Method for Soil Resistivity Measurement Based on Neural Network

During the cathodic protection of underground metal constructions, the soil resistivity measurement is mostly important to determine the position of sacrificial anodes. However, the measurement data are inevitably influenced by temperature variations and power fluctuations. In order to reduce these adverse impacts and enhance the accuracy of measurement, this paper introduces a novel application of neural network algorithm and data fusion, which adopts multi-dimensional calibration experimental data as its study samples. The experiment result shows that under the same condition of working temperature variation and power fluctuation, the measurement accuracy has been significantly enhanced after data fusion process.