Experimental Comparison of STR and PI Controllers on a Nonlinear Liquid-Level Networked Control System

This paper discusses the design, practical implementation, and comparison of the adaptive controller using the indirect self-tuning regulator (STR) method and PI controller for a liquid-level networked control system. The system model is identified by using the least squares (LS) method and design a PI controller tuned by Ziegler-Nichols method. Also, the system parameters are estimated online, using the recursive least squares (RLS) method to implement the indirect STR adaptive controller. With the help of the estimated parameters, the STR controller should be designed through pole placement. To compare the performance of the controllers, the dynamics of the liquid-level control system have been changed by placing an external heterogeneous object inside the tank. Adaptive controllers are used in systems with variable parameters to detect dynamic changes and control the system appropriately. The results show that the indirect STR adaptive controller performs better than the PI controller in both transient and steady-state responses.