基于人工神经网络的预腐蚀铝合金疲劳性能预测

 通过对BP神经网络算法分析和收敛性改进,从获得的预腐蚀和疲劳试验数据中通过训练建立了LY1 2CZ铝合金腐蚀性能和疲劳特性与预腐蚀温度和时间的映射模型,从而可预测铝合金在一定预腐蚀环境谱下的最大腐蚀深度和疲劳特性。神经网络算法采用 BP算法 ,网络结构采用2-4-2形式。结果表明 ,神经网络用于预腐蚀铝合金的腐蚀状况和疲劳性能预测是可行的

PREDICTIVE MODEL BASED ON ARTIFICIAL NEURAL NET FOR FATIGUE PERFORMANCES OF PRIOR-CORRODED ALUMINUM ALLOYS

A prediction model for corrosion and fatigue performances of the prior corroded aluminum alloys under a varied corrosion environmental spectrum based on artificial neural net was developed and the non linear relationship between maximum corrosion depth,fatigue performance and corrosion temperature,time was established based on BP learning algorithm analysis and convergence improvement.The maximum corrosion depth and fatigue performances of prior corroded aluminum alloys can be predicted by means of the trained neural net from the testing data. The learning algorithm for neural net is BP(back\|propagation) algorithm with 2 4 2 structure.The results show that,for multi\|factor corrosion prediction,the prediction model based on BP learning algorithm for corrosion and fatigue performances of the prior corroded aluminum alloys is feasible and effective.Thus,by virtue of the prediction model,the future corrosion status and fatigue performances of aluminum alloys can be evaluated under random complicated environmental spectrum.