Akademik Veri Yönetim Sistemi
Mechanics Based Design of Structures and Machines, cilt.54, sa.1, 2026 (SCI-Expanded, Scopus)
This study presents a novel two-layer Electronic Stability Control (ESC) system that integrates the learning-based capability of the Adaptive Neuro-Fuzzy Inference System (ANFIS) in its first layer as a yaw stability controller and the efficiency of the offline Robust Model Predictive Control (RMPC) in the second layer as wheel slip controller. As the tire-road friction coefficient has different values in various road conditions, ESC parameters have to be adapted to different road types. The ANFIS-based yaw stability controller is trained based on the independent PID controllers, each designed for a specific road surface. The second layer controller contains four individual offline RMPC controllers on wheels to keep the slip ratios at their optimal values to maximize tire-road friction. The generated corrective yaw moment from the upper-layer is distributed between wheels using torque distribution method to generate braking forces across the wheels to enhance the stability of the vehicle during maneuver, leading to oversteer or understeer. The proposed approach is evaluated using a seven degrees of freedom (7-DoF) vehicle dynamic model under high and low adhesion surfaces. Simulation results demonstrate that the integrated two-layer “ANFIS-offline RMPC” control structure of ESC tracks the desired yaw rate and sideslip angle properly and keeps the optimal slip ratios compared to conventional methods. Comparing the relative CPU times shows that the low computational burden of the offline algorithm makes the approach proper for real-time automotive applications. This combination of learning-based adaptability and an offline predictive control approach introduces an effective solution for vehicle stability management.