The work described in this paper proposes a new strategy for personalizing lane change assistance systems (LCAS) in intelligent vehicles. This strategy takes into account the driver's behaviour and characteristics for decision making and path planning in manual driving mode based on the driver's lane change intention, in order to adapt and personalize the system's decision making characteristics as well as the generated path planning. The system decision making is based on real-time adaptive cost functions, these cost functions take into account the risk of the environment, the legal aspect, and a learning terms calculated from incremental stochastic gradient descent approach, and the trajectory planning is based on adaptive polynomial planning for lateral movement, where the trajectory is also adapted with the same approach. The results of the simulation, carried out with the SHERPA dynamic car simulator in real driving situations, for lane change scenarios, show the effectiveness of the proposed approach to adapt and personalise the decision and planning mechanisms, and to ensure the safety of the vehicle as well as to increase the driver's confidence and acceptability.
Personalized Decision Making and Lateral Path Planning for Intelligent Vehicles in Lane Change Scenarios
2022-10-08
2704018 byte
Conference paper
Electronic Resource
English
Lane split decision techniques for vehicle lateral control path planning
European Patent Office | 2024
|Lane change risk assessment and personalized lane change decision method
European Patent Office | 2023
|LANE SPLIT DECISION TECHNIQUES FOR VEHICLE LATERAL CONTROL PATH PLANNING
European Patent Office | 2022
|