混凝土钢筋锈蚀量预测的智能模型及应用

Intelligence model for predict corrosive degree of concrete embeded rebar

由于混凝土开裂后钢筋的锈蚀伴随着很多不确定性,实际工程中钢筋锈蚀程度的离散性又很大,另一方面,由于混凝土中的钢筋锈蚀机理非常复杂,影响影响因素很多,因此,采用各种计算模型很难得到钢筋锈蚀量,且与实测值误差较大。将径向基经网络技术应用于锈蚀开裂后混凝土中钢筋锈蚀量的评估,建立了评估钢筋锈蚀量的智能信息模型。最后通过实际工程检测结果验证了该方法的实际可行性,为钢筋锈蚀量的预测提供了一条新的途径。

The corrosion of reinforced concrete cracking was influenced by the many uncertainties, and the mechanism of steel corrosion in concrete is very complex.Radial basis function neural network is used for assessing corrosive degree of concrete embeded rebar after corrosion crack was produced After the main factors affecting the corrosive degree of rebar is analyzed, a artificial intelligence model was established.The adapti0n of the model is discussed from the angles of networks structural optimization and learning parameters.The feasibility is verified according to the data from practical engineering investigation.It provides a new way to predict corrosive degree of concrete embeded rehar after corrosion crack.