摘要
为准确预测连续管的多轴疲劳寿命,分析其疲劳寿命的影响因素,建立基于GRA-PSO-BP的连续管疲劳寿命预测模型。采用灰度关联法(GRA)分析连续管疲劳寿命的影响因素,建立经过粒子群算法(PSO)优化后的BP神经网络预测模型即PSO-BP模型,分析连续管直径、壁厚、弯曲半径以及内压与连续管疲劳寿命的关系并进行预测。研究结果表明:基于GRA分析结果可知,弯曲半径对连续管的疲劳寿命影响最大;通过与5种机器学习模型进行对比可知,GRA-PSO-BP模型具有较高的预测精度和较强的泛化性能。
In order to accurately predict the multi-axial fatigue life of coiled tubing and analyze the influencing factors of its fatigue life,a fatigue life prediction model of coiled tubing based on GRA-PSO-BP was established.Firstly,the gray correlation method(GRA)was used to analyze the factors affecting the fatigue life of CT.Then,a PSO-BP prediction model,namely the BP neural network optimized by particle swarm optimization(PSO),was established,and the relationship between the diameter,wall thickness,bending radius,internal pressure of CT and the fatigue life of CT was analyzed to carry out the related prediction.The results show that the bending radius has the greatest influence on the fatigue of CT by GRA analysis.The established GRA-PSO-BP model has high prediction accuracy and strong generalization performance,and the prediction results of this model are reliable and accurate compared with multiple models.
作者
吴远灯
刘少胡
马卫国
WU Yuandeng;LIU Shaohu;MA Weiguo(School of Mechanical Engineering,Yangtze University,Jingzhou Hubei 434023,China)
出处
《中国安全生产科学技术》
CAS
CSCD
北大核心
2023年第6期135-142,共8页
Journal of Safety Science and Technology
基金
国家自然科学基金项目(51974036,51604039)
湖北省高等学校优秀中青年科技创新团队计划项目(T2021035)。
