A data processing method for a field mobile data collection terminal

Abstract

The present invention relates to the field of agricultural information technology, and in particular to a data processing method for a field mobile data collection terminal. The present invention not only has fixed and accurate meteorological and soil data, but also has a mobile data collection terminal for collecting mobile meteorological data of a large area. By constructing a time and space asynchronous data processing method and performing self-verification, data acquisition of the entire region is achieved, and a relationship between meteorological data and soil moisture data changes is constructed through a machine learning model. The change in soil moisture data at an observation point can be obtained through the meteorological data collected by mobile movement. The provided method can reduce the number of fixed observation meteorological stations in a region and can implement data inspections over a large area.

Description

Technical Field

[0001] The present invention relates to the field of agricultural information technology, and particularly relates to a data processing method for a field mobile data acquisition terminal. Background Art

[0002] For agricultural data acquisition systems or platforms, they mainly rely on fixed observation points. Through fixed weather stations and soil moisture sensors, farmland information is collected. When the farmland area is large, the number of observation stations needs to be increased. Especially for soil moisture sensors, not only does the cost increase, but after the number increases, it will significantly affect farmland mechanization operations. And mobile data acquisition is not suitable for a large acquisition range. During the movement, the data acquisition time and location are different, resulting in a lack of information on different plots. Moreover, automated mobile acquisition equipment can only travel on the road and cannot enter the field to collect soil information.

[0003] In summary, developing a data processing method for a field mobile data acquisition terminal is still a key problem urgently to be solved in the field of agricultural information technology. Summary of the Invention

[0004] In view of the above-mentioned drawbacks of the prior art, a data processing method for a field mobile data acquisition terminal is provided. The provided method can reduce the number of regional fixed observation weather stations and can implement large-area data inspection.

[0005] To achieve the above object, the present invention provides the following technical solutions:

[0006] A data processing method for a field mobile data acquisition terminal, comprising the following steps:

[0007] (1) Two sets of data acquisition systems: Using a mobile data acquisition platform and a fixed-point data station data acquisition system for data acquisition;

[0008] (2) Preparation of mobile data: First, determine the data acquisition plot, where the roads in the plot meet the normal driving requirements of autonomous driving. Then, determine the observation point positions and GPS data. The position of a fixed observation point coincides with the position of one of the mobile observation points. Then, according to the position layout form route, all observation points are marked in a grid in the plane coordinate system;

[0009] (3) Time asynchronous data processing: Simulating the actual data of the whole day through the data at different time points of a single observation point, where:

[0010] The time asynchronous data processing includes time asynchronous data processing of temperature and relative humidity, time asynchronous data processing of solar radiation intensity data, change amount of soil moisture data and meteorological data processing, and spatial asynchronous data processing.

[0011] The present invention is further configured as follows: in step (1), the mobile data acquisition platform adopts a Beidou navigation automatic driving system and a GPS positioning system to automatically drive on farmland roads, and acquires collected data through installed meteorological data sensors, and the collected data includes air temperature, relative air humidity and solar radiation intensity.

[0012] The present invention is further configured as follows: in step (1), the fixed point data station data acquisition system performs data acquisition through a fixed point data acquisition station, and the collected data includes air temperature, air relative humidity, solar radiation intensity, soil moisture content, soil EC, soil pH, and soil nitrogen, phosphorus and potassium nutrient concentrations.

[0013] The present invention is further configured as follows: in step (2), the route includes all observation points and forms a closed loop, and data is collected at least three times at the same observation point location every day.

[0014] The present invention is further configured as follows: in step (3), the time asynchronous data processing of temperature and relative humidity adopts a first model for data processing, and the calculation formula of the first model is: θ=α·sin(ωt+β)+δ, at different times t, the simulation data θ is calculated by a sine function, wherein α, ω, β, δ are model parameters, and the calculation process of the first model is as follows:

[0015] By collecting and observing data from all observation points, the time of occurrence of the maximum and minimum values ​​is found, which are the peak time and trough time of the sine function respectively. Then, the period of the sine function is calculated as the difference between twice the peak time and the trough time. Through at least 3 sets of data from each data observation point, all model parameters are determined using the minimum square root error method to complete the data model calibration for the day. Then, by bringing in the time point, the temperature and relative humidity data at all time points of each data observation point on the day are calculated.

[0016] The present invention is further configured as follows: in step (3), the time asynchronous data processing of the solar radiation intensity data adopts a second model for data processing, and the calculation formula of the second model is: At different times t, the simulated radiation intensity data Rad is calculated by the sine function. is the average of the daily cumulative radiation measured at the fixed observation point and the regional daily cumulative radiation measured at the mobile observation point, l is the length of the day, and γ is the model parameter.

[0017] The present invention is further configured as follows: in step (3), the variation of the soil moisture data and the meteorological data are processed using a third model, and the third model simulates the variation of the soil moisture data at the observation point Δ, Δ = f(θ Tem ,θ RH ,Rad), where θTem is the temperature, θ RH is the relative humidity and Rad is the simulated radiation intensity.

[0018] The present invention is further configured as follows: in step (3), the spatial asynchronous data processing adopts the fourth model to perform data processing as follows:

[0019] The data of mobile observation points are used to simulate the changes in meteorological data and soil moisture data of the measurement area surrounded by all mobile observation points; through the constructed plane coordinate system, the lower left corner is selected as the coordinate far point, and the actual position and distance of all points are determined by GPS information to determine the position coordinates; the single linear interpolation A, the nearest neighbor algorithm interpolation B and the cubic interpolation C are calculated respectively, and then different weight ratios are assigned to them respectively; the internal data are simulated and calculated through the edge observation values; the internal observation values ​​are used to verify the interpolation data and calibrate the weight ratio coefficient; the optimal weight coefficient is determined, and then the data of the internal unmeasured points are simulated through all observation points to realize the processing of spatial asynchronous data.

[0020] Beneficial Effects

[0021] Compared with the known public technology, the technical solution provided by the present invention has the following beneficial effects:

[0022] The present invention not only has fixed and accurate meteorological and soil data, but also has a mobile data acquisition terminal to collect mobile meteorological data of a large area. By constructing a time and space asynchronous data processing method and performing self-verification, data acquisition of the entire area is achieved, and the relationship between the meteorological data and the change in soil moisture data is constructed through a machine learning model. The change in soil moisture data at the observation point can be obtained through the meteorological data collected through mobile operation. The provided method can reduce the number of fixed observation meteorological stations in the region and can realize data inspection over a large area. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 A schematic diagram of the preparation process for mobile data collection according to the present invention. DETAILED DESCRIPTION

[0024] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0025] The present invention will be further described below in conjunction with the embodiments.

[0026] Example

[0027] The present invention provides a data processing method of a field mobile data collection terminal, comprising the following steps:

[0028] (1) Two data collection systems: Data collection is performed using a mobile data collection platform and a fixed-point data station data collection system.

[0029] Furthermore, the mobile data collection platform uses the Beidou navigation automatic driving system and the GPS positioning system to automatically drive on farm roads and obtain data through the installed meteorological data sensors. The collected data includes air temperature, relative humidity and solar radiation intensity.

[0030] Furthermore, the fixed point data station data acquisition system collects data through fixed point data acquisition stations, and the collected data include air temperature, air relative humidity, solar radiation intensity, soil moisture content, soil EC, soil pH and soil nitrogen, phosphorus and potassium nutrient concentrations.

[0031] In this step, the integrity of the data can be ensured by complementing and calibrating the data of the two systems.

[0032] (2) Preparation of mobile data: First, determine the data collection site, whose roads meet the normal driving requirements of autonomous driving. Then determine the observation point location and GPS data. The fixed observation point location coincides with the location of one of the mobile observation points. Then, based on the location layout form route, all observation points are grid-marked in the plane coordinate system.

[0033] Furthermore, the route includes all observation points and forms a closed loop, and data is collected at least three times at the same observation point location every day.

[0034] In order to better explain this step, Figure 1 As shown, Figure 1 (a) is the determined data collection site, which can meet the normal driving needs of autonomous driving, (b) is the observation point location and GPS data, (c) is the route formed according to the location layout, and (d) is the grid annotation effect of all observation points in the plane coordinate system.

[0035] (3) Time-asynchronous data processing: The actual data of the whole day is simulated by using the data at different time points of a single observation point, where:

[0036] Time asynchronous data processing includes time asynchronous data processing of temperature and relative humidity, time asynchronous data processing of solar radiation intensity data, change of soil moisture data and meteorological data processing, and space asynchronous data processing.

[0037] Furthermore, the time-asynchronous data processing of temperature and relative humidity adopts the first model for data processing. The calculation formula of the first model is: θ = α·sin(ωt+β)+δ. At different times t, the simulation data θ is calculated by the sine function. In the formula, α, ω, β, and δ are model parameters. The calculation process of the first model is as follows:

[0038] By collecting and observing data from all observation points, the time of occurrence of the maximum and minimum values ​​is found, which are the peak time and trough time of the sine function respectively. Then, the period of the sine function is calculated as the difference between twice the peak time and the trough time. Through at least 3 sets of data from each data observation point, all model parameters are determined using the minimum square root error method to complete the data model calibration for the day. Then, by bringing in the time point, the temperature and relative humidity data at all time points of each data observation point on the day are calculated.

[0039] Furthermore, the time asynchronous data processing of solar radiation intensity data adopts the second model for data processing, and the calculation formula of the second model is:

[0040] At different times t, the simulated radiation intensity data Rad is calculated by the sine function. is the average of the daily cumulative radiation measured at the fixed observation point and the regional daily cumulative radiation measured at the mobile observation point, l is the length of the day, and γ is the model parameter.

[0041] Furthermore, the variation of soil moisture data and meteorological data are processed by a third model, and the third model simulates the variation of soil moisture data at the observation point Δ, Δ = f(θ Tem ,θ RH ,Rad), where θ Tem is the temperature, θ RH is the relative humidity and Rad is the simulated radiation intensity.

[0042] It should be noted that, through the fitting model of the changes in soil moisture data (soil water content, soil EC, soil pH and soil nitrogen, phosphorus and potassium nutrient concentrations) and meteorological data (air temperature, air relative humidity and solar radiation intensity), the meteorological data (temperature, relative humidity and radiation intensity) at different observation points can be used to simulate the change Δ of the soil moisture data at the observation point.

[0043] The fitting model provided is Δ=f(θ Tem ,θ RH,Rad) is constructed by using the historical data of meteorological data and soil moisture data measured at fixed observation points using machine learning algorithms. The machine learning algorithms include but are not limited to naive Bayes classification algorithm, linear regression, logistic regression, K-means clustering algorithm, decision tree and Gaussian regression algorithm. By calculating the minimum error, determining the relationship model, and continuously accumulating data, the model accuracy can be continuously updated to achieve the change Δ of the soil moisture data at the observation point simulated by meteorological data.

[0044] Furthermore, the spatial asynchronous data processing adopts the fourth model to perform data processing as follows:

[0045] The data of mobile observation points are used to simulate the changes in meteorological data and soil moisture data of the measurement area surrounded by all mobile observation points; through the constructed plane coordinate system, the lower left corner is selected as the coordinate far point, and the actual position and distance of all points are determined by GPS information to determine the position coordinates; the single linear interpolation A, the nearest neighbor algorithm interpolation B and the cubic interpolation C are calculated respectively, and then different weight ratios are assigned to them respectively; the internal data are simulated and calculated through the edge observation values; the internal observation values ​​are used to verify the interpolation data and calibrate the weight ratio coefficient; the optimal weight coefficient is determined, and then the data of the internal unmeasured points are simulated through all observation points to realize the processing of spatial asynchronous data.

[0046] In this step, the plane coordinate system is constructed by preparing the moving data in step (2). The present invention combines three commonly used spatial interpolation methods, namely, unilinear interpolation, nearest neighbor algorithm and cubic interpolation, to calculate unilinear interpolation A, nearest neighbor algorithm interpolation B and cubic interpolation C respectively, and then assign different weight ratios, such as a, b and (1-ab), respectively.

[0047] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements will not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A data processing method for a field mobile data acquisition terminal, characterized in that, it includes the following steps: (1) Two sets of data acquisition systems: Use a mobile data acquisition platform and a fixed-point data station data acquisition system to collect data; (2) Preparation of mobile data: First, determine the data acquisition plot, where the roads in the plot meet the normal driving requirements of autonomous driving. Then, determine the observation point positions and GPS data, and make the fixed observation point position coincide with one of the mobile observation point positions. Then, according to the position layout form route, grid-label all the observation points in the plane coordinate system; (3) Time asynchronous data processing: Simulate the actual data of the whole day through the data at different time points of a single observation point, where: The time asynchronous data processing includes time asynchronous data processing of temperature and relative humidity, time asynchronous data processing of solar radiation intensity data, change amount of soil moisture data and meteorological data processing, and spatial asynchronous data processing; The time asynchronous data processing of the temperature and relative humidity adopts a first model for data processing. The calculation formula of the first model is: θ = α·sin(ωt + β) + δ. For different times t, the simulated data θ is calculated through the sine function. In the formula, α, ω, β, and δ are model parameters. The calculation process of the first model is as follows: Through the collection and observation of all observation point data, find the occurrence times of the maximum value and the minimum value, which are the peak time and the trough time of the sine function respectively. Then calculate the sine function period as the difference between 2 times the peak time and the trough time; Through at least 3 groups of data of each data observation point, use the method of least square root error to determine all model parameters and complete the calibration of the data model for the day; Then, through the substitution of time points, calculate the temperature and relative humidity data at all time points of each data observation point for the day; The time asynchronous data processing of the solar radiation intensity data adopts a second model for data processing. The calculation formula of the second model is: At different times t, the simulated radiation intensity data Rad is calculated through a sine function. is the average of the daily cumulative radiation measured at a fixed observation point and the regional daily cumulative radiation measured at a moving observation point. l is the day length of that day, and γ is a model parameter. The variation of soil moisture data and meteorological data are processed by a third model, which simulates the variation of soil moisture data at the observation point Δ, Δ = f(θ Tem ,θ RH ,Rad), where θ Tem is the temperature, θ RH is the relative humidity and Rad is the simulated radiation intensity.

2. The data processing method for a field mobile data acquisition terminal according to claim 1, characterized in that, in step (1), the mobile data acquisition platform adopts a Beidou navigation autonomous driving system and a GPS positioning system, drives autonomously on the farmland roads, and obtains the collected data through the installed meteorological data sensors. The collected data includes air temperature, air relative humidity, and solar radiation intensity.

3. The data processing method for a field mobile data acquisition terminal according to claim 1, characterized in that, in step (1), the fixed-point data station data acquisition system collects data through a fixed-point data acquisition station. The collected data includes air temperature, air relative humidity, solar radiation intensity, soil water content, soil EC, soil PH, and soil nitrogen, phosphorus, and potassium nutrient concentrations.

4. The data processing method for a field mobile data acquisition terminal according to claim 1, characterized in that, in step (2), the route includes all the observation points and forms a closed loop, and at least 3 times of data are collected at the same observation point position every day.

5. The data processing method of a field mobile data collection terminal according to claim 1, It is characterized in that In step (3), the spatial asynchronous data processing adopts the fourth model to perform data processing as follows: The data of mobile observation points are used to simulate the changes in meteorological data and soil moisture data of the measurement area surrounded by all mobile observation points; through the constructed plane coordinate system, the lower left corner is selected as the coordinate far point, and the actual position and distance of all points are determined by GPS information to determine the position coordinates; the single linear interpolation A, the nearest neighbor algorithm interpolation B and the cubic interpolation C are calculated respectively, and then different weight ratios are assigned to them respectively; the internal data is calculated by simulating the edge observation values; the internal observation values ​​are used to verify the interpolation data and calibrate the weight ratio coefficient; the optimal weight coefficient is determined, and then the data of the internal unmeasured points are simulated through all observation points to realize the processing of spatial asynchronous data.

Images