一种新的鲁棒自适应广义预测控制算法及鲁棒性分析

A Novel Robust Adaptive Generalized Predictive Control Algorithm

针对对象的不确定性干扰，提出一种新的鲁棒参数自适应算法，证明了这种算法具有自适应验后误差和对象参数估值均值渐近小性质；根据确定性等价原理，将这种普棒参数自适应算法和鲁棒广义预测控制律相结合，提出了一种鲁棒自适应广义预测控制方案，并建立了相应的鲁棒稳定性定理。

Clarke, well known internationally, in 1993 presented a method concerning robustnessof adaptive GPC (Generalized Predictive Control). We, in this paper, present an algorithmbased on our new method [2], which is different from Clarke's.The present paper is the most recent one of a series of papers written by us and published in professional journals [1, 2, 3]. In the present paper, we present a robust parameter adaptive algorithm, which is described mathematically by eqs. (15) through (21). Theorem 1 in thiS paper proves that the new algorithm has three propertics, expressed as eqs.(24) through (26). In addition the new algorithm is combined with the GPC law given in ourprevious work [3] to provide a new robust adaptive GPC algorithm, which is proved by us topossess robust stability. This proof requires the use of Theorems 1 and 2 presented in thispaper.