Event-Based Adaptive Control for Uncertain Nonlinear Systems With Actuator Nonlinearities and Experimental Verification

This article proposes a dual-channel adaptive event-triggered control scheme for nonlinear systems with actuator nonlinearities, unmeasured states, and external disturbances. First, a Nussbaum lemma is constructed to handle the actuator nonlinearities. Second, an radial basis function (RBF) neural network-based state observer and disturbance observer are established to estimate unmeasured states and unknown disturbances, respectively, and an event-based adaptive law is designed. Third, a new dual-channel adaptive event-triggered control scheme is proposed to minimize network resource consumption. The event-triggered control will cause discontinuity of the signals, which will make the differentiation of the virtual control signal undefined. Therefore, by utilizing the command filter, this issue is addressed, and the error between the command filter and the virtual controller is compensated by establishing the compensation signals. With the derived control scheme, all the closed-loop signals are semiglobal bounded, and Zeno behavior can be eliminated. Finally, the designed control scheme is verified on the piezoelectric positioning stages to demonstrate its effectiveness.