Prediction and Control Method for the Burning-Through Point in Sintering Process Based on Multi-Mechanism Time Alignment and LPV-TS Fuzzy Modeling

The sintering process features long delays and strong nonlinearity, making accurate prediction and control of the burn-through point (BTP) challenging. To address the problem of temporal misalignment among process variables, this study proposes an integrated framework combining multi-mechanism time alignment, LPV-TS fuzzy modeling, and predictive control. First, fast Fourier transform (FFT) and wavelet transform (WT) are used to extract periodic and disturbance features, and a fuzzy time-alignment algorithm with exponential time decay corrects asynchronous data. Then, a linear parameter-varying Takagi–Sugeno (LPV-TS) model with fuzzy c-means clustering dynamically adjusts parameters according to system states, improving nonlinear adaptability and interpretability. Finally, a fuzzy controller based on prediction error and trend realizes real-time speed regulation of the sintering machine. Simulation results show that the proposed framework effectively enhances prediction accuracy and system stability, reducing RMSE and MAE while improving R². This approach provides a unified, interpretable, and efficient solution for intelligent sintering, achieving precise BTP regulation, energy conservation, and process optimization.