Research On Strategies And Methods Of Platform Assignment Optimization For High-speed Railway Large-scale Passenger Station

China’s high-speed railway is playing an increasingly important role in the transportation system.With the development of technology and the improvement of train speed and intensity,high-speed railway system becomes increasingly sensitive to complex operating environments.Meanwhile,the need to adapt to diversified operating scenarios is also increasingly urgent.It is inadequate for the traditional single static scene analysis method and “optimization” research framework to improve the efficiency of station technical operations.As the core work content of station technical operation,the optimization of railway platform assignment is one of the leading research directions of technical operation automation in railway stations.Based on the operation practice of high-speed railway,it aims at the optimization of the railway platform assignment under certain and uncertain conditions.it studies the method of the allocation planning and reschduling of the railway platform assignment and the route in the throat area,which is oriented to operation efficiency and anti-interference ability based on mathematical programming theory and evolutionary algorithm.The main research contents and innovations of this dissertation are as follows:Firstly,a time-divided and multi-objective optimization model of railway platform assignment in different periods is constructed to adapt to the diversified operation scenarios.The optimization objectives of different railway platform assignment plans vary in different periods.An improved Non-dominated Sorting Genetic Algorithm II(NSGA-II)based on individual survival scores is designed by analyzing the multi-objective characteristics of the model.A heuristic algorithm is proposed to increase the number of feasible solutions during population initialization.Besides,the calculation of crowding distance is replaced by the individual survival score at a later stage of evolution to enhance the convergence and diversity of the solution by adding accessory chromosomes to the coding process to indicate the correlation between the railway platform assignment and the train routing.The chapter also uses the Taguchi method to adjust the parameters of the algorithm to improve the applicability of the algorithm to the problem.The simulation is carried out with the traffic data of the actual station.The results show that the time-division multi-objective optimization method effectively improves the robustness of the scheme on the premise of ensuring the balance of railway platforms.At the same time,the improved algorithm is superior to the NSGA-II algorithm in terms of solution quality and solution speed.Secondly,a train arrival delay distribution model is obtained by analyzing the randomness of the railway platform assignment plan using Kernel Density Estimation and Kolmogorov-Smirnov goodness of fit test(K-S test).On such a basis,the chance-constrained planning method is introduced to establish a stochastic optimization model of railway platform assignment that allows violations of constraints under a certain confidence level.Combined with stochastic simulation technology,an improved genetic algorithm based on stochastic simulation is designed to solve the problem.It helps in adapting to high-speed railway lines,which are more sensitive to random disturbance.In order to ensure the feasibility of the solution,the algorithm uses a stochastic repair mechanism and a sequential repair mechanism to deal with the infeasible solutions.Besides,it designs a selection operator based on sorting,crossover,and mutation operator for the problem to improve the optimality and convergence of the solution.At the same time,it deals with the constraints of the model containing random variables through random simulation.Experimental results show that the proposed method improves the anti-interference ability of railway platform assignment under random disturbance scenarios.Lastly,On the basis of the study of railway platform assignment planning,a rescheduling strategy based on model predictive control(MPC)is proposed.It uses the adaptive rolling horizon mechanism to adjust the allocation scheme continuously.Meanwhile,the railway platform assignment rescheduling models based on integer programming and chance-constrained programming are designed to describe the deterministic and random situation of the train arrival time,respectively.Aiming at the above model,this chapter designs a heuristic genetic algorithm with excellent real-time performance.Using the crossover mutation operator in the neighborhood,the convergence speed of the heuristic genetic algorithm is guaranteed.At the same time,the feasibility of the solution is guaranteed by designing the heuristic algorithm and using the static penalty function method to construct the fitness function to deal with the constraint conditions.Meanwhile,the calculation time consumption caused by the feasibility of the repair solution is reduced and the real-time performance is improved.The experimental results show that the predictive control approach based on the chance-constrained programming model has better stability and optimization effects.