首页|基于IAOA-SVM模型结构时变可靠性研究

基于IAOA-SVM模型结构时变可靠性研究

Time-varying Reliability Based on IAOA-SVM

扫码查看
原文链接
  • NETL
  • NSTL
  • 万方数据
目的 为有效解决使用传统代理模型进行结构时变可靠性研究中存在流程复杂、计算效率低等问题.方法 提出以改进算术优化算法(Improved Arithmetic Optimization Algorithm,IAOA)优化支持向量机模型(Support Vector Machine,SVM)进行时变可靠性研究的方法,结合IAOA-SVM模型和极值理论,以某塔式起重机回转支承为研究对象,对其进行动态确定性分析获取样本数据,建立IAOA-SVM可靠性模型,采用蒙特卡洛法求解得到其可靠度结果,并与EKM和ERSM算法对比分析其仿真精度和效率.结果 当回转支承径向变形许用值为0.278× 10-3m时,采用蒙特卡洛法求解得到其可靠度为99.68％,IAOA-SVM模型相比EKM和ERSM方法仿真效率有所提升,建模精度分别提高了 10.42％和9.23％.结论 IAOA-SVM方法在建模和仿真精度与效率方面具有较明显的优势,IAOA-SVM方法为求解机构时变可靠度难题提供了一种新的解决思路.
Objective To effectively solve the problems of complex process and low computational efficiency in the time-varying reliability study of structure using traditional agent models.Methods The Improved Arithmetic Opti-mization Algorithm(IAOA)s proposed to optimise the Support Vector Machine(SVM)model for time-varying reliability studyith the IAOA-SVM model and the extreme value theory and a tower crane slewing bearing as the research object,dynamic deterministic analysis s carried out to obtain sample data establish the IAOA-SVM relia-bility modelhe reliability results,and compare and analyse with the EKM and ERSM algorithms.Results When the permissible value of radial deformation of the slewing bearings 0.278x10-3m,the reliability of the slewing bear-ings 99.68％,and the simulation efficiency of the IAOA-SVM model s improved compared with that of the EKM and ERSM methods,the modelling accuracy improved by 10.42％and 9.23％respectively.Conclusion The IAOA-SVM method has obvious advantages in modelling and simulation accuracy and efficiency,and the IAOA-SVM method provides a new solution to the problem of solving the time-varying reliability of the.

time-varying reliabilitysupport vector machinesarithmetic optimization algorithmsslewing rings

郑建校、张小康、王亮亮、张锦华

展开 >

西安建筑科技大学机电工程学院,陕西 西安 710055

陕西省科学技术厅陕西省纳米材料与技术重点实验室,陕西西安 710055

时变可靠性 支持向量机 算术优化算法 回转支承

陕西省自然科学基础研究计划基金资助项目陕西省自然科学基础研究计划基金资助项目陕西省重点研发计划基金资助项目陕西省秦创原"科学家+工程师"队伍建设基金资助项目

2023-JC-YB-3132023-JC-YB-2942024GX-YBXM-1782024QCY-KXJ-140

2024

10.3969/j.issn.1672-1098.2024.03.002

安徽理工大学学报(自然科学版)
安徽理工大学

安徽理工大学学报(自然科学版)

影响因子：0.331
ISSN：1672-1098
年,卷(期)：2024.44(3)
  • 6