SYNTHESIS AND RESEARCH OF THE EFFICIENCY OF COMBINED REGULATORS FOR CONTROLLING ENERGY-CONSUMPTIVE PROCESSES

Abstract

The article presents the synthesis and a comprehensive evaluation of the effectiveness of combined controllers designed for controlling energy-intensive technological processes. As an example, the study considers the development and comparative analysis of modern automatic control systems for regulating the outlet temperature of an industrial shell-and-tube heat exchanger. It is intended for heating crude oil before it is supplied to an electrostatic desalting unit (EDU). Maintaining the temperature at 80 °C is critical for the efficiency of the desalting process, as deviations lead to a decrease in the degree of emulsion separation and an increase in demulsifier consumption. A simplified first-order distributed-parameter model with a time constant τ = 720 s is proposed. The model provides an adequate description of the object’s dynamics under significant disturbances in oil flow rate. Two control systems have been synthesized and investigated: a classical cascade PID controller with feedforward compensation and offset-free model predictive control (Model Predictive Control, MPC) based on this model. The simulation was performed with a +45% step disturbance in oil flow. The results showed that the PID controller provides a maximum temperature deviation of - 8.4 °C with an Integral Absolute Error (Integral of Absolute Error, IAE) of 1820 °C·s, while the offsetfree MPC reduces the deviation to -1.8 °C (IAE = 378 °C·s), which corresponds to an improvement of 4.7 times. The proposed MPC has a computational complexity acceptable for implementation on industrial PLCs (calculation time < 60 ms). The results obtained confirm the advantage of MPC for objects with high inertia and significant disturbances. They can be applied in the modernization of heat exchanger control systems before the ELOU in oil refineries.