Study On The Optimization Of Train Operation And Control

In recent years,Chinese high-speed railway has developed rapidly,its running distance and scale continue to become greater.At the same time,people have a higher requirements for characteristics in a train running process such as security,energy-saving,punctuality,etc.Therefore,it’s meaningful to study on train running process and optimize control strategy.Aiming at the optimization of train running process,based on previous works relative,this paper intends to optimize train running control strategy,and improves performance indexes of energy consumption,punctuality,stopping precision and passenger comfort while ensuring running security.Train running process is complicated and non-linear,traditional methods are difficult to achieve good results.Intelligent optimization algorithm offers a feasible way to solve this kind of problem,its application in train running optimization problem can be explored.Thus,this paper introduces intelligent optimization algorithms into the optimization of train running control strategy.The content of the paper is carried out from the following aspects.Firstly,train running single-particle model is built based on mechanical energy theory and train traction calculation theory,length of train is then taken into consideration to improve single-particle model,which makes the model conform better to actual running situations.Afterwards,performance indexes that need to be optimized and constraints are defined,the problem model of train running control optimization is summarized.Next,two population based intelligent optimization algorithms that need few parameters and are easy to programming,firefly algorithm and cuckoo search algorithm,are applied to the problem above.According to their simulation results,a hybrid algorithm combining those two algorithms called improved firefly-cuckoo search algorithm is proposed,which incorporates the characteristics of fast convergence rate and strong local search ability of firefly algorithm with the advantages of wide search range and not easy to be trapped in local optimum of cuckoo algorithm.Weighted sum method is applied to deal with multiple optimization objectives of train running problem,a control strategy is regarded as a solution.Firefly algorithm,cuckoo search algorithm and their hybrid algorithm are applied to optimize the problem.Besides,multi-objective optimization theories that include dominant relationship and leader select strategy method are introduced to improve the hybrid algorithm in order to eliminate the influence on result caused by different weight coefficients while using weighted sum method.Then,based on an actual test case and simulated in MATLAB,this paper analyzes from three aspects,population distribution,fitness value and final solutions during optimization process,simulation results of algorithms show that the improved algorithm achieves better optimization of energy consumption,running time,passenger comfort and stopping position,solutions with greater fitness value are obtained,global and local search ability of the algorithm also are improved.Finally,with the help of simulation test platform of our laboratory,the train running control strategy simulation system is designed using C#programming language.A test case is designed based on Beijing-Shanghai high-speed railway line,optimization results of train control strategy verify feasible practical value of those algorithms in actual system,satisfied control strategies can be provided for train operation.