Local Linearisation Using Fuzzy (Soft) Input Space Partition for Nonlinear System Modelling

摘要

Fuzzy local linearisation (FLL) is a useful divide-and-conquer method for coping with complex problems such as nonlinear system modelling. In this paper, a modified adaptive spline modelling (MASMOD) algorithm for FLL is introduced, and a hybrid learning scheme which combines MASMOD and expectation-maximisation (EM) algorithms is proposed based on a probabilistic interpretation of FLL. Application examples on nonlinear time-series analysis and nonlinear trajectory estimation using FLL models are presented to testify the algorithms developed in the paper.