Abstract: Unmanned aerial vehicle（UAV） have the potential to be deployed as vehicle data collectors in future intelligent transportation systems（ITS） due to their high maneuverability and low operational costs. To ensure the timeliness of the collected data, the freshness of the data received by UAVs from vehicles is measured in terms of age of information（AoI）. Therefore, this paper proposes a UAVassisted ITS scheme based on minimizing the Age of Information, by jointly optimizing the UAV’s trajectory and the association strategy between the UAV and vehicles, to minimize the weighted sum of AoI for all vehicles. The optimization problem is modeled as a MARKOV decision process（MDP）, and the curiosity-driven deep Q-learning network（C-DQN） algorithm is employed to solve it. Extensive simulation results demonstrate that this algorithm outperforms the traditional deep Q-learning network（DQN） algorithm in terms of exploration capability and performance gain.