ON THE DESIGN OF A CONTINUOUS-TIME MODEL ESTIMATOR FOR BIAS DISTRIBUTION

Modelling of physical processes always involves approximations. This u ndermodelling, coupled with noisy process data, results in biased freq uency response estimates of models. The bias error cannot be eliminate d, but can be distributed over frequencies by careful design of the es timator. This paper aims at analysing the bias error in the estimates from an equation-error algorithm based on tau-time-ahead prediction us ing a continuous model of the process with filtered input/output signa ls. From the analysis, the roles played by the design variables of the estimator in affecting the bias distribution are identified and the c hoice of these is discussed. Finally, a procedure to optimize the desi gn variables is indicated, Numerical simulation results are presented. (C) 1997 by John Wiley & Sons, Ltd.