is proposed for estimating the times of jerk-wise change in the parameters of the random process where the probabilistic characteristics of jerks are unknown and only one state of the parameter change process is known. With reference to an autoregression model the procedure is studied and compared with optimal methods for complete knowledge.

UDC: 62-501.72:519.272


Received: 14.04.1982

 English version: Automation and Remote Control, 1983, 44:11, 1438–1443

Bibliographic databases:

•