Temperature Field Prediction Model for Zinc Oxide Rotary Volatile Kiln Based on the Fusion of Thermodynamics and Infrared Images

Abstract

The timely and accurate measurement of temperature field is of great significance for the low-carbon and high-efficiency operation of the zinc oxide rotary volatile kiln (ZORVK). Due to the large axial length, closed internal space and complex reaction mechanism, it is difficult to measure complete temperature field data. In this study, a novel temperature field prediction model based on the fusion of thermodynamics and infrared images is proposed for the first time. First, a thermodynamic model involving complex chemical reaction heat is established. Then, an industrial infrared thermal imager is developed. In order to obtain high-precision solid fluidized bed temperature from infrared images, a temperature extraction algorithm based on infrared image processing and YOLOv5s is designed. Finally, a parameter optimization model is built by minimizing the error between the predicted temperature of the thermodynamic model and extracted actual temperature. Moreover, five common thermodynamic parameters applicable to any ZORVK are extracted, which reduces the optimization cost of unknown parameters. Results show that our proposed fusion model can effectively improve the prediction accuracy and optimization efficiency of the pure thermodynamic model.