| Popis: |
In this paper, the authors aimed to control the concentration variable of the CSTR process by fuzzy scheduling. Here, the PID parameters of the controller are scheduled by fuzzy c-means (FCM) clustering approach. The input–output data of the scheduler are grouped into clusters, and an appropriate Fuzzy Inference System (FIS) is generated based on the clusters. Thus, the output of the scheduler changes the PID parameters, and hence, the control system adapts to the shift in operating point of the process. The proposed scheme shows an improved performance over a PID controller tuned based on single operating point. The simulation studies conducted on the differential equation model of the CSTR process depict the vital role played by fuzzy c-means clustering in optimizing membership function of the FIS that models the input–output data behavior. |
