论文部分内容阅读
为实现钢-混组合结构PBL剪力键极限承载力的准确预测,通过分析既有推出试验结果,对PBL剪力键的作用机理及破坏模式进行总结,确定PBL剪力键纵向抗剪承载力的主要影响因素为开孔直径、钢板厚度、混凝土抗压强度、贯穿钢筋直径、钢板屈服强度等。以神经网络理论为基础,选用误差反向传播(BP)神经网络算法模型,选取钢板厚度、开孔直径、贯穿钢筋直径和混凝土抗压强度为输入因子,用C语言编写出基于BP神经网络的钢-混凝土组合结构PBL剪力键极限承载力预测系统。利用国内外既有试验资料对PBL剪力键极限承载力神经网络预测系统进行网络训练及预测验证。结果表明:利用BP神经网络可以建立主要影响因素与PBL剪力键极限承载力之间的非线性映射关系,实现PBL剪力键极限承载力预测。
In order to accurately predict the ultimate bearing capacity of PBL shear bond, the mechanism and failure mode of PBL shear bond were analyzed by analyzing the results of existing tests. The longitudinal shear capacity of PBL shear bond The main influencing factors are the diameter of open hole, the thickness of steel plate, the compressive strength of concrete, the diameter of steel bar, the yield strength of steel plate and so on. Based on the neural network theory, BP neural network algorithm model of error back propagation is selected, and the thickness of steel plate, the diameter of open hole, the diameter of reinforced concrete and the compressive strength of concrete are used as input factors, and the BP neural network Steel - concrete composite structure PBL shear bond ultimate capacity prediction system. Based on the existing experimental data at home and abroad, the neural network prediction system of PBL shear bond ultimate bearing capacity was trained and forecasted. The results show that the BPNN can establish the nonlinear mapping relationship between the main influencing factors and the ultimate bearing capacity of PBL, and realize the prediction of the ultimate bearing capacity of PBL shear bond.