Research On Multi-objective Optimization Of Train Automatic Driving Process

With the rapid development of economy,China's railway industry has entered the golden age of development.The improvement of the train running speed and the increase of the train operation density make the traditional manual driving mode difficult to meet the demand of the further development of the train operation.As a running model,automatic driving model improves the train running performance index while replaces driver,which is often used to replace the artificial model,and will become the future development trend.As a core subsystem of automatic driving mode,ATO(Automatic Train Operation)system is a key equipment to improve the efficiency of the train operation and increase the density of the train operation.Due to the complexity of the train operation process,the train running speed is usually subject to the line conditions,speed limit conditions,the driver proficiency,and a lot of other uncertain factors.The replacement of artificial driving not only can meet the demand of each performance index which in the train running process,but also able to avoid the man-made factors.Therefore,it is necessary to study the ATO system.Firstly,the structure,function and working principle of ATO system were introduced,and the idea of optimizing the operation process of train from the running curve and the controller of automatic driving was proposed;Secondly,for the optimization of train running curve,the smart NPSO(Niche Particle Swarm Optimization)algorithm was selected,and the solving the multi-objective model can get the ideal curve of train operation;in the aspect of the design of controller,combing the advantages of the grey prediction,fuzzy control and PID control,the core algorithm of ATO controller based on gray prediction fuzzy PID control was designed;Finally,according to the generation of the ideal running curve and the control effect of the ATO controller,the relevant simulation lines and models are selected,and the simulation analysis is carried out to verify the rationality and validity.By comparing and analyzing the train operation curve through the simulation,its result can realize the comprehensive optimization of various performance indicators,and it is proves that the NPSO algorithm can achieve the goal of optimizing train operation target curve.Taking the optimized train operation curve as the input and after the controller action,the error between the output curve and the input curve is lesser,so the train automatic driving controller,which core is grey predictive fuzzy PID control algorithm,can achieve good control effect.In conclusion,through optimizing both the operation curve and the controller of the train,the result can achieve the goal of optimizing train automatic operation process.