Data-Driven Rationalizability Analysis of Simple Platoon Games: A Revealed Preference View

We investigate the problem of platoon matching through the lens of revealead preference. Specifically, we consider a noncooperative game among a number of vehicles that decide to form or not to form a platoon on a single road. Given a collection of observed vehicles’ decisions, we provide tractable procedures to decide whether such decisions are rational under some game and some model of rationality, i.e., whether the decisions are rationalizable. We further show that for each model of rationality in our consideration, there exists a collection of observed vehicles’ decisions that are not rationalizable, showing that our model of decision making in vehicle platooning is not too general as to admit any vehicle’s decision as rational under some game and some model of rationality.