Prediction of sintering burn-through point position based on Takagi–Sugeno fuzzy rules and interval type-2 fuzzy neural network

The burn-through point position during the sintering process is a key parameter that affects both the sintering quality and energy consumption. Due to the strong nonlinear coupling relationship among the characteristics of the sintering process, traditional soft measurement models have difficulty fully exploring the interaction among the features. Therefore, this study proposes a hybrid TS-IT2FNN model, combining Takagi–Sugeno fuzzy rules with an interval type-2 fuzzy neural network for high-precision BTP prediction. Enhanced features generated through fuzzy rules are integrated with original process variables in the neural network. The experimental results show that the Takagi–Sugeno-interval type-2 fuzzy neural network hybrid model achieves 0.0078, 0.0084 and 0.92 in mean absolute error, root mean square error, and coefficient of determination ( R 2 ) , respectively, with error ranges of [−0.25, 0.25]. The model outperforms both data-driven and mechanistic models, maintaining high accuracy and stability under disturbances, highlighting its suitability for real-time industrial sintering applications.