Multi-objective Optimization-Based Design of Robust Fractional-Order PIλDμ Controller for Coupled Tank Systems

摘要

In this paper, optimal control of coupled tank systems has been proposed using H∞ fractional-order controllers. Controller tuning has been posed as multi-objective mixed sensitivity minimization problem for tuning the fractional-order PID (FOPID) controllers and multi-objective variants of bat algorithm (MOBA) and differential evolution (MODE) has been used for optimization. Use of fractional-order controllers provides better characterization of dynamics of the process and their tuning using multi-objective optimization helps in attaining the robust trade-offs between sensitivity and complementary sensitivity. Both the FOPID controllers tuned with MOBA and MODE present robust behavior to external disturbance and the compared results show that MOBA-tuned controller presents efficient tracking of the reference.