Batch scheme Hartley modulating functions method for detecting gradual parameter changes in the identification of integrable nonlinear continuous-time systems

Abstract

In this work, a batch scheme nonrecursive algorithm for gradual or smooth changes of the parameters in the identification of integrable nonlinear continuous-time systems using the Hartley modulating functions (HMF) method is investigated. It is implemented by shifting a fixed window size or record length one step forward at each sampling time. The batch scheme HMF-method assumes the a priori knowledge of the system structure with unknown parameters and uses a frequency weighted least-squares (FWLS) algorithm to identify the unknown gradually or smoothly changing model parameters. The performance of the proposed identification method is investigated on a Van der Pol oscillator system structure. The illustrative simulation studies suggest that this method lies in the potential of being able to accurately identify gradual changes of the parameters of integrable nonlinear continuous-time system in the presence of output measurement disturbances.