Energy-saving Oriented Optimization Method For Automatic Train Speed Control In High-speed Railway

As the most sustainable transport model, railway transportation has been the artery of the national economy, the popular traffic tool in China and also the backbone for modern cities’operation. It is the critical infrastructure and the important basic industry of China, which plays an irreplaceable global support for the economic society development, the improvement of people’s livelihood and the national safety. Meanwhile, as a means of safe and reliable, fast and comfortable, high-speed railway systems have now become one of the first choices for people’s journey. Though the high-speed railway system performs much better on environmental benefits compared to highways, waterways and airlines, it is still a department with relatively larger energy consumption. With the increasingly seriousness for global energy problem, the huge energy consumption caused by railway systems will be the key issue that restricts the sustainable development for itself. Automatic train speed control technology can improve the operative efficiency, and optimize the operational energy consumption under constraints of safety, punctuality and comfort, therefore it gradually becomes one of hot research areas in the subject of high-speed train control.Based on the characters of practical high-speed train operation, this thesis studies on the train motion model and the estimation strategy for train state with uncertain parameters, aiming to develop an update solution of train motion model for both the procedures of on-line train speed trajectory planning and tracking control. An integrative energy-saving optimization approach for automatic train speed control is proposed based on the train motion model update solution, and then it is expanded to the application for multiple trains’movement. The innovations of this thesis are listed as follows.(1) A train motion status estimation algorithm is proposed for the application of automatic train speed control. The Hammerstein model is used to demonstrate the multiple-mass model of high-speed train motion, which is with non-affine and complex internal and external structure features. The state estimation framework that can provide hierarchical status prediction models for both on-line train speed trajectory planning and tracking control is established based on the federal filter theory.(2) A moving horizon on-line train trajectory planning approach with an adaptive weight allocation mechanism is developed based on the proposed real-time train status prediction model. The proposed approach can make full use of redundant running time to lower the energy consumption, and improve the flexibility and energy saving level of train speed trajectory planning procedure.(3) An integrative energy-saving optimization approach for train speed control is proposed based on hybrid train motion model, which makes breakthrough for traditional research methods that treated the train speed trajectory planning and tracking control process independently. The proposed control approach uses the unbiased estimation algorithm to update the single particle model for on-line train speed trajectory planning and the multi-particle model for train speed trajectory tracking control synchronously. This solves the problem that the planned objectives are always hard to achieve due to the different model coefficients using in train motion model and to further improve the energy-saving high-speed train operation control level.(4) A distributed energy-saving optimization algorithm of automatic speed control for multiple train movements is proposed. The directed acyclic graph is applied to describe the train speed control strategy model, and the uncertainty of train position estimation is incorporated into train speed control process. Based on partially observable Markov decision theory, a distributed collaborative optimization model for automatic train speed control is developed, while the algorithm flow is also designed.In this thesis, simulations and field experiments are both employed for validation, which demonstrate the effectiveness of the proposed energy-saving oriented optimal control approaches for automatic high-speed train operation. Theoretical results in this thesis will lay the foundation for the development and designing of automatic high-speed train control system in China.