改进BP神经网络的混凝土构件承载力预测仿真  

Improved BP Neural Network for Predicting the Bearing Capacity of Concrete Components

作　　者：夏运生 白鑫 马三蕊 XIA Yun-sheng;BAI Xin;MA San-rui(Huanghe Jiaotong University,Jiaozuo Henan 454950,China;Ningbo University,Ningbo Zhejiang 563000,China)

机构地区：[1]黄河交通学院,河南焦作454950 [2]宁波大学,浙江宁波563000

出　　处：《计算机仿真》2024年第4期436-440,共5页Computer Simulation

摘　　要：受到多种因素的影响,大直径混凝土受弯构件在使用期间其承载力将发生变化,为此提出大直径混凝土受弯构件承载力预测方法。确定影响大直径混凝土受弯构件承载力的五大因素,将影响因素作为输入建立用于大直径混凝土受弯构件承载力预测的BP神经网络模型,通过模拟退火-粒子群混合算法优化BP神经网络模型参数,并使用优化后BP神经网络模型完成大直径混凝土受弯构件承载力预测。实验结果表明,所提方法的大直径混凝土受弯构件承载力预测精度和效率更高,整体应用效果更好。Due to various factors,the bearing capacity of large-diameter concrete flexural members will change during use.To address this issue,this article put forward a method for predicting the bearing capacity of large-diameter concrete flexural member.Firstly,five major factors affecting the bearing capacity of large-diameter concrete flexural member were identified.Then,these factors were served as inputs to construct a BP neural network model for predicting the bearing capacity of large-diameter concrete flexural members.Finally,the parameters of the BP neural network model were optimized by the simulated annealing-particle swarm algorithm,and then the optimized model was used to complete the prediction of the bearing capacity of large-diameter concrete flexural members.The experimental results show that the proposed method has higher prediction accuracy and efficiency for the bearing capacity of largediameter concrete flexural member and better overall application effect.

关 键 词：大直径混凝土 承载力预测 受弯构件 神经网络模型 模拟退火-粒子群混合优化算法 

分 类 号：TP391[自动化与计算机技术—计算机应用技术] TP183[自动化与计算机技术—计算机科学与技术]