부산대학교 도서관

The advancement in vehicle-to-vehicle and vehicle- to-infrastructure technologies makes it possible for vehicles to obtain the real-time information related to transportation and traffic infrastructure. This paper presents the development of an optimal velocity generation algorithm that leverages the availability of traffic and road information. The objective of this optimization problem is to generate a velocity trajectory within a prediction horizon to reduce tractive force while monitoring the overall travel time required for the trip. The developed algorithm reduces energy consumption by avoiding wasteful driving maneuvers and utilizes the opportunities to recuperate kinetic energy with regenerative braking capability. This non-linear constrained optimization algorithm is implemented by an automatic control and dynamic optimization (ACADO) toolkit for real-time execution. The energy reduction is observed in the evaluation results obtained with a vehicle model for the 2nd generation of GM Chevrolet Volt, developed at Michigan Technological University. An experimentally validated vehicle dynamic model is used for the assessment of energy consumption and vehicle performance.