Interative Learning Control (ILC) for Nonlinear System with Slow Variable Parameter and Noise, Applications for Wastewater Treatment Plantcontribution

This study focuses on developing algorithm and control structure for managing dissolved oxygen (DO) levels in tank 5 at a wastewater treatment plant using the Benchmark Simulation Model No.1 (BSM1). Initially, the paper outlines the principles of Iterative Learning Control (ILC), including its basic principles and its applications to nonlinear objects. It then introduces the wastewater treatment system and the standard BSM1 simulation model, a nonlinear object with slow variable parameters and noise. The emphasis is on proposing a control solution that combines feedforward Iterative Learning Control (ILC) with feedback PI control for the BSM1 wastewater treatment system. Results indicate that the control of dissolved oxygen levels in tank 5 has been significantly optimized, particularly as the number of learning iterations increases and the \({K}_{p}\) , \({K}_{d}\) parameters of the PD learning function are appropriately selected.