Adaptive neural control of non-affine pure-feedback non-linear systems with input nonlinearity and perturbed uncertainties

Adaptive neural dynamic surface control of strict-feedback nonlinear systems with full state constraints and unmodeled dynamics

“…Many significant research results have been obtained (He, Dong, & Sun, 2015;Liu, Lai, Zhang, & Philip Chen, 2015;Qiu, Liang, Dai, Cao, & Chen, 2015;Ren, Ge, Tee, & Lee, 2010;Tee, Ge, & Tay, 2009;Tee, Ren, & Ge, 2011). Backstepping approach (Ge & Wang, 2002;Kanellakipoulos, Kokotovic, & Morse, 1991;Krstic, Kanellakopoulos, & Kokotovic, 1995;Zhang, Ge, & Hang, 2000) and dynamic surface control (Swaroop, Hedrick, Yip, & Gerdes, 2000;Wang & Huang, 2005;Yip & Hedrick, 1998;Zhang, Zhu, & Yang, 2012) were used to design the controller of nonlinear systems with constraints (Guo & Wu, 2014;Kim & Yoo, 2014;Liu et al, 2015;Meng, Yang, Si, & Sun, 2015;Qiu et al, 2015;Ren et al, 2010;Tee et al, 2009. Four adaptive control schemes were developed by using barrier Lyapunov function (BLF) for strictfeedback nonlinear systems with static output constraint or timevarying output constraint or partial state constraints and known ✩ This work was supported by the National Natural Science Foundation of China under Grants 61573307, 61473250 and 61473249.…”

Robust adaptive control with unmodeled dynamics and unknown dead-zones

“…(ii) By introducing first-order filtering, the proposed DSC method can avoid the circular arguments in traditional backstepping design based on fuzzy logic system or neural networks, compared with the existing results in [24]. (iii) The assumption with respect to the derivative of the virtual control coefficients is removed in [25,26]. (iv) By combining Young's inequality with RBF neural network technique, the assumption with respect to unmodeled dynamics is relaxed in [17,18].…”

“…In [24], adaptive controller was investigated for a class of strictfeedback systems which contains only one adaptive parameter that needs to be updated online. In [25,26], two adaptive DSC schemes were proposed for pure-feedback systems with dead-zones by approximation of multi-layered neural networks and RBF neural networks respectively, the dead-zone forms were different from each other. In this paper, RBF neural network-based adaptive control is investigated by combining dynamic surface control with adaptive control for a class of pure-feedback systems with unmodeled dynamics and unknown dead-zones.…”

Adaptive neural prescribed performance tracking control for near space vehicles with input nonlinearity

“…For the dead-zone problem, a new adaptive fuzzy output feedback control scheme was developed for a class of nonlinear MIMO systems with unknown nonsymmetric dead-zone inputs which are treated as system uncertainties in [29]. There are also some researches on integrating dead-zone with saturation [30][31][32]. In [32], a neural network (NN) adaptive control based on variable structure control (VSC) was investigated for uncertain nonlinear systems with nonsymmetric input nonlinearities of saturation and dead-zone.…”