Comparative Evaluation of PID and Fuzzy Electronic Stability Control via Differential Braking in a Double Lane Change Maneuver

Abstract

Electronic Stability Control (ESC) is central to preventing loss-of-control crashes, yet its on-road behavior still reflects design trade-offs between rapid error correction and ride comfort. This study builds a high-fidelity ESC evaluation platform by coupling a spatial vibration model with a two-track handling model in a CarSim–MATLAB/Simulink co-simulation. Two representative controllers—a classical PID and a fuzzy-logic design—utilize identical inputs (driver steering angle and measured yaw rate) and a common supervisory layer (a dead-zone around minor errors and left/right brake distribution logic). Interventions are applied as wheel-specific brake pressures; the fuzzy controller employs nine membership sets for yaw rate and steering and five for brake pressure. Performance is assessed in an ISO-style Double Lane Change at 40 km/h using trajectory RRMSE, yaw-rate IAE, control-effort IACA (from brake pressure), and peak lateral acceleration.
Both controllers keep the vehicle close to the reference path (RRMSE < 10%). The PID achieves faster yaw-rate convergence and smaller IAE but at the cost of higher control activity (larger IACA) and more pronounced oscillations, particularly in lateral acceleration. The fuzzy controller yields smoother, step-like brakeactions that reduce oscillatory behavior and control effort, while converging more slowly. These head-to-head results quantify a practical trade-off relevant to calibration: prioritize PID when rapid stabilization is paramount, or favor fuzzy logic when comfort and restraint of intervention are critical. The framework provides a reproducible basis for ESC benchmarking and suggests targeted tuning of PID gains and fuzzy rule bases, as well as extensions to advanced controllers (e.g., LQR, MPC), for improvedstability and comfort balance