Hardware-in-the-Loop Testing Platform of a Predictive Controller Applied to a Simulated Regrinding Circuit

摘要

Real-world testing of advanced controller applications in a mineral processing plant can be complex and expensive and generate process instability. The hardware-in-the-loop (HIL) simulation solution allows testing of the behavior of a plant being controlled by advanced controllers in a virtual environment before being applied in the real plant. The objective of this paper is to create a hardware-in-the-loop testing platform for the study of the application of a model predictive controller (MPC) in a simulated regrinding circuit of an iron ore processing plant of Samarco, located in Mariana, Brazil. The predictive controller is available in the distributed control system (DCS) of the studied plant. A present problem in milling process is excessive milling of the material, which increases the power consumed, hinders the recovery of very small particles in the concentration steps and impairs the filtration of the pelletizing process due to clogging caused by material smaller than specified. To improve the milling performance and reduce its operational cost, it is necessary to study adequate control strategies that control better the various variables in this process and allow the control of variables that are not usually used in automatic plant control. Most of these circuits are controlled only by PID controllers. With the HIL platform, it was possible to simulate actuators and sensors, in addition to allowing the development of a supervisory model predictive controller (MPC) to generate remote setpoints (SP) for the PID controllers using the industrial control blocks built into the DCS. The correct development has made the HIL platform efficient, and the results of the simulation of the MPC application showed an improvement in the particle size control of the final product, which is the main variable of the regrinding process.