Satisfactory Driving Mode Classification Based on Pedal Operation Characteristics

Abstract

Since the vehicle's response according to the driver's gas pedal operation varies greatly depending on the driving mode, the selection of the driving mode significantly affects the driver's satisfaction. This paper presents a satisfactory driving mode classification that enhances the driver's satisfaction by providing a suitable driving mode to the driver. Unlike the conventional algorithm based on driving style recognition, the proposed approach determines the changes required for the current driving mode, such as mode-up, -stay, -down. Features suitable for classification are extracted from the driver's pedal operation characteristics during specific situations, such as launch, acceleration while driving. The performance of the proposed algorithm is evaluated through nested cross-validation, compared with conventional algorithms based on driving style recognition, demonstrating its superiority, generality. The proposed algorithm is event-based, operates in real-time while driving. As a result, it provides a more reliable, effective solution for enhancing driver satisfaction by providing an appropriate driving mode.