BP神经网络在铜及铜合金海水腐蚀预测中的应用

BP Neural network approach in the prediction of copper and copper alloy corrosion in seawater

根据实海环境数据及材料腐蚀数据,利用BP结构神经网络建立了铜及铜合金在实海环境中腐蚀速度与环境因素、材料成分之间神经网络预测模型。利用建立的预测模型分析了环境因素对铜及铜合金的腐蚀速度的影响。分析结果表明,温度的升高及生物污损促进铜及铜合金的腐蚀,而pH、盐度和氧浓度的升高对浸泡1年的材料腐蚀速度有明显的抑制作用。

Based on the environment data and material corrosion data, using BP artificial neural network method, models for prediction of corrosion rates of copper and copper alloy were proposed to describe the relationship between corrosion rate, environment factors, and component of material. Use the model to analyse the affect of environment factors on corrosion rates of copper and copper alloy. The result of analyses is that higher temperature and bio-fouling accelerate copper and copper alloy corrosion, but the pH, oxygen solubility and salinity decelerate corrosion clearly only in one-year test.