Variable Servo Characteristic Brake System Matching and Implementing Method Based on Driving Style Identification

This article proposes a variable servo characteristic matching and implementing method based on driving style identification for the self-designed electrobooster (Ebooster) brake system. First, 103 drivers were invited to collect data under the typical urban following drive cycle, and 11 characteristic parameters related to driving, braking, and environmental factors were selected to identify driving styles combined with factor analysis, Gaussian mixture model, and neural network, which had high identification accuracy and simple model configuration. Then, the indices of variable servo characteristic brake system (VSCBS) were extracted based on a brake four-phase map and matched the relationship between driving styles and brake servo characteristics. Next, a variable servo characteristic control strategy was presented for Ebooster by transforming the complex mechanical properties changes under the millimeter-level deformation of the reaction disk into a calibration diagram with a simple and easy-to-understand structure. Finally, the hardware-in-the-loop and real-vehicle tests results show that the deviation between the brake servo characteristics of the VSCBS and the driver’s expected value of cautious, normal, and aggressive driving styles are all controlled within 3.5%, manifesting that the proposed VSCBS can help drivers with different styles match appropriate vehicle braking characteristics.