Fuzzy adaptive-event-triggered control for semi-linear parabolic PDE systems with stochastic actuator failures

Highlights：

• A novel two-sided looped-functional improved form the existing results is constructed in this paper. Unlike the traditional looped-functional or discontinuous LKF, the two-sided looped-functional (TSLF) proposed in this paper not only considers the information in [tp, t), but also takes full advantages from [t, tp+1). With the help of the proposed TSLF and some advanced inequalities, a less conservative result is obtained in this paper, and the examples show a comparison between our result and some existing results.

• To better fit the engineering practice, stochastic actuator failures are considered in our work. In comparison with the Markov jump failures, the proposed semi-Markov jump model allows sojourn-time to be non-exponential distributed, which includes the normal Markov jump failures model as a special case. In the simulation part, a Weibull distributed actuator failures model is introduced, which makes our research more practical.

摘要

•This work considers a networked control strategy based on pointwise measurements for fuzzy parabolic PDE systems with a novel AET communication mechanism. Unlike the distributed controllers, the controller designed in this paper only requires finite sensors and actuators. By defining a time-varying threshold that can be adjusted adaptively as needed, the AET control scheme introduced in our work can further reduce the communication frequency without destroying the dynamic performance, which will be demonstrated in the simulation part.•A novel two-sided looped-functional improved form the existing results is constructed in this paper. Unlike the traditional looped-functional or discontinuous LKF, the two-sided looped-functional (TSLF) proposed in this paper not only considers the information in [tp, t), but also takes full advantages from [t, tp+1). With the help of the proposed TSLF and some advanced inequalities, a less conservative result is obtained in this paper, and the examples show a comparison between our result and some existing results.•To better fit the engineering practice, stochastic actuator failures are considered in our work. In comparison with the Markov jump failures, the proposed semi-Markov jump model allows sojourn-time to be non-exponential distributed, which includes the normal Markov jump failures model as a special case. In the simulation part, a Weibull distributed actuator failures model is introduced, which makes our research more practical.