Research On Parameter Calibration And Intellectualization Of Soil Sensor

Soil is an important environment for plants to survive,and it plays an important role in the entire process of plant growth.Among them,the temperature and humidity parameters in the soil have a greater impact on the growth environment of plants.The accurate detection of soil temperature and humidity parameters is an important guarantee for scientific farmland management.Therefore,the detection of soil temperature and humidity parameters is very necessary.The soil sensor is an important tool for the detection of soil temperature and humidity parameters.It can quickly detect changes in temperature and humidity parameters.However,when it works in harsh environments for a long time,the detected temperature and humidity parameters will produce errors,causing inaccurate detection.Based on the above problems,in order to achieve online calibration of soil sensor parameters,reduce measurement errors,and improve measurement accuracy,the following work has been done:(1)Analyzing the static characteristics of the soil sensor,the results show that the humidity parameter error is large,and the temperature parameter error is small.Therefore,this thesis focuses on the humidity parameter error and determines the two-level calibration scheme.The first-level calibration solution solves the problem of soil sensor’s own probe corrosion and aging of internal components,which cause errors in humidity parameters.The secondary calibration solution solves the problem that the dielectric constant inside the soil will change under different soil temperatures,resulting in errors in the humidity parameters of the soil sensor.(2)In order to solve the problem of humidity error,the realization process of the two-level calibration scheme is as follows: The first level calibration is to add a calibration circuit inside the sensor,and use cubic spline interpolation and polynomial curve fitting methods to correct the corrosion of the soil sensor probe.And the error caused by the degree of aging.The second-level calibration scheme is to establish a temperature compensation model for the soil sensor parameters.In order to improve the compensation effect of the humidity parameters,the Particle Swarm Optimization(PSO)algorithm is used to calculate the parameters c and SVR of the Support Vector Regression(SVR).The optimization of g is carried out,and the result is c=10.509,g=0.353,the mean square error of the temperature compensation model is 0.0109,and the coefficient of determination can reach 0.9833.(3)Using the architecture model of the Internet of Things,combined with software and hardware technology,an online calibration system for soil sensor parameters is designed.Use STC89C52 and STM32 as the microprocessor(Micro Control Unit,MCU)of the system node,build a Lo Ra(Long Range)one-master multi-slave communication method,and use the Narrow Band-internet of things(NB-iot)to connect The original parameters of the soil sensor are transmitted to the calibration terminal,and the calibration terminal software developed by QT is used to realize the two-level calibration algorithm,so that the online calibration of the soil sensor parameters can be intelligent.(4)Test the key nodes of the soil sensor parameter calibration system,and the results show that the wireless network transmission data is stable,and the first-level calibration algorithm and the second-level calibration algorithm are implemented correctly.Finally,through comparative experiments,the average absolute error and relative error between the two-level calibration data and the standard value are 0.007 and 1.107%,and the calibration effect is good.