A game theory-based controller approach for identifying incidents caused by aberrant lane changing behavior

Abstract

Aggressive driving is a key contributor to traffic incidents which deteriorate traffic flow, increase traffic congestion, and pose serious threats to driver and passenger safety. This paper presents a methodology for the estimation of driver aggressiveness and detection of traffic incidents using a game-theory based controller. We first present a game theory-based controlling mechanism, in which a witness vehicle (vehicle A) interacts with an aggressive vehicle (vehicle B) to estimate the aggressiveness of and to predict the future behavior of vehicle B. Second, we use a probe vehicle framework to detect incidents. Third, we apply shockwave theory to identify the location of the incident. Results show that the proposed method can estimate the aggressiveness of vehicle B with a high degree of accuracy. Numerical results obtained through simulation show that the proposed method obtains a better incident detection rate with more than 90% of the incidents detected, on average, with a nearly 91% classification rate and lower false alarm rate than three commonly used methods. It also requires less time to clear the traffic incident. The information obtained from the proposed system can be used to reduce traffic accidents caused by aggressive driving, thereby improving the safety of both drivers and passengers.