Reliability analysis of telescopic arm of pieline-catching vehicle based on semi-supervised deep neural network
YUAN Guozhi
LIU Wei
YAN Zilong
ZHANG Ruilin
ZHAO Mingxuan
SANG Jianbing
Abstract：The telescopic arm,a pivotal component in the pipeline grabbing vehicle,links the lifting platform and the mechanical claw,shouldering the majority of the load.Conducting a reliability analysis is imperative.Traditional methods for reliability face challenges like high computational costs and low accuracy dealing with multidimensional uncertainties.To overcome these,our study proposed an engineering mechanical reliability analysis method,leveraging Adams dynamic simulation,semi-supervised learning,deep neural networks,and Monte Carlo method.In this study,a virtual prototype model of the pipeline grabbing vehicle was established,identifying hazardous operating conditions.Combining the telescopic arm model's geometric parameters and overall structure,uncertain factors influencing the maximum von Mises stress were determined,conducting a sensitivity analysis was conducted.Utilizing optimal Latin hypercube sampling based on uncertain parameter distributions,Ansys Workbench was employed to build a finite element model,obtain output results for the sample size.Semi-supervised learning processed the finite element simulation data,enhanced deep neural network training accuracy.Finally,based on the fourth strength theory,a failure criteria for the telescopic arm component was determined.Combining deep neural networks and Monte Carlo method,the reliability and failure probability were predicted.Results show that this method surpasses actual engineering precision requirements,provides a certain guiding significance.
Keywords：Telescopic armReliability analysisSemi-supervised learningDeep neural networksOptimal Latin hypercube sampling
Publication Date：2025-08-15
Online Publishing Date：2026-09-12(First online date of this platform, not the publication date of the document)
Pages：9( 159-167 )