Fractional Order Controller Design using Biogeography Based Optimization Algorithm for Quadruple Tank Process

Abstract

Designing control algorithms that can handle multivariable and non-linear process with parametric variations is a challenging task. With the restrictions on memory and the processing power in the industrial controllers such as programmable logic controllers, the control algorithms are required to be simple in design, yet capable of handling disturbances and variations in the process parameters. This investigation presents a fractional order controller (FOC) design for a benchmark process called quadruple tank process (QTP) that exhibits multivariable and non-linear characteristics. To tune the FOC parameters, we use an evolutionary computing technique named Biogeography based optimization (BBO) algorithm. To illustrate the performance benefits obtained with the FOC design tuned by BBO algorithm, we compare it with the conventional PID algorithm. Our results indicate that FOC design provides better performance than the PID with improvements up to 42% in the transient performance and 29.6 % in the controller performance metrics.