Iterative Learning Robust Fuzzy Security Predictive Control for Nonlinear Batch Processes With Deception Attacks

With the rapid integration of the Industrial Internet of Things into modern manufacturing, nonlinear batch processes have undergone transformative development via cyber-physical systems. Large-scale deployment of networked control systems enables real-time coordination and intelligent resource allocation, while exposing batch processes to deception attacks that threaten control security and stability. Unlike continuous processes, batch processes feature time-varying control objectives and parameters within each batch, resulting in inherent state fluctuations or nonlinear characteristics that can mask attack behaviors. This phenomenon allows attackers to continuously inject false data, degrade control effectiveness, and potentially destabilize the system. To address this issue, an iterative learning robust fuzzy security predictive control strategy is proposed. This approach introduces a robust secure invariant set to constrain the states to within a safe region given deception attacks. Furthermore, attack-related information is incorporated into the iterative learning mechanism, enabling adaptive correction of the control inputs and improved control performance. The simulation results verify the stability and effectiveness of the proposed method.