The invention belongs to the field of railway
train operation dispatching and commanding, and relates to a novel
train operation intelligent adjusting method. According to the method, fault information in a disaster prevention alarm
monitoring system, temporary speed limiting information sent by a driving dispatching
desk and running diagram information of a
train under a fault are fully utilized;a minimum
operation time division calculation module is added in the RBC
database and is responsible for calculating the minimum
operation time division of the remaining intervals and the train operation curve with the minimum
energy consumption under the time division requirement; the result is fed back to the driving dispatching
desk; a running situation prediction module predicts the time foreach train to arrive at a subsequent
station, and the time, the speed and the position during running in the intervals; by taking the minimum
total delay time of the trains on the line as an optimization target, the departure sequence and time of each train based on a first-in first-out or
reinforcement learning method are determined, offline training is carried out continuously, and after the training is completed, the training parameters are stored in the model for intelligent adjustment of the train running diagram under the influence of the
delay and inhibition of
delay propagation on theline.