Autonomous Lane-Change Controller via Mixed Logical Dynamical

This paper focuses on the design of a model based methodology to optimize lane change maneuvers based on the predictions of neighboring agents. In particular, the dynamics of vehicles are modeled as double integrators, and the lane change actions are indicated by boolean variables while neglecting the lane change dynamics. The objective of the optimal control is to minimize the travel time, maintain saftey distances between the target vehicle and neighboring ones, satisfy the operational constraints of the target vehicle, and comply with the speed limit. The control problem is formulated as a mixed logic dynamic system, and solved by Cplex, a commercial mixed integer optimization solver. Finally, the effectiveness of the proposed methodology is demonstrated by two simulated scenarios in this paper.