AUTOMATED PARAMETER-IDENTIFICATION AND ORDER REDUCTION FOR DISCRETE-TIME SERIES MODELS

The development of an algorithm for the parameter identification of discrete time-series models for dynamic structural systems is presented. The estimated models are used to predict the transient response of structures subject to arbitrary input. The linear, structural transfer function is expressed in the form of an autoregressive moving-average (ARMA) discrete time-series model. The accurate estimation of the system parameters is often limited due to the influence of measurement noise. Overspecification of the system order has proven useful in reducing the bias errors, but the resulting model may be excessively large and will contain noise related modes. A completely automated approach for parameter identification and model reduction is presented. Experimental response data from a subscale sailplane is used to demonstrate the technique. The resulting reduced order models provided accurate prediction of the transient response of the sailplane to arbitrary inputs.