用BP模型预测地铁杂散电流腐蚀危险性

根据埋地金属结构受杂散电流影响极化电位正向偏移的数据和环境中氯离子的浓度，采用人工神经网络的方法，建立了地铁杂散电流腐蚀危险性等级的预测模型，并选取了一组样本进行了预测模型的训练，利用此模型对广州地铁一号线整体道床排流网和侧壁结构杂散电流腐蚀情况进行预测，结果表明，该模型能够预测地铁杂散电流腐蚀的危险性等级.

PREDICTION OF DANGER DEGREE OF CORROSION INDUCED BY STRAY CURRENT FOR METRO BY USING BP ARTIFICAL NEURAL NETWORK

On the basis of chlorine ion concentration in the environment and the polarity voltage positive offset data of the buried metal structure influenced by stray current,by adopting artificial neural networks' method,a prediction model of dangerous degree of stray current corrosion for metro was proposed.Carrying out training with one group of selected swatches,the prediction model was established,and then by using the model,we predicted the stray current corrosion of the whole roadbed drainage net and side-wall structures of Guangzhou metro line No.1.The result showed that the model can predict the dangerous degree of stray current corrosion for metro.