eduzhai > Applied Sciences > Engineering >

Lane-Change Initiation and Planning Approach for Highly Automated Driving on Freeways

  • king
  • (0) Download
  • 20210506
  • Save

... pages left unread,continue reading

Document pages: 6 pages

Abstract: Quantifying and encoding occupants preferences as an objective function forthe tactical decision making of autonomous vehicles is a challenging task. Thispaper presents a low-complexity approach for lane-change initiation andplanning to facilitate highly automated driving on freeways. Conditions underwhich human drivers find different manoeuvres desirable are learned fromnaturalistic driving data, eliminating the need for an engineered objectivefunction and incorporation of expert knowledge in form of rules. Motionplanning is formulated as a finite-horizon optimisation problem with safetyconstraints. It is shown that the decision model can replicate human drivers discretionary lane-change decisions with up to 92 accuracy. Further proof ofconcept simulation of an overtaking manoeuvre is shown, whereby the actions ofthe simulated vehicle are logged while the dynamic environment evolves as perground truth data recordings.

Please select stars to rate!


0 comments Sign in to leave a comment.

    Data loading, please wait...