采用两种方法对日仪管道沿线土壤腐蚀性能进行了评价。一是利用实测极化曲线求得腐蚀电流密度，再根据腐蚀电流密度与实际失重率之间的对应关系得到试片的实际失重率，对土壤腐蚀性进行等级评价; 二是根据土壤的理化性质，构建BP神经网络，选择一定容量的样本进行训练后，根据训练好的网络进行腐蚀性评价。结果表明，两种方法评价的结果一致。通过对比两种评价过程发现: 当测试点较多时，两种评价方法均可使用; 但是，当测试点较少时，方法二误差很大，应采用方法一。

Two methods were used to evaluate the soil corrosivity along the Riyi pipeline. One was measuring polarization curves to get corrosion current density, then according to the relationship between corrosion current density and actual weight loss rate to get the actual weight loss rate of the specimen, and to evaluate the soil corrosivity. Another method was constructing the BP neural network based on soil physical and chemical properties. Selecting a certain capacity of the training samples, and then according to the trained network, the corrosirity evaluation was conducted. The results show that two methods have good coherence. Comparison of the two methods, indicated that when the number of the test points was large, two evaluation methods all could be used. However, when the number is less, the second method show large errors, the first method should be used in that case.