Automatic Train Operation Control Algorithm And Its Implementation In Urban Mass Transit Train

As an integral part of the urban transportation, rail transport system is facing growing population pressure and traffic demands. Advanced automatic traffic management system is urgently needed. Automatic train operation system, an important part of the rail traffic automation, can ensure the train run automatically with perfect energy saving effect under the safety and punctuality requirements. Automatic train operation is an inevitable trend of the development of future urban rail transit. Its research is of great significance.The main principle of Automatic Train Operation system (ATO) is very complicated. Firstly, ATO algorithm generate the train operation curve offline, with considering the ATO bound rules and optimized operation principles of the train’s running process. And then, ATO can control the train run according to optimized curve with the online control algorithm. This paper established a model of train traction and introduced train operation strategy, proposed a fuzzy predictive control for solving the real-time control problem of train operation, and finally build an embedded ATO vehicle platform.ATO control algorithm should have characteristics of small amount of calculation and the abilities of real-time control, subject to the computing capacity constraints of embedded platform. This paper used the fuzzy control rules to improve the shortcomings of online optimization in the predictive control, the dynamic matrix method to create predictive models and train feedback control link on the goal of comfort and accurate stop, and designed Reasonable membership functions and fuzzy control rules, built train control flowchart based on the combination of predictive and fuzzy control. On this basis, the ATO control algorithm is programed and a simulation is completed with the Xi’an Metro Line. By comparing with conventional PID controller, the comfort and parking indicators have improved greatly.Finally, the algorithm is translated into WinCE embedded operating systems with C++, which completed the primary building of the embedded ATO vehicle platform. Based on the HLA/RTI, the embedded ATO vehicle platform accessed to the train operation integrated simulation platform for distributed simulation.According to the result, Fuzzy Predictive Control can improve the accuracy of the train stop and traveler’s comfort index. Primary implementation of embedded ATO vehicle platform can work well with the integrated simulation platform, which reached the expected functionality. Furthermore, this platform provides a good foreshadowing for the development of the vehicle platforms and research of the train swarm optimization control.