The invention relates to a running diagram compilation method based on a GRU neural network. The method comprises the following steps: predicting a predicted value y of the passenger flow in a future predetermined time period based on a GRU neural network; judging whether the predicted value of y meets the current operation requirement or not, and if so, carrying out operation according to the current actual operation diagram; and if the predicted value of y cannot meet the current operation requirement, predicting a predicted value of y+N again through the GRU neural network, and dynamically adjusting a train working diagram; wherein N is a positive integer. According to the method, the train operation interval is dynamically adjusted by calculating the passenger flow in a plurality of periods in the future, the train operation frequency is reduced, the train operation energy consumption is reduced, and the operation efficiency and the passenger taking experience are greatly improved.
本发明涉及一种基于GRU神经网络的运行图编制方法。本发明基于GRU神经网络预测出未来预定时段内的客流量的预计值y;判断y的预计值是否满足当前运营要求,若是,则按照当前实际运行图运营;若所述y的预计值不能满足当前运营要求,则通过GRU神经网络重新预测出y+N的预计值,并动态调整列车运行图;其中,N为正整数。本发明通过计算未来数个周期内的客流量,动态调整列车运营间隔,减少了列车运行次数,节省列车运行能耗,并且大大提高了运营效率以及乘客乘坐体验。
GRU neural network-based running diagram compilation method and system
基于GRU神经网络的运行图编制方法及系统
2021-09-10
Patent
Electronic Resource
Chinese
Ad-hoc network train group-based working diagram compilation method and system
European Patent Office | 2023
|Train running simulation method and system based on running diagram
European Patent Office | 2023
|The Application of TUXEDO Middleware in a Network-Based Train Working Diagram Compilation System
British Library Conference Proceedings | 2009
|European Patent Office | 2023
|Highway running state evaluation method based on macroscopic fundamental diagram
European Patent Office | 2016
|