Optimizing calcining kiln operations: A multivariable GTS integrated model for predictive-driven quality control

The inherent inability to obtain real-time quality measurements in calcined kilns poses significant challenges for maintaining optimal production stability. This research introduces an intelligent control method for calcined kiln product quality, leveraging a multivariable integrated deep learning model to predict and regulate quality by manipulating production actuators. A feature screening process was conducted by combining grey relative correlation, feature extraction was performed using Time Series Convolutional model, quality classification was carried out by Support Vector Machine model, and the GTS model was constructed to predict the quality of the calcined kiln product. An intelligent fuzzy control system was established to fine-tune operational parameters based on quality forecasts, with the goal of direct quality. Experimental validation using industrial furnace data demonstrates the effectiveness of the model, achieving 87.55 % prediction accuracy, and confirms the operational viability of the control strategy. This study provides a scientific approach to managing calcining kiln quality, enhancing production efficiency and facilitating the implementation of intelligent control in calcining operations. • GTS fusion model. • 87.55 % accuracy in predict quality. • Intelligent control strategy.