Research On Coordinated Train Control By Using Agent And Cellular Automata

In the21st century, increased traffic demand challenges the sustaina-ble development of urban rail transit system from inside, Energy-saving and ejec-tion-decreasing put pressure on the sustainable development of urban rail transit system from outside. In order to achieve sustainable development, urban rail transit system needs to further increase line capacity and reduce energy consumption on the premise that safety is guaranteed. Study on novel train operation control method is an important way to increase line capacity and reduce energy consumption. As a train operation con-trol method, coordinated train control method coordinates multi-train operation to op-timize system performance. It can limit train delay propagation and reduce energy con-sumption. Coordinated train control has become a hotspot for the research on train op-eration control method.For the purpose of research on coordinated train control, the dissertation proposes a train operation control model by using agent and cellular automata (CA). The agent and CA model are typical bottom-up models. Compared with the existing top-down models, the proposed model focuses on the interactive rules between trains. The proposed model becomes a powerful tool for design, simulation and analyses of coordinated train control. Then, an analytical based coordinated train control method is introduced to reduce train delay and a coordinated train control method is given to maintain headway regularity at bottlenecks in railway network. Finally, the effectiveness of the proposed methods is demonstrated through simulation based on the proposed train operation control model.The innovation of the dissertation is as follows:●A train operation control model with two-layer framework is proposed. The upper layer is an agent based train operation functional model and the lower layer is a CA based train-following model. Compared with the existing top-down models, the proposed model focuses on the interactive rules between trains. System functions are achieved and the dynamic of system state is de-scribed by using the characteristics of bottom-up model. The proposed model is suitable for the research on coordinated train control.●A train operation control model in complex environment is proposed in order to make model more realistic. From the point view of the number of platforms and the end-to-end communication delay, an improved station model and a model with considering end-to-end communication delay is given. The im-proved station model can simulate complicated phenomena of rail traffic flow, such as overtaking. And two different schemes for trains passing through sta-tion are compared. The model with considering end-to-end communication delay can be used for analyzing the impact of end-to-end communication delay on rail traffic flow. The improved model becomes a powerful tool for the re-search on coordinated train control.●An analytical based coordinated train control method is introduced. This me-thod uses advisory speed as guidance for delayed train entering station. An al-gorithm is proposed to calculate the minimum of running time for train enter-ing station. By using the proposed method, the delay propagation is absorbed and the train delay is reduced.●A cooperative control method for train coordination is proposed. The method is used to maintain headway regularity at bottlenecks in railway network by adjusting the travel time on each segment for each train. The computation time of the proposed method is analyzed and the analysis result shows that it can satisfy the requirement on real-time. The proposed model is tested and the consensus trend of headways can be observed through simulation. The simula-tion results demonstrate that the model can preliminarily solve the problem of train delay reduction.