A Novel Approach to Reduce Arrival Interval at High-speed Railway Stations

By Samudrapom DamReviewed by Susha Cheriyedath, M.Sc.Oct 17 2023

In an article recently published in the journal Mathematics, researchers proposed an integrated operation method to reduce the arrival interval through train operation and route setting optimization.

Background

In recent years, the high-speed railway has witnessed rapid development and has become a crucial passenger transportation mode. However, the frequent occurrence of track occupancy conflicts when trains enter a station due to emergencies, such as storms and rains, is a major problem. The succeeding train drivers are forced to halt their trains as they remain unaware of these conflicts, leading to schedule instability and longer train following intervals.

The failure of the preceding train to complete the operation plan as per the scheduled timetable is one of the major single train delay reasons at the station caused by disturbances, leading to the entering station conflict, which influences the moving authority of the following train.

The following train has to halt before the home signal and await entry clearance, resulting in extended train arrival intervals and avoidable delays. Thus, effectively adjusting the train operation strategy to compress the arrival interval during station conflicts is crucial.

The proposed method 

In this paper, researchers proposed an integrated operation method that can compress/reduce the arrival interval for the entry conflict scenario by avoiding unnecessary stops in front of the home signal and increasing the train running/passing speed through the throat area.

This two-step optimization method combined mathematical–theoretical analysis and intelligent optimization algorithms. The mathematical analysis algorithm was used for the entry operation strategy determination, while the optimization algorithm was used for optimal adjustment trajectory calculation.

In the first step, the recommended approaching speed and position were obtained by analytical calculation, while in the second step, the speed profile from the current position to the position corresponding to the recommended approaching speed was optimized using intelligent optimization algorithms.

The recommended trajectory can be solved by the improved optimization algorithm after obtaining the adjustment time within a short period. Researchers evaluated the feasibility and validity of the proposed method through a field test and a simulation experiment.

A three-dimensional (3D) problem model was constructed and the optimal operation scenario and the normal operation scenario were replicated on a distance–time–speed spatial rectangular coordinate system to describe the succeeding/following train operation adjustment problem due to entering station conflict.

The operation control optimization considering the home signal opening time was divided into two stages. In the first stage, the optimal adjustment point was obtained through entry operation strategy analysis, and then, the meta-heuristic algorithm was used to design the train trajectory before the adjustment point.

Three entry operation strategies were proposed based on Pontryagin’s maximum principle, including maximum acceleration (MA), maximum acceleration–maximum braking (MA-MB), and maximum acceleration–cruise–maximum braking (MA-CR-MB).

In the second stage, both the Q-learning algorithm in reinforcement learning (RL) and the meta-heuristic algorithm were improved based on the constraints and objectives and applied to solve the optimal adjustment trajectory of the succeeding train based on the operation strategy calculation.

The meta-heuristic algorithm is a crucial optimization strategy that can solve complex optimization problems by simulating the search behavior of nature using a core of swarm intelligence algorithms. This study used the classic particle swarm optimization (PSO) algorithm of heuristic algorithms and the popular grey wolf optimizer (GWO) algorithm to solve the problem. Q-learning is a classic value-based RL algorithm where the Q value represents the expected cumulative reward in a given action and state.

Experimental evaluation and findings

Two distinct case studies were used to experimentally verify the proposed integrated operation method. The first case study used data from the Beijing–Shanghai high-speed railway line, while the second case study was based on the field test on a railway test track in Beijing.

Researchers set a simulation scenario for the first case study, with the line parameters utilized in the simulation including station speed limit, line speed limit, train traction acceleration, train braking deceleration, initial train position, the adjustment time, and the length of line, throat area, and station. The line speed limit and the speed limit in the station were 300 km/h and 80 km/h, respectively. Additionally, the succeeding train starting adjustment point/beginning position of the operation optimization was set to the start of the simulation line.

The MA-MB pattern displayed the best performance based on the time saved the three entry operation patterns. The original running time of the line was 106.2 s, and the pattern saved the entry running time by up to 12.5 s, which translated into an 11.7% reduction in entry running time.

In the second case study, researchers considered various stopping times of the delayed train, excluded the influence of the train and line parameters on the optimization model, and observed the change in entry running time saved by the proposed method. The integrated operation method effectively saved 30% and 14.2% of entry running time in the first and second case studies, respectively, leading to a shorter arrival interval. To summarize, the findings of this study demonstrated the feasibility of using the proposed method to mitigate controllable conflict events occurring at the station.