铁道科学与工程学报2024，Vol.21Issue(7)：2699-2706,8.DOI:10.19713/j.cnki.43-1423/u.T20231607
稳定车-轨道耦合作用下轨排响应特性试验研究
Experimental research on response characteristics of track panel under stabilizer-track coupling
摘要
Abstract
As a kind of large-scale rail transit operation and maintenance equipment,the stabilizer is important to ensure the safety of railway line operation.But it have no ability to online percept lateral resistance of the sleeper,and it is difficult to control operation parameters intelligently in real time.In order to explore the method to online percept lateral resistance of the sleeper during stabilizer operation,the track panel dynamic test was carried out to measure lateral load and displacement of the track panel under stabilizer-track coupling on the test line.The lateral resistance of the marked sleepers were measured according to existing standard and in-situ test method.According to the measured test data,the function models were selected to fit the relationship between lateral resistance of the sleeper,lateral load and displacement of the track panel.Furthermore,an RBF neural network was constructed and trained by orthogonal least squares learning algorithm.The nonlinear model was built,and the accuracy of the function fitting and the RBF neural network model were compared.The results are shown as follows.In the test of lateral resistance of the sleeper,lateral displacement of the sleeper increases nonlinearly with the increase of lateral load on the test sleeper.In the track panel dynamic test,there is a nonlinear relationship between lateral load of the track panel,lateral displacement of the track panel and lateral resistance of the sleeper.The relationship obtained by function fitting reflects the nonlinear variation trend of lateral resistance of the sleeper to some extent,but the fitting error is large.The model based on RBF neural network has higher accuracy in test set verification.The off-line detection method of lateral resistance of the sleeper is inefficient,while the on-board sensing method based on RBF neural network model can improve the detection efficiency.The research results can provide a theoretical basis for real-time percept of lateral resistance of the sleeper and control operation parameters intelligently during stabilizer operation.关键词
稳定车/有砟轨道/道床横向阻力/动力学试验/径向基函数神经网络Key words
stabilizer/ballast track/lateral resistance of the sleeper/dynamic test/radial basis function neural network分类
交通工程引用本文复制引用
陈春俊,江浩..稳定车-轨道耦合作用下轨排响应特性试验研究[J].铁道科学与工程学报,2024,21(7):2699-2706,8.基金项目
国家自然科学基金资助项目(U2034210) （U2034210）
