| According to current using situation of water resource at home and abroad and the utilization rate of irrigation on current agricultural production in China, this article analyses the study implications and practical value on measurement of soil moisture. And it also introduces the development trends of measurement of soil moisture at home and abroad, the theoretical models about soil moisture content, mainly expands the types and basic principle of existing sensor, the advantage and disadvantage of each sensor and comes up with new equipment which based on impedance for measurement of soil moisture. Based on existing research methods, this article discusses the idea and technology of this design, and lists the potential problems and resolve methods.This article analyses the mathematical model, the electro circuit precision, the feature and size of the sensor, expounds the design idea of the soil moisture measurement sensor and types chosen method, lists the advantages of soil moisture measurement sensor (Type E) in the testing time, the examination precision, the working conditions, the feature and size, the weight, the cost price and so on. Based on theory model, this article also analyses the effects of soil structure on reactance of the sensor, explains the switching from soil dielectric measurement constant to the electrical quantities that computers can recognize which provides the basis for incremental design. Based on the relationship between dielectric coefficient and soil volume moisture content proposed by Top et al, we proposed the relational expression for this. In this research, we determined the selective conditions of testing objects which are used for testing the soil moisture measurement sensor and adopted the representational Black soil, Yellow soil as testing soil. We made the fitting test of the sensor under the different experimental object condition and the different experiment frequency condition by MATLAB7.0, according to the fitting analysis of the two specific sensors, we proposed the linear equation of the dielectric constant and the soil water content for each sensor separately, and calculates the error by using the concept of absolute error and relative error to the general characteristic fitting equation. Our conclusion is that model error is less than±1.9% which meets the demands of design. |