Velocity and position forms of self-tuning Fuzzy PI controller applied to jacketed CSTR

Abstract

This paper presents an application of self-tuning Fuzzy PI controller (STFPIC) in velocity form (VF) and position form (PF) to control the jacketed Continuous Stirred Tank Reactor (CSTR) in a cascade configuration. CSTR is a highly non-linear process and requires non-conventional intelligent control techniques for stable and efficient operations. Self-tuning fuzzy logic controllers are one such option, usually employed for control of CSTR. STFPIC generates the gain updating factor according to the dynamics of the process, which modifies the output scaling factor. Two configurations have been investigated in cascade mode i.e., STFPIC in VF in primary loop and conventional proportional (P) controller in secondary loop, and in the other case STFPIC in PF in primary loop and conventional P used in secondary loop. Further, to tune the performance of the controllers a grey wolf optimizer has been utilized for obtaining minimum Integral of Absolute Error (IAE) for set-point tracking. The overall approach has been simulated in LabVIEW™ environment and a comparative study with the PID controller have been presented in this paper. Based on the intensive simulation it is found that, for set-point tracking, STFPIC in VF outperformed STFPIC in PF by 31.79% and PID by 69.40%, whereas for disturbance rejection STFPIC in VF outperformed STFPIC in PF by 26.37% and PID by 72.95%, in terms of IAE. Based on these investigations it is recommended that VF of STFPIC should be employed for efficient control of non-liner jacketed CSTR.