Distributed Data-driven Model Predictive Control for Air-ring Systems with Application to the Blown Film Industry

The blown plastic film industry is constantly evolving, and the simpler products of just a few years ago are being displaced by new technical films. Market pressures have demanded a lower cost of plastic film with higher quality and thinner gauges. As a result of these trends, the demand for improvements in gauge control of the film increases dramatically, especially for manufacturers with older generations of equipment. The proposed MITACS project named “Thickness AI” will develop a distributed learning-based model predictive control (MPC) algorithm to improve film quality and save energy consumption. More precisely, this project will develop a data-driven method to learn the system model of the air-ring system in the blown film product line based on the offline collected input-output trajectories. Then, the distributed MPC (DMPC) algorithm is designed for the air-ring system to improve the gauge and uniformity of the film. The success of this project will not only fulfill the control objectives of the film but also extend the life of the manufacturing equipment.

Faculty Supervisor:

Yang Shi

Student:

Partner:

Macro Engineering & Technology Inc

Discipline:

Engineering

Sector:

Manufacturing

University:

University of Victoria

Program: