Advanced methods of model structure determination from test data
N. GUPTA, W. HALL, JR., Thomas L. Trankle · Guidance and Control Conference · 1977
The technology of system identification from input/output test data has been developed into an operational procedure for improving and validating mathematical models upon which state-of-the-art guidance and control systems are based. A wide application of this technology has focused attention on one particular phase of the system identification methodology, the model structure determination phase, as an essential step. This step consists of processing the input/output data to determine the significant linear and nonlinear equations and associated parameters that are necessary to represent an observed system response. Although the model structure determination problem has been recognized for a long time, it has been attacked systematically only recently. This paper summarizes various aspects of this problem and results of some recent research. A unified approach is developed for this problem and is applied to test data from a submarine, a missile, and an aircraft engine.