Work

Real-Time Evaluation of a Robust Adaptive Cerebellar Model Articulation Controller for an Uncertain Coupled-Tank System on OPAL-RT

Year: 2025
Type: article
Abstract: This paper addresses the challenge of maintaining a desired water level in a coupled tank system, a process that is inherently affected by nonlinearities, disturbances, and time-varying parameters. To overcome the limitations of model-based controllers, such as PID, a Robust adaptive cerebellar model articulation controller (RACMAC) integrated with sliding mode control is proposed as an effective solution. The research contributions include the development of a nonlinear disturbance estimator based on a cerebellar neuron model and a robust adaptive law that approximates both the ideal controller and the compensating controller. Methodologically, RACMAC uses an online CMAC structure to approximate complex nonlinear disturbances, with adaptive parameters continuously updated using Lyapunov stability theory to ensure global stability. The controller is evaluated through MATLAB/Simulink simulation and experimentally validated on the OPAL-RT OP5707XG real-time hardware platform. The simulation results show significant improvements in performance, including improved tracking accuracy, reduced overshoot, and better adaptability to parameter fluctuations compared to the FOPID and SMC controllers. The controller maintains stable water level regulation even under strong disturbances and uncertain operating conditions. Real-time experiments further verify its robustness and consistency with simulated responses. The simulation and experimental results demonstrate that RACMAC reduces RMS error by 16–17%, MAE by 15–17%, and average steady-state error by up to 30% compared to FOPID and SMC.
Language: English
Open Access status: diamond
OpenAlex ID: W7160946199
FWCI: 0.229
Cites: 60
Cited by: 1
Related to: 10