hrvatski

Optimiranje parametara autoregresijskih modela pomoću genetičkog algoritma

Soft sensors for estimation kerosene distillation end point are developed using linear and nonlinear identification methods. Experimental data are acquired from the refinery distributed control system (DCS) and include on-line available continuously measured variables and laboratory assays. In present work development of AutoRegressive Moving Average with eXogenous inputs (ARMAX) and Nonlinear AutoRegressive model with eXogenous inputs (NARX) are developed. Matlab system identification toolbox allows searching for optimal zero and model pole order for given range only for ARX model, while for other models cumbersome trial and error procedure should be applied. Also, the number of network units in nonlinear block of NARX model is selected in an ad hoc manner. To overcome mentioned problem of finding optimal parameters, the genetic algorithms based code for optimizing model with large number of parameters is developed. Genetic algorithms are used for multiobjective optimization of ARMAX and NARX model parameters in a way to find the best model fits with minimal prediction error for given parameters range. Based on developed soft sensors it is possible to estimate fuel properties continuously as well as apply the methods of inferential control.

Soft senzor; Optimiranja; Genetički algoritmi

nije evidentirano