Development of Dye Exhaustion Behavior Prediction Model using Deep Neural Network

The textile dyeing process consumes a significant quantity of energy as it is necessary to maintain the water temperature between 60–120 °C during the dyeing of reactive dyes. Therefore, to reduce the overall cost of the process through reducing the quantity of waste energy, it is crucial to increase the right first time (RFT) rate, which corresponds to the rate at which the target color is imparted through a single dyeing process (Park et al., 2009). To improve the RFT rate, the proper operation with following the optimal dye exhaustion behavior in consideration of the color difference and dyeing uniformity is a critical factor. The color difference is determined according to maximal absorption and the dyeing uniformity is decided by dye exhaustion behavior [Bouatay et al., 2016]. In this study, we developed a model for predicting dyeing exhaustion behavior, and utilized the model to predict optimal dye exhaustion behavior under various dyeing conditions. A deep neural network-based on the dye exhaustion behavior prediction model was developed through regression analysis, the model was further developed and evaluated by dividing the entire dataset into learning and evaluation data. The model's performance was evaluated using the root mean square error (RMSE) parameter alongside the coefficient of determination (R²) which acted as performance evaluation metrics. Using these performance metrics, it was found that the proposed DNN regression exhibited the highest performance and the smallest error in comparison with established models, with root mean square error RMSE and R² values of 0.016 and 0.994, respectively. The results reported in this study demonstrate that the proposed model exhibits superior performance in predicting the dye exhaustion behavior.