Skip Navigation
Skip to contents



29권 11호 987-993 2021 [한국자동차공학회 논문집 ]

제목 다층 신경망을 이용한 통합형 Axle 캠버 횡강성 시험-해석 코릴레이션
분야 변속기 및 동력전달장치
언어 Korean
저자 조호선(서한기술연구소), 오병기(서한기술연구소), 김학영(현대모비스)
Key Words Back propagation(오차역전파법), Camber strength(캠버 횡강성), Correlation(상관관계 분석), Finite element analysis(유한요소해석), Forward propagation(순전파법), Multi-layer neural network(다층 신경망)
초록 One of the chassis products, namely, the integrated axle, must meet the camber strength criteria in the development stage. Camber strength performance verification is mainly conducted by testing and finite element analysis. Recently, the test was replaced by the finite element analysis to verify integrated axle camber strength performance and reduce product development cost and time. However, in order to replace the test ith a simulation, the test/analysis correlation must be done beforehand. Therefore, this study proposes the integrated axle camber strength test/analysis correlation method through a multi-layer neural network. The result of the test/analysis correlation using a multi-layer neural network showed that the average error between the test and the analysis is about 10 %. This result proves the effectiveness of this research method, and suggests that the camber strength test could be replaced by the finite element analysis.
원문(PDF) 다운로드 Journal Site

사단법인 한국자동차공학회

  • TEL : (02) 564-3971 (사무국 업무시간 : 평일 오전 8시~)
  • FAX : (02) 564-3973
  • E-mail :

Copyright © by The Korean Society of Automotive Engineers. All rights reserved.