A sequential learning algorithm for RBF networks with application to ship inverse control

An improved minimal resource allocating network (IMRAN) learning algorithm is developed for constructing radial basis function (RBF) network. The RBF network is adjusted on-line for both network structure and connecting parameters. Based on the proposed sequential learning algorithm, a direct inverse control strategy is introduced and applied to ship course-keeping control. Simulation results of ship course-keeping experiment demonstrate the applicability and effectiveness of the sequential learning algorithm and the RBF network-based inverse control strategy.