中文摘要：

对150组粉煤灰双掺混凝土试块进行试验室腐蚀试验的同时，还进行了25组宁夏地区普通混凝土试块的常规抗压试验，获得了大量对比试验数据，部分该数据还用作BP网络训练样本。通过试验数据统计和分析得到了超声一回弹综合法检测宁夏地区粉煤灰双掺耐腐混凝土的强度测强曲线，在此基础上，采用BP神经网络分别对宁夏地区考虑河水腐蚀和碳化影响条件下的粉煤灰双掺耐腐混凝土进行了智能预测分析。试验及预测结果可为高卤地区（宁夏等）粉煤灰混凝土的无损质量检测提供参考依据。

英文摘要：

The normal compression test for 25 suites of routine concrete test blocks was conducted while 150 sets of corrosion resistant test for fly ash concrete in laboratory has been finished.Lots of check experimental data based on the work were obtained.Part of the data was used as the sample of BP neural networks.Strength curve of corrosion resistant fly ash concrete in Ningxia checked by ultrasound-rebound method was achieved according to the experimental data.Then,the intellectual technology of BP neural network system was applied to predict the intensity of corrosion resistant fly ash concrete in order to improve the mechanical performance of the normal concrete corroded and carbonized usually by the water of river and soils in Ningxia.Nondestructive examination of corrosion resistant fly ash concrete can be verified by the conclusions from the work mentioned above.