Iterative learning active disturbance rejection control based on extended state observer for batch process

To address the robustness issues in a class of batch processes characterized by uncertainty and non-repeatability, a proportional-derivative （PD） type iterative learning control （ILC） strategy based on an extended state observer （ESO） was designed. The ESO was employed to estimate the total disturbances acting on the system, and an indirect PD-type ILC law was applied to iteratively update the disturbance-rejection controller along the batch axis. This enabled joint optimization of the control law in both the time and batch domains. Simulation results on a batch reactor and a rubber internal mixing process demonstrated that the proposed ESO-based iterative disturbance-rejection learning strategy effectively suppressed external uncertain disturbances and improved the system's robustness and control accuracy.