Train Dispatching Models With Probabilistic Conflicts On A High-speed Railway Line

With continuous construction and operation of high-speed railway, the train speed increases, which greatly shortens people’s travel time. However, at the same time, the demand of people on travel time reliability becomes more and more prominent. High quality of train dispatching plays an important role in inproving the on-time rate of passenger trains.The optimization models of train dispatching are mostly based on the assumption that the information of conflict between trains is determined, and get schemes on the basis of this assumption. However, due to the uncertainty of the external disturbance information in the real world, the conflict between trains is uncertain, which lead to the limit of existing train dispatching optimal model.With the background of high speed railway, this thesis establishes a train dispatching model considering the uncertainty of the conflict between trains. And then an algorithm based on Lagrangian relaxation is designed. Finally, experiments are taken to verify the above models and algorithm in this paper.Specific research can be summarized as the following 5 parts:(1) In this part, the research achievements at home and abroad are analyzed, and the research status and development trend of this field are summarized as well. Aslo,this part defines the theoretical significance and application value of building train dispatching optimization model considering the probabilistic conflict.(2) This part analyzes the train dispatching optimization model considering the probabilistic conflict. In details, this part describes the meaning of probabilistic conflict, influencing factors of probabilistic conflict, the expression of probabilistic conflict, the characteristics and difficulties of the train dispatching problems considering the uncertainty of the conflict and some basic measures dealing with train conflicts.(3) Based on the train delay correlation analysis, this part fits train running state distribution curve, and establish the prediction model of train running state. According to the correlation coefficient between trains, the model of probablistic conflict between trains is established.(4) Based on ased on ased on ased on network flow theory and cumulative flow variable theory, a model to describe infrastructures on railway network is developed. Then this chapter builds a train dispatching model with probabilistic conflicts (PI) and train dispatching model with probabilistic and stochastic conflicts(P2).(5) In view of the limitation of commercial optimization software ong model sulotion, this paper designs a heuristic solution algorithm based on Lagrange relaxation. Then, this thesis uses performance data of Beijing Shanghai high speed railway as the background, mainly does the following experiment:① differnece between train dispatching model with/without probabilistic conflicts;② differnece on secenarios settings of train dispatching model with probabilistic conflicts;③ differnece between train dispatching model with probabilistic and/without stochastic conflicts.Through the above experiments, the feasibility of the proposed model is verified, and the high feasible solution quality of the heuristic algorithm is obtained. The following conclusions are obtained by experiments:① based on the actual network information and historical data, train dispatching model with probabilistic conflicts (PI) gets experimental results better than train dispatching model without probabilistic conflicts (P3);② with the increase of the number of scenarios, we found that the smaller the result of the upper and lower bounds of the objective function, the higher the quality of the train dispatching scheme is;③ With the continuous advance of the predicting time, the distance of the train’s location is closer to the end point. The upper bound and lower bound of the objective function is on a decreasing trend, and the quality of train operation adjustment scheme is improved continuously.