Reducing Overshoot In Process Control For Efficient Recovery Of Natural Gas Liquid (Ngl) Using Model Predictive Control Scheme (Mpc)
Osondu I. ONAH
Department of electrical and electronic engineering, Enugu state university of science and technology
Patrick U. Okafor
Department of electrical and electronic engineering, Enugu state university of science and technology
Stella N. Arinze
Department of electrical and electronic engineering, Enugu state university of science and technology
Keywords: Cost Function, Flow rate, Genetic algorithm, Kalman filter, Open loop control, Turbo expander
Abstract
This presents a methodological approach at reducing overshoot in process control for efficient recovery of Natural Gas Liquid (NGL), using the MPC scheme. Reducing process variability to achieve energy-efficient recovery of NGL requires a robust control system. Process variables such as overshoot, rise time, settling time, steady-state error, etc., are controlled to improve system stability, speed of response and prevent mechanical oscillations in control systems. Here, the control problem was formulated in state space, in which a quadratic cost function was used. Kalman filter was used for state estimation with all information necessary to predict future plant behavior. The quadratic cost function was achieved using a genetic algorithm to obtain the optimal control signal. The control system was modeled and implemented in Simulink. A simulation was carried out using the proposed scheme and Proportional Integral Derivative (PID) controller to determine improvement. Results showed that the adaptive MPC scheme outperformed the PID controller at set point following and also, the MPC controller achieved an average of 46.31% overshoot than the PID controller in the reduction of variability in inlet gas flow to the turbo expander