Interval type-2 fuzzy reservoir cerebellar model articulation controller design for antilock braking systems

• A novel ABS control scheme integrating IT2FRCMAC with WOA-based optimization. • New IT2FRCMAC design combining type-2 fuzzy logic, reservoir computing, and CMAC. • Adaptive laws derived from Lyapunov theory ensure stability and parameter convergence. • WOA optimizes learning rates, improving adaptation speed and tracking accuracy. • Simulations under dry, wet, snow, and ice confirm superior accuracy and robustness. The antilock braking system is a key safety component that enhances vehicle stability and steerability by preventing wheel lockup during braking. Designing a reliable antilock braking system controller is, however, a nontrivial task due to highly nonlinear braking dynamics, parameter uncertainties, and rapidly changing road conditions. This study introduces a novel control framework based on the interval type-2 fuzzy reservoir cerebellar model articulation controller. By embedding a reservoir mechanism into the type-2 fuzzy cerebellar model articulation controller structure, the proposed controller improves nonlinear approximation capability while effectively addressing system uncertainties. Adaptive learning laws are derived using Lyapunov stability theory to guarantee convergence and robustness. The choice of learning rates in the adaptation process strongly influences the overall performance. To automatically determine these rates, we incorporate the whale optimization algorithm, a bio-inspired metaheuristic that models the bubble-net hunting strategy of humpback whales. Whale optimization algorithm provides an efficient balance between global exploration and local exploitation, enabling the controller to achieve optimal adaptation speed and accuracy. Simulation results under diverse road surfaces demonstrate that the proposed controller with whale optimization algorithm optimized learning rates maintains the wheel slip around its optimal region, reduces tracking error, and suppresses oscillations in the control input. Compared with conventional approaches, the method offers enhanced robustness and adaptability, making it a promising solution for advanced active safety systems in modern vehicles.