With the advancement of the regional economic integration process, the inter-city passenger dedicated line plays an increasingly important role in the overall economic development of the region. Train operation program as an important part to link the railway operation organization and passenger flow,need to have a good passenger adaptability .Now,the research on the train program is mostly based on the given passenger flow, which is actually in the dynamic change. Therefore, it is necessary to study the optimization method of the inter-city passenger dedicated line.First of all, based on the retrieval and analysis of the problems related to train routing at home and abroad, this paper focuses on the analysis of the relevant factors of passenger flow demand and intercity train operation program, and analyzes the matching relationship among the elements. Based on this, a matching evaluation function is established to evaluate the matching scheme of train operation program and passenger flow.Secondly, through the analysis of space-time characteristics of passenger flow and feedback control theory, this paper puts forward a method to adjust the train operation program based on dynamic passenger flow. In this method,the program adjustment mode is divided into single train adjustment and overall adjustment. Cluster analysis and principal component analysis are used to analyze and evaluate the train travel program evaluation model. And as a basis for the definition of the train operation program adjustment factor,according to the adjustment coefficient is given the program to adjust the program to adjust the method. Then this paper analyzes the single train adjustment method.Finally, in order to make the optimization system of intercity train that based on dynamic passenger flow more complete,this paper studies the optimization model of train program, and establishes a multi-objective planning model with the minimum cost of railway operation and the minimum travel cost of passengers. And the genetic algorithm is designed to solve the model. |