Search

Railway construction organization is a crucial part of modern railway operation system and the decision-making of construction plan is a complex and difficult problem. Compounding with computer science, information technology and group decision support system theory, a web-based distributed group support system, called Railway Construction plan Decision Support System (RCDSS) is designed and developed. The key issues of RCDSS are discussed including the system solution and the main framework, the data support service platform, the model base, the knowledge base and the inference engine. Based on computer-aided compiling of construction plan, RCDSS provides train schedule influence analysis, railway technical operation influence analysis, railway macro-network operation influence analysis and multi-criteria comprehensive evaluation for further decision support. RCDSS has been implemented in Shenyang Railway Administration in China in a Web-based environment and is accessible by authorized decision makers without geographical constraints. The result demonstrates the system enable better performance in railway construction decision.

the train scheduling is a crucial part of modern railway operation system. The train schedule optimization problem is exceptionally large-scale and hardly solved via the conventional mathematical method. In this paper, the double-track railway train schedule optimization is concerned. Based on the characteristics of the double-track railway operation system, a novel train schedule optimization model based on clonal selection differential evolution (CSDE) algorithm is proposed. In comparison to the traditional differential evolution method, the modifications include: (1) clonal selection mechanism with an auxiliary pool is introduced to provide the best individuals with more evolution opportunities; (2) an apoptosis process is employed in which worst individuals will be either extinct or regenerated in probability; (3) the population is in a dynamic balance. The novel method contains more control variables and can be converted into ordinary differential evolution problem with specific parameters setting. In a computational instance, the implementing result indicates that the proposed model and algorithm are practical and feasible for double-track train schedule optimization problem.