TWO-DEGREE-OF-FREEDOM PIDA CONTROLLERS DESIGN OPTIMIZATION FOR LIQUID-LEVEL SYSTEM BY USING MODIFIED BAT ALGORITHM

In this paper, an optimal design of the two-degree-of-freedom proportional-integral-derivative-accelerated (2DOF-PIDA) controllers for the liquid-level system based on the modern optimization by using the modified bat algorithm (MBA) is proposed. The MBA is the new modified version of the original bat algorithm (BA) developed from the echolocation behaviour of micro bats. To improve its exploration and exploitation properties, the random number drawn from a Levy-flight distribution and new loudness and pulse emission rate functions are proposed. Performance of the MBA over the BA investigated against 10 benchmark functions is presented in this paper. Results of the 2DOF-PIDA controllers designed by the proposed MBA for the liquid-level system are compared with those of the one-degree-of-freedom proportional-integral-derivative (1DOF-PID) controller designed by Ziegler-Nichols (ZN) tuning rule, 2DOF-PID designed by Araki-Taguchi (AT) tuning rule and 1DOF-PID and 1DOF-PIDA controllers designed by the MBA. As results, it was found that the liquid-level controlled system with the 2DOF-PIDA controllers designed by the MBA can provide very satisfactory responses superior to that with the 2DOF-PID controllers designed by AT and MBA and 1DOF-PID controller designed by ZN, significantly.