Fault Information Recognition for On-board Equipment of High-speed Railway Based on Multi-neural Network Collaboration

Author(s):  
Lu-Jie Zhou ◽  
Jian-Wu Dang ◽  
Zhen-Hai Zhang
2017 ◽  
Vol 873 ◽  
pp. 220-224 ◽  
Author(s):  
Young Chan Kim ◽  
Mosbeh R. Kaloop ◽  
Jong Wan Hu

The performance prediction of High-speed railway bridges (HSRB) is vital to detect the behavior of bridges under different train’s speeds. This study aims to design a prediction model using the artificial neural network (ANN) to assess the performance of Yonjung high-speed bridge. A short-term health monitoring system is used to collect the behavior of bridge with different high-speed train’s speeds. The statistical analysis is utilized to evaluate the bridge under speeds 165 to 403 Km/h. The evaluation of bridge and prediction model showing that the bridge is safe, and the ANN is shown a good tool can be used to estimate a prediction model for the displacement of bridge girder.


1996 ◽  
Vol 11 (2) ◽  
pp. 740-747 ◽  
Author(s):  
T. Dalstein ◽  
T. Friedrich ◽  
B. Kulicke ◽  
D. Sobajic

2013 ◽  
Vol 779-780 ◽  
pp. 731-738 ◽  
Author(s):  
Ke Xin Zhang ◽  
Jian Wei Yao ◽  
Ze Ping Zhao

The principal aim of this paper is to determine the reasonable design parameters of high-speed railway vibration attenuation. The orthogonal test method is used to design the test of ground vibration induced by high-speed train. Four main factors that impact the maximum ground vertical vibration level are selected, and different values are given to each factor, so 8 groups of combinations can be obtained by using orthogonal test technique. Each group test data of the maximum ground vertical vibration level can be obtained by conducting vehicle testing on-track. In this paper, the primary and secondary factors that impact the maximum ground vertical vibration level are determined by range analysis. Moreover, the neural network theory is used to establish a model of the ground vertical vibration level, and this model can be trained and verified by the test data. The impact factors can be predicted by the method of combining orthogonal test and neural network concerning the specified vibration limit, and the value of maximum ground vertical vibration level with the predicted factors meets the requirement of accuracy. The conclusions provide a valuable reference to the vibration attenuation design of the high-speed railway.


2012 ◽  
Vol 178-181 ◽  
pp. 1956-1960
Author(s):  
Xiao Yan Shen ◽  
Hao Xue Liu ◽  
Jia Liu

In order to scientifically decide the percentage of vehicle entering expressway rest area, based on analyzing the influencing factors relating to the percent of mainline traffic stopping, a BP neural network prediction model for it was put forward. Finally, The Xinzheng Rest Area (XRA) was taken as an example for verifying the feasibility of the prediction model and determining the influence degree of the Shijiazhuang-Wuhan high-speed railway on the percentage of mainline vehicles entering XRA. The result shows that the model had a high precision and reliability.


Sign in / Sign up

Export Citation Format

Share Document