27 June 2017 in IEEE Transactions on Automatic Control
In this paper, we develop a novel extended state observer (ESO), in terms of tracking error only, for output tracking of a class of multi-input multioutput systems with mismatched uncertainty. A novel ESO is constructed from the nonsmooth function “fal” to estimate both uncertainty and state of the system. An ESO-based output feedback controller is then designed to compensate (cancel) the uncertainty and to achieve the output tracking. The convergence of the closed-loop system is proved. The effectiveness of the proposed method is demonstrated by numerical results of trajectory tracking for a practical autonomous underwater vehicle model. We show that in the presence of measurement noise, this novel ESO leads to better performance than the linear ESO. Moreover, this type of ESO has much smaller peaking value than the linear ESO under the same tuning gain.
12 October 2017 in IEEE Transactions on Fuzzy Systems
In this paper, event-triggered reliable dissipative filtering is investigated for a class of T-S fuzzy systems. First, a reliable event-triggered communication scheme is introduced to release sampled measurement outputs only if the variation of the sampled vector exceeds a prescribed threshold condition. Second, an asynchronous premise re-construct method for T-S fuzzy systems is presented, which relaxes the assumption of the prior work that the premises of the plant and the filter are synchronous. Third, the resulting filtering error system is modeled under consideration of event-triggered communication, sensor failure and asynchronous premise in a unified framework. By adopting the Lyapunov functional method and integral inequality approach, a delay-dependent criterion is developed to guarantee asymptotic stability for the filtering error systems and achieve strict $(Q, S,R)-\alpha$- dissipativity. Consequently, suitable filters and the event parameters can be derived by solving a set of linear matrix inequalities (LMIs). Finally, an example is given to show the effectiveness of the proposed method.
4 March 2013 in International Journal of Robust and Nonlinear Control
3 November 2015 in Journal of Dynamical and Control Systems
31 July 2020 in Nonlinear Dynamics
28 January 2021 in IEEE Transactions on Systems, Man, and Cybernetics: Systems
In this article, we apply the active disturbance rejection control (ADRC) to the output tracking of a class of lower triangular nonlinear systems subject to mismatched bounded stochastic disturbances of unknown statistic characteristics and nonvanishing at the origin. A major obstacle is that the paths of the stochastic disturbances are nowhere differentiable almost surely which causes that the stochastic disturbances cannot be refined into the control input channel by the usual way of state transformation. To overcome this obstacle, a set of second-order extended state observers is first designed to estimate, in real time, the disturbance in each channel, and then a backstepping ADRC based on feedforward compensation and a constructive backstepping procedure is developed, guaranteeing that the closed-loop output tracks a time-varying reference signal in practically mean square and the closed-loop states are practically bounded in probability first defined in this article. Finally, some numerical simulations are presented to validate the effectiveness of the proposed backstepping ADRC approach.