机载非接触式近红外土壤墒情检测系统研制.

A non-contact and online detection system was developed for the soil moisture using Fabry Perot interference near-infrared chip. A large number of points were reduced for the cost-saving and non-destruction of the cultivated layer after the probe of the soil moisture sensor was inserted into the soil. The hardware of the system consisted of an airborne automatic detection device, an electrical control box, and Beidou dual antenna real-time differential positioning. The specific sensors were then selected to detect the soil water content. The packaging of the module was designed to protect the internal structure of the sensor from damage. A lifting detection device was used to control the soil moisture sensor onto the soil surface for the detection. The obstacle avoidance of the device with the ultrasonic sensor was installed to realize the automatic positioning of the measured height. The spectral data in the range of 1 750-2 150 mm was collected from the soil samples with different water content and types. A partial least squares (PLS) prediction model was established for the water content of the soil surface. The determination coefficients of all prediction models were above 0.9. There was a high curve fitting degree of the overall soil model, where the R2 reached 0.933 4. Although there was slightly lower than that of the single soil prediction model, there was a high universality of the overall model suitable for the prediction of most soil water content. Furthermore, the secondary development of the sensor data acquisition was realized to embed the prediction model of soil water content into the original near-infrared sensing chip system. As such, the measured value of soil water content and the corresponding spectral data curve were more intuitively displayed on the industrial computer in real time, when measuring soil samples. The measured distance information by the ultrasonic sensor was transformed into the voltage signals using the cooperated Xinjie PLC and analog module. The closed-loop stepping motor was utilized to adjust the height of the soil moisture sensor in real time, according to the PLC feedback signals, thereby realizing the coordination between the positioning system and the detection device. Correspondingly, a trial prototype was fabricated to integrate the near-infrared detection system of soil moisture and the automatic self-propelled transplanter. The field test results show that the soil moisture sensor after online calibration was dropped onto the soil surface with the lifting detection mechanism for measurement when the transplanter moved at the inspection speed of 0.3 m/s. After that, the moisture content value rose within 5 s, after the real-time display of the moisture content on the industrial computer. The measured water content of surface soil was combined with the positioning of the Beidou RTK system, thereby calculating the soil moisture content under the accurate longitude and latitude. The distribution map of soil moisture content was generated on the measured plot. Subsequently, the sampling points after the test were sampled, pretreated, dried, and calculated to obtain the actual moisture content of the sampling points, where the relative error was calculated between the measured and actual moisture content. Consequently, the measured soil moisture content in the test was basically consistent with the actual one, where the relative error of continuous detection was less than 10%. The distribution map of soil surface water content can be expected to directly match the early warning level of soil moisture for the visualization of the abstract information. Therefore, the soil moisture information of the field can be detected to accurately display in real time. The finding can provide a strong reference and practical significance for the variable irrigation, such as the sprinkler and drip irrigation, thereby optimizing the regional water and soil adaptation in the spatiotemporal pattern of water productivity. Intelligent agricultural machinery can also be created to fully realize the “border inspection and side management” of field management [ABSTRACT FROM AUTHOR]