压弯钢筋混凝土柱正截面极限承载力的预测——基于BP神经网络技术  

Forecasting of the Terminal Bearing Capacity in Sections of Reinforced Concrete Column Under Bending and Compression: On the Basis of BP Neural Network

作　　者：焦俊婷[1] 叶英华[1] 刁波[1] 于霖冲[2] 

机构地区：[1]北京航空航天大学土木工程系,北京100083 [2]嘉应学院土木工程系,广东梅州514015

出　　处：《华南理工大学学报（自然科学版）》2005年第8期80-82,94,共4页Journal of South China University of Technology(Natural Science Edition)

基　　金：国家自然科学基金资助项目(50178008)

摘　　要：提出双向压弯钢筋混凝土柱正截面极限承载力的预测模型.以影响钢筋混凝土柱正截面极限承载力的主要因素(如:截面尺寸、混凝土强度、加载角度及配筋率等)为参数,用数值模拟结果为训练样本,建立了柱正截面极限承载力的BP神经网络预测模型.经验证,该模型对双向压弯钢筋混凝土柱正截面极限承载力具有良好的预测效果.A model is presented to forecast the terminal bearing capacity in the sections of reinforced （RC） columns under bi-axial bending and compression. By taking the main factors affecting the bearing concrete capacity, such as the section dimension, the concrete strength, the loading angle and the reinforce ratio, as the model parameters, and by using the numerical simulation results as the training specimens, a forecasting model is established based on BP neural network. It is verified that the proposed model is of excellent forecasting ability for the terminal bearing capacity of RC columns under bi-axial bending and compression.

关 键 词：钢筋混凝土 双向压弯 承载力 预测 神经网络 

分 类 号：TU375[建筑科学—结构工程] TP183[自动化与计算机技术—控制理论与控制工程]