Development of an Automation and Intelligent Control System for the Drying Process of Polyvinyl Chloride

This study develops a structured automation and intelligent control concept for suspension-grade polyvinyl chloride (S-PVC) drying. The proposed architecture combines industrial instrumentation, a model-assisted outlet-moisture soft sensor, online bias adaptation, and constrained model predictive control (MPC). A control-oriented drying model represents moisture removal by an effective rate coefficient and first-order outlet dynamics with transport delay. A reproducible four-hour numerical experiment compares a conventional PI controller with the bias-adaptive MPC under stepwise variations in wet-cake moisture. For the explicitly assumed nominal parameters, MPC reduced the calculated mean utility-demand proxy from 303.19 to 300.31 kW (0.95%) and reduced the proportion of simulated time above the illustrative 0.43% dry-basis moisture limit from 20.4% to 0.0%. A ±10% kinetic-parameter sensitivity test showed both the value and the limitations of online model correction. These are model-based findings, not results of industrial commissioning or laboratory validation. The paper specifies an implementation and validation route for production-scale deployment.