Research On Real-time High-speed Railway Rescheduling Under Station Disturbance And Interval Disturbance Scenarios

The train will inevitably be disturbed by various emergencies during the operation process.How to minimize the impact of the disturbance on passengers and rationally plan the arrival and departure of the train at each station after the disturbance is of great significance.At present,Chinese railway enterprises dispatching and adjustment of trains under emergencies are all done manually by dispatchers relying on experience,while the actual decision-making time is short,and dispatchers are under great pressure.Using scientific and reasonable train rescheduling models and algorithms to solve the train’s arrival and departure reschedule plan can provide decision-making support for dispatchers,making decision-making more intelligent and scientific.In this paper,the train rescheduling is researched under the two disturbance scenarios of the disturbance at the originating station and the loss of the passing ability caused by the interval disturbance.The mathematical model of train rescheduling is established with the minimum the sum of the product of delays and speed levels of all trains at all stations as the optimization objective.By determined the trains affected by disturbance in advance,the problem scale is effectively reduced and the calculation efficiency is improved.In this paper,a firework algorithm based on the group loser out tournament-optimum replace selection(GLOTORS)is designed to solve the problem.This algorithm effectively ensures that the computational search resources are inclined to high-quality solutions through the competitive elimination of fireworks groups,which accelerates the convergence speed of the algorithm and improves the optimization effect.Taking the Beijing-Guangzhou high-speed railway as an example,numerical experiments are carried out under two interference scenarios.The solution results show that when the disturbance duration is within 120 minutes,the fireworks algorithm designed in this paper can solve a feasible rescheduling scheme within 2 minute,and the solution quality is relatively stable.Comparing the rescheduling schemes of Genetic Algorithm(GA)and Particle Swarm Optimization(PSO)in the same disturbance scenarios,the results show that the solution quality of the fireworks algorithm is the best,which proves the powerful optimization ability of the fireworks algorithm.In addition,by comparing 8 selection strategies derived from the two ideas,the results show that the " group loser out tournament-optimum replace selection(GLOTORS)" proposed in this paper has the fastest convergence speed and the best solution quality,which proves the superiority of this strategy.The research results can provide support for real-time train scheduling and adjustment decisions.