| Summary: | Event-triggered control is applied to adaptive dynamic programming (ADP) as an effective method to reduce the computational cost, in which sampling only happens when a specific event occurs. In several cases of industrial applications, however, the sampling rate should be reduced in the given ranges where the prior knowledge of the process has been obtained. Correspondingly a greater sampled data are needed to observe the change of the control system more carefully if the controlled variables lie outside the boundary of the desired region. In response to this industrial demand, a novel adaptive sampling strategy according to a priori knowledge for different system states is proposed to reduce the sensitivity of the ADP-based methods. To implement the strategy into nonlinear continuous-time systems, an adaptive sampling condition is given. Furthermore, the stability analysis for the close-loop system is explicitly provided by using Lyapunov approach. The experimental results on the rare earth extraction process (REEP) verify the effectiveness of the proposed method.
|
|---|
