Characteristic Curve Fitting of Capacitive Rainfall Sensor Based on BP Neural Network

Rainfall is a kind of common weather phenomenon. Accurate measurement of rainfall is of great significance for weather and natural disasters forecasting. In this paper, the capacitive sensor was applied to the measurement of rainfall and the working principle was introduced. Furthermore, the output characteristic curve of the sensor was fitted by BP neural network. The sensor's calibration data were taken as training samples and BP network model was established. The results showed that the fitting algorithm based on BP neural network had faster convergence speed and higher accuracy, and its fitting error was much smaller than that of the least square method.