Ke Zhang wins the Best Paper Award of CICTP 2020-21
2021-12-28
source:unknown
clicks: 578 作者:未知
On December 19, 2021, Ke Zhang’s paper, “Reinforcement Learning for Shortest Path Problem on Stochastic Time-dependent Road Network”, won the best paper award of the 20th and 21st Joint COTA International Conference of Transportation Professionals. The conference received more than 1700 papers and awarded 15 Best Paper Awards, with an award proportion of about 0.9%.
Finding a shortest path between two locations on a stochastic time-dependent road network is an important constituent in vehicle guidance system. However, it is difficult for traditional heuristic algorithms to handle the complexity and stochasticity within the road network. In this paper, the stochastic time-dependent routing problem is modeled as a Markov decision process and utilize several reinforcement learning methods to solve this problem, such as Sarsa, Q-learning and Double Q-learning method. Sarsa method uses the actual Q-value for iteration instead of the maximum value function used by Q-Learning, while Double Q-learning utilizes two estimators to compute the value function, which can overcome the shortcoming of overestimation. Evaluated on ten stochastic time-dependent road networks, it can be concluded that Double Q-learning method outperforms other methods.
010-62785569