A study on structural deformation of platform screen door based on artificial neural network

• Published in: 2015 International Conference on Fluid Power and Mechatronics (FPM) pp. 370 - 373
• Main Authors: Zhifei Wang, Tingming Kan, Yu Wang, Yanfang Zuo, Dechun Xia, Dong Chen, Ansheng Sun, Pengfei Han
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01-08-2015
• Subjects: Artificial neural networks; Biological neural networks; BP neural networks; Neurons; Platform screen door; Rail transportation; Rails; Rigidity prediction; Training
• DOI: 10.1109/FPM.2015.7337142

In order to analyze the influence of flexural rigidity of platform screen door in cities, the BP artificial neural network is used, and through optimization of calculation method and training of the samples and after several efforts dereferencing of hidden layer and all hidden units the trainlm training function is selected and a artificial neural network for rigidity predication is established. The analysis results show that: the biggest relative error between the predicated value and measured value obtained through experiment of flexural rigidity of urban railway PSD is 6.7% and the artificial neural network model has higher prediction accuracy.