TY - JOUR
T1 - Support vector machine-based similarity selection method for structural transient reliability analysis
AU - Chen, Jun Yu
AU - Feng, Yun Wen
AU - Teng, Da
AU - Lu, Cheng
AU - Fei, Cheng Wei
N1 - Publisher Copyright:
© 2022 Elsevier Ltd
PY - 2022/7
Y1 - 2022/7
N2 - The transient reliability analysis of structures enduring complex loads from multiple sources originating from extraordinarily tangled and complicated operation situation, plays a leading role in the operational safety and design cost of system. In this work, a support vector machine-based similarity selection genetic algorithm (SVM-SSGA) for structural transient reliability analysis is developed by integrating support vector machine (SVM), similarity selection strategy and genetic algorithm (GA). The transient reliability analysis of nose landing gear (NLG) shock strut outer fitting stress is performed to verify the modeling and simulation performance of SVM-SSGA. The results show that (i) the developed SVM-SSGA method holds eminent modeling features, resulting from that average absolute error is 0.7493 × 106 Pa and modeling time is 0.1847s, and (ii) the SVM-SSGA method is superior to other methods in simulation characteristics, since simulation time is 0.2347s of 104 MC samples and precision reaches 99.99% compared to direct simulation, (iii) the reliability degree of the NLG shock strut outer fitting stress is 0.9972 when the allowable stress is σ=1.5020 × 109 Pa. The efforts of this study provide a promising method in transient structural reliability analysis, which is prospective to improve the operational safety and reliability of the system besides the NLG.
AB - The transient reliability analysis of structures enduring complex loads from multiple sources originating from extraordinarily tangled and complicated operation situation, plays a leading role in the operational safety and design cost of system. In this work, a support vector machine-based similarity selection genetic algorithm (SVM-SSGA) for structural transient reliability analysis is developed by integrating support vector machine (SVM), similarity selection strategy and genetic algorithm (GA). The transient reliability analysis of nose landing gear (NLG) shock strut outer fitting stress is performed to verify the modeling and simulation performance of SVM-SSGA. The results show that (i) the developed SVM-SSGA method holds eminent modeling features, resulting from that average absolute error is 0.7493 × 106 Pa and modeling time is 0.1847s, and (ii) the SVM-SSGA method is superior to other methods in simulation characteristics, since simulation time is 0.2347s of 104 MC samples and precision reaches 99.99% compared to direct simulation, (iii) the reliability degree of the NLG shock strut outer fitting stress is 0.9972 when the allowable stress is σ=1.5020 × 109 Pa. The efforts of this study provide a promising method in transient structural reliability analysis, which is prospective to improve the operational safety and reliability of the system besides the NLG.
KW - Genetic algorithm
KW - Landing gear
KW - Similarity selection
KW - Support vector machine
KW - Transient reliability
UR - http://www.scopus.com/inward/record.url?scp=85128313679&partnerID=8YFLogxK
U2 - 10.1016/j.ress.2022.108513
DO - 10.1016/j.ress.2022.108513
M3 - 文章
AN - SCOPUS:85128313679
SN - 0951-8320
VL - 223
JO - Reliability Engineering and System Safety
JF - Reliability Engineering and System Safety
M1 - 108513
ER -