Research On Train Tracking Operation Based On Fuzzy Predictive Control

Nowdays,with the rapid development of the two-way,large capacity and real-time communication technology between trains,how to make full use of the advantages to improve the efficiency of railway transport needs further research.So the train track operation under this condition is studied in this thesis.A new train tracking strategy is proposed and a fuzzy predictive controller is designed for the strategy.On the premise of guaranteeing safety,the existing given tracking interval time is reduced to enhance the railway operation efficiency in this strategy.The train operation environment is described more accurate by using more date which come from real-time communication between trains to analyze and calculate the running status and trajectory of the front train.A real-time interval time of train tracking index is presents in this thesis through analyzing the existing train tracking interval time and considering the real-time demand of tracking control.Then,the change of the index is analyzed during the train operation,especially in the case that trains through the temporary speed restriction zone.A new strategy is proposed for tracking train in the situation of real-time communication between trains.Firstly,by changing the tracking strategy,the speed of the train decreases as the distance decreases,and the change of the real time interval is slowed down,the minimum real time tracking interval is increased to improve the safety of train tracking.Then on the premise of guaranteeing safety,the existing given tracking interval time is reduced to enhance efficiency.In order to realize the train operation strategy that is proposed in this thesis,the automatic speed regulation function of ATO is designed.Fuzzy and predictive hierarchical control is used to design the expected speed curve and real-time control law.The fuzzy control is adopted in the upper layer,and the corresponding fuzzy rules are designed according to the train tracking strategy.The generalized predictive control is used to control the train speed,which has good robustness to model inaccuracy and is real-time online local optimization.In the process of train running,firstly,according to the real-time data,the corresponding control strategy is given.Then the expected curve is generated according to the control strategy.Finally,the real-time and on-line moving horizon optimization is used to control the train running according to the excepted curve.