Semi-parametric adaptive control of discrete-time systems using extreme learning machine
Hao Zhou, Hongbin Ma, Nannan Li, Chenguang Yang · 2017
In this paper, we investigate a novel semi-parametric adaptive design approach for discrete-time systems as a further study of the challenging work on dealing with both parametric and nonparametric uncertainties. An extended version of information concentration (IC) estimator other than traditional recursive identification algorithm is adopted to estimate unknown parameters with a priori knowledge considered. To best utilize input-output history data, an improved version of extreme learning machine is developed to approximate nonparametric part. According to accurate estimates of uncertainties, control signal is established and subsequent simulation examples indicate that the designed adaptive control strategy can guarantee the boundedness of all the closed-loop signals and achieves asymptotic tracking performance.