Research On Intelligent Optimization Control Method Based On Operation Dispatching Of Train Group

With the development and improvement of China’s high speed railway, by the end of 2014, railway operating mileage of 112000 kilometers breakthrough, while high-speed railway operating mileage has exceeded 16000 kilometers, more than half of the world’s total operating mileage of high-speed rail, ranking first in the world. According to the state approved the implementation of the "long-term railway network planning", by 2020, China will form in the "four vertical and four horizontal" high speed rail as the backbone of the rapid railway network. The continuous improvement of the railway network brings new challenges to the train operation dispatching work.At present, train operation dispatching works base on the railway zone as a unit. This way of dispatching is to keep the dispatchers focused on the zone within the limits of their own duty, which cross-zone dispatching can be implemented using information exchange with other zones. But this approach can’t perfectly adapt to the needs of train group operation efficiency adjustment under the railway network condition. Therefore, it is urgent to study intelligent optimization methods to improve the train group dispatching efficiency, which is a realistic and complex system puzzle in the academic research in the area of railway transportation. The study of this thesis put the emphasis on the train group dispatching optimization and adjustment in the network condition from the following four parts: train group running path searching and distributing, train group operation sequence determination, multiple-target optimization of the train group operation control and adjustment based on coordinate-particle swarm. These four parts of the gradually progressive study compose a point, line, surface structure.The main research harvest of this paper includes:Firstly, train group running path searching and distributing. Based on summarizing the characteristics of our country’s train dispatching command system and relevant literatures, we put forward the train group dispatching connotation and relevant problem. According to the stochastic character of the spare capacity of the train running path in the railway network, we founded a stochastic-constrained train group running path distribution model which is then transferred to be a definite-constrained model.Secondly, train group operation sequence determination. Through analyzing conflict forms in the process of train group running and factors of arriving and leaving sequence, discussing the sequence problem of which train group passing through, arriving and leaving each station within a railway zone, we proposed an algorithm to calculate the train group running sequence and founded a model to solve the arriving, leaving, or passing time of each train on each station.Thirdly, multiple-target optimization of the train group operation control. We divided the problem of train group operation control into two levels to: one is the sequence of the train passing through each station and the other is the train running time. We simulated the scene of the train group operation dispatching control, designed the most optimal algorithm of train operation dispatching control based on Pareto, and proposed a higher efficiency intelligent optimal algorithm-particle swarm algorithm which is used to founded a multiple-target optimal train group dispatching control model.Finally, adjustment based on coordinate-particle swarm. Because the accomplishment of the train dispatching plan should base on the algorithm which has the higher degree of precision and processing-capacity, we studied the train group dispatching problem deeply, improved the basic-particle swarm algorithm, processed the coordinate-particle swarm algorithm which divides the particle swarm into three sub swarms, proposed three separate optimal targets according to each sub swarm, and founded the train dispatching model based on the coordinated-particle swarm algorithm.