An Online Updating Robust Soft Sensor for Nonstationary Industrial Processes Based on Bidirectional Long Short-Term Memory and Integrated Gradients
This study proposes an online-updating robust soft sensor framework based on bidirectional long short-term memory (BiLSTM) with integrated gradients (IG) and smoothed quantile loss (SQLoss) that is consistent with field experience, offering valuable insights for optimizing the desulfurization control system.