Train station parking (TSP) accuracy is important to enhance the efficiency of train operation and the safety of passengers for urban rail transit, while TSP is always subject to a series of uncertain factors. To increase the parking accuracy, robustness and self-learning ability, we propose a new train parking approach by using the reinforcement learning (RL) theory. Three algorithms were developed, involving a stochastic optimal selection algorithm (SOSA), a Q-learning algorithm (QLA) and a fuzzy function based Q-learning algorithm (FQLA) in order to reduce the parking error in urban rail transit. Meanwhile, five braking rates are adopted as the action vector of the three algorithms and some statistical indices are developed to evaluate parking errors. Parking results show that the parking errors of the three algorithms are all within the ±30cm, which meet the requirement of urban rail transit.