Two bi-level models to reconstruct origin-destination (O-D) demand under congested network are explored in terms of the observed link and route travel times, where one model inputs the known trajectories of observed route travel times and the other model uses both known and unknown trajectories of observed route travel times. The proposed models leverage both the link and route traffic information to determine the network O-D demand that minimizes the distances between the observed and estimated traffic information (O-D, link and route travel times) in the upper-level, and optimize the stochastic user equilibrium (SUE) in the lower-level. Meanwhile, the observed information of travel time can capture the relationships between traffic flow and travel cost/time in congested network. The K-means (hard assignment) and Gaussian mixture model (GMM, soft assignment) clustering methods are presented to identify the trajectories of observed route travel times. An iterative solution algorithm is proposed to solve the built O-D reconstruction models, where the method of gradient descent, the method of successive average and Expectation-Maximization (EM) algorithm are used to solve the upper-level model, lower level model, and GMM, respectively. Results from numerical experiments demonstrate the superiority of the travel time based model over the traditional flow based method in congested traffic network, and also suggest that using both the route and link information outperforms only using link information in the reconstruction of O-D demand.