Recursive nonlinear identification using multiple model algorithm

In this paper, the multiple model algorithm is used in deriving recursive algorithms for the identification of nonlinear systems. The radial basis function (RBF) networks with only linear weights requiring estimation combined with the Kalman filter algorithm forms the essence of the identification algorithm. Multiple networks are used to identify the multi-modes of the system under a Markovian assumption, the model estimation and selection being carried out on-line. Both, `hard´ and `soft´ competition based estimation schemes are developed where in the former, the most probable network is adapted by the Kalman filter and in the latter all networks are adapted by appropriate weighting of the observation