Vehicle Motion Control Using Data-Driven Varying Road Friction Map

Purpose and goal:

In this project, methods to handle real-world vehicle traffic safety problems due to the variation of road friction will be developed. This includes first the road friction estimation and prediction, concerning variation, uncertainty, accuracy and confidence. Secondly, we will explore the use of this information in the state-of-art vehicle motion control functions assisting the driver, manually or autonomously, to adapt the driving in a safe and non-intrusive way. This will contribute to improved performance of active safety and autopilot systems on board.

Expected results and effects:

This project aims to gain new knowledge which will bridge the gap between friction estimation and optimal vehicle motion operation, so that the quality measures of friction information become relevant for vehicle control. Based on the knowledge, we will deliver algorithms and tools upon which new, or improved driver assistance systems and autonomous driving functionalities can be based on.

The goal is to be able to reduce 50% of all fatal accidents happened due to the driver’s misjudgment of low road friction. Approach and implementation:

Road friction measurements will be acquired and annotated at different road and weather conditions, for data-driven method development in road friction estimation. Friction map criteria will be established as input to friction estimation. Motion planning and control will start when friction estimation is reached at its first application status.

Simulation environment will be developed when friction map is defined. Vehicle testing is planned to verify the control design with the integrated motion planning algorithm, aiming for a demo at proving ground.