Static game approach for solving lane-merging conflict between autonomous vehicles

Abstract

The right-of-way conflict such as lane-merging problem between Autonomous Vehicles (AVs) is an inescapable issue. Related studies formalize the problem using centralized decision-making models of “Reservation” or “Auction” from the perspective of intersection management, that are suitable only for the scenarios with a centralized intersection agent; and the fiat currency spent on bidding may trigger certain controversial issues concerning law or tax. This paper presents the first-stage results of a long-term work: (i) establishes a prototype of 2-player static game within distributed decision-making paradigm, to formalized and solve the lane-merging conflict between 2 AVs, in order to adapt to scenarios with or without a centralized decision-maker, i.e. intersection or road segment; (ii) designs the game's dynamic payoffs which result from the space-time status of AV rather than any financial currency, to avoid possible social controversies. Numerical results show that AV could promisingly make compromised decisions to avoid the potential deadlock of right-of-way conflict.