A Robust Self-Tuning Fuzzy Controller for Integrating Systems

We present a robust non-fuzzy self-tuning scheme for fuzzy PI controllers (FPIC). The output scaling factor (SF) of the proposed controller (NF-STFPIC) is continuously adjusted by an updating factor (beta) which is defined on the normalized change of error of the controlled variable and the number of fuzzy partitions of the input linguistic variables. Performance of the proposed controller is tested for a pure integrating plus dead-time (IPD) and first-order integrating plus dead-time (FOIPD) processes with a considerably large perturbation in dead-time under both set-point change and load disturbance. Performance comparison of NF-STFPIC with conventional Fuzzy PI controller (FPIC) and the well-known self tuning fuzzy PI controller (STFPIC) reported in the leading literature, is made in terms of several performance indices. Unlike STFPIC, which uses 49 expert's defined self-tuning fuzzy rules in conjunction with 49 control rules, our NF-STFPIC uses only 25 control rules and a single non-fuzzy self-tuning relation. Experimental results justify the effectiveness of the proposed NF-STFPIC.