Soil moisture retrieval using fuzzy logic based on UWB signals

Abstract

A fuzzy logic approach to retrieve the soil moisture from the ultra-wideband (UWB) radar measurements is investigated in this paper. A monostatic UWB radar module is applied to collect the reflected signals from subsurface of bare soil and sand with different volume water contents (VWCs). Meanwhile, VWCs' data are collected by a time domain reflectometer (TDR). Traditionally, the channel modeling, from the statistical point of view, is able to retrieve the soil moisture. However, the statistical information lacks the ability to disclose the direct relation between the waveform and soil moisture. In this work, we try to model the problem as a time-series forecasting problem. The reflected signals are divided into training data and testing data. The training data are imported into the type-1 fuzzy logic system (T1FLS) to forecast the testing data. After that, we extract the parameters of membership functions in the final iteration and obtain the VWC value based on a recognition strategy.