System identification and conditional control for an optimal operation of a pilot plant binary distillation column

Abstract

In this paper a new control structure has been proposed that minimizes the actuator energy by optimally reducing the error at steady state of closed loop response. A dynamic switching phenomena is employed such that effective actuator input manipulates the process variable. Parameter identification using PRBS input signal is carried out for MISO and MIMO processes. Pilot scale binary distillation column is considered for the study with tray temperature as variable of interest, heater voltage and reflux flow rate as its manipulated variables. The usage of proposed control structure with existing controller improves the system performance and reduces the operational cost. To depict the efficacy of the control structure, system is subjected to uncertainty by perturbing plant parameters of 30% from their actual values. Result shows significant improvement in actuator energy consumption. Control methodology has been experimentally validated on both MISO and MIMO schemes.