Model for estimating soil moisture content based on ultrasound velocity under dry-wet alternating conditions

Soil moisture content is an important parameter for the management of farmland irrigation and research of regional hydrological conditions. It is an important prerequisite for scientific irrigation management to acquire the information of soil moisture and understand the dynamic change of soil moisture timely and accurately. However, the existing soil moisture content detection methods have some disadvantages, such as complex analysis process, long cycle, high cost and difficult to be widely used in practical applications. As a good information carrier, ultrasonic wave carries information related to soil moisture content in the receiving wave when it propagates in soil medium and interacts with soil. So, Ultrasonic testing technology is a way, which can monitor soil moisture quickly and in timely. In this study, taking the soil of farmland tillage layers as research specimens, the velocity of ultrasonic pulse through the soil sample from wet to dry was measured by ZBL-U510 non-metallic ultrasonic detector. The relationship between the velocity of ultrasonic wave and the soil moisture content was studied under different soil moisture conditions. The initial moisture contents of soil samples were 15%, 20%, 25% and 30%, respectively, and then dried to 5% in the natural environment. Each treatment sample experienced 4 dry-wet alternations. The results showed that the soil bulk density and ultrasonic velocity increased nonlinearly with the decrease of soil moisture content in the process of soil sample from wet to dry and the change relationship was similar in each wet-dry process. The change of soil moisture content had the greatest influence on the ultrasonic velocity in the first wet-dry process. In order to consider the effect of the dry-wet alternation on the ultrasonic velocity, an adaptive weighted data fusion algorithm was adopted. Firstly, the ultrasonic velocity values corresponding to soil sampling points for moisture content measurement in the processes of dry-wet alternation were fused in cross batch, and four ultrasonic velocity fusion values and mean square deviation were obtained. Then, the weighted factors of each group were calculated according to the multiple extreme value theory. Finally, the final fusion value of ultrasonic velocity was obtained according to the adaptive weighting model and the introduced weighting factor. Through the analysis of the relationship curve between the fusion value of ultrasonic velocity and the soil moisture content, the ultrasonic velocity increased nonlinearly with the continuous decrease of the soil moisture content in four different types of dry and wet alternation process, and the ultrasonic velocity increased rapidly with the decrease of the soil moisture content in the stage of high moisture content. The increasing rate of ultrasonic velocity was slowed down when the moisture content of soil samples ranged from 7%, 9%, 11% to 12%, which was basically consistent with the corresponding moisture content when the ultrasonic pulse velocity slowed down in each wet-dry alternate process. According to the relationship between the fusion values of ultrasonic velocity and moisture content, a polynomial model suitable for different initial moisture content was established. The relative error between the predicted value and the measured value of the model was less than 6%, which showed that the composite polynomial model can be used to accurately describe the relationship between soil moisture content and ultrasonic velocity. The results of this study can provide valuable information for the continuous monitoring and forecasting of soil moisture information in farmland by using the characteristics of ultrasonic velocity. © 2020, Editorial Department of the Transactions of the Chinese Society of Agricultural Engineering. All right reserved.