Adaptive Soft Sensor Using Stacking Approximate Kernel Based Bls For Batch Processes

This article proposes an adaptive stacking approximate kernel based broad learning system to improve the accuracy and generalization of predictions for batch processes. The model combines the strengths of the Broad Learning System (BLS) and the Adaptive Stacking framework, while also addressing the uncertainty and computation time issues of the BLS algorithm. The model is shown to have strong nonlinear fitting ability, excellent generalization ability, and adaptive ability, making it suitable for industrial online applications.