Soil element spectrum detection analysis method and system based on machine learning

Abstract

The invention discloses a soil element spectrum detection analysis method and system based on machine learning, and belongs to the technical field of smart agriculture. Field soil element data, plant growth data and environment data are obtained, and an element mineralization release model, an element loss model, a plant absorption model and a pollution element influence model are constructed; the method comprises the following steps: evaluating element mineralization release amount, plant element absorption amount and influence of harmful elements on plant absorption, constructing a soil element content prediction model, introducing the mineralization release amount, element loss amount, plant absorption amount and pollution element influence into the soil element content prediction model to calculate future soil element content, and constructing a fertilization time prediction model. The future soil element content is introduced into the fertilization time prediction model to evaluate the fertilization time, the fertilization time is predicted by coupling the soil element dynamic evolution and plant growth absorption and combining the nutrient demand characteristics of the plant in different growth states, and the plant planting fertilization efficiency is improved.

Description

Technical Field

[0001] This application belongs to the technical field of intelligent agriculture, specifically a method and system for soil element spectral detection and analysis based on machine learning. Background Art

[0002] In existing agricultural production, fertilization management mainly relies on regular manual sampling and soil testing, combined with agronomic experience to judge the fertilization time. However, this method has significant problems such as information lag, weak prediction ability, and disconnection from the actual fertilizer demand status of crops. Traditional fertilization schemes are often based on static soil data and plant growth status, lacking the trend of soil nutrients changing over time during the planting cycle, failing to consider the influence of environmental factors such as temperature, humidity, and soil type on processes such as soil element migration, mineralization, and leaching, not quantifying the element absorption laws of plants at each growth stage, and not achieving an accurate match between the soil element supply capacity and the dynamic fertilizer demand of crops. This leads to a dislocation between the fertilization timing and the actual fertilizer demand peak, low fertilizer utilization rate, and even resource waste and environmental pollution.

[0003] This application couples the dynamic evolution of soil elements with plant growth and absorption to dynamically predict the concentration change trend of key elements in the soil over a period of time in the future. Combining the nutrient demand characteristics of plants in different growth states, it predicts the most suitable fertilization time, realizes fertilization on demand and on time, improves the fertilization efficiency of plant planting, and realizes the development of green, efficient, and sustainable agriculture. Summary of the Invention

[0004] In view of the deficiencies of the prior art, this application proposes a method and system for soil element spectral detection and analysis based on machine learning.

[0005] To achieve the above object, this application provides the following technical solutions: A method and system for soil element spectral detection and analysis based on machine learning, which includes the following specific steps: Obtain field soil element data, plant growth data, and environmental data; Construct an element mineralization and release model, and import soil environmental data and soil element data into the element mineralization and release model to evaluate the element mineralization and release amount; Construct an element loss model, and import rainfall intensity, vegetation coverage, and soil structure parameters into the element loss model to evaluate the element loss amount; Construct a plant absorption model, and import soil element content and plant growth status into the plant absorption model to evaluate the element absorption amount of plants; Construct a pollution element influence model, and import the harmful element content into the pollution element influence model to evaluate the influence of harmful elements on plant absorption; Build a prediction model for soil element content, and import the mineralization release amount, element loss amount, plant absorption amount, and the influence of polluting elements into the prediction model for soil element content to calculate the future soil element content; Build a prediction model for fertilization time, and import the future soil element content into the prediction model for fertilization time to evaluate the fertilization time.

[0006] Preferably, the acquisition of field soil element data, plant growth data, and environmental data includes the following specific steps: S11. Real-time collect the spectral signals of soil elements through spectrometers deployed in the field, preprocess the spectral data in combination with partial least squares regression, extract the characteristic wavelength range, identify the spectral characteristic bands related to the soil element content, and obtain the soil element content and organic matter content; S12. Real-time monitor the plant growth through sensors to obtain environmental data, take pictures of the plant surface through drones, and obtain the basic information of plant planting.

[0007] Preferably, the construction of the element mineralization release model, and the import of soil environmental data and soil element data into the element mineralization release model to evaluate the element mineralization release amount includes the following specific steps: Substitute the information of the current soil temperature, soil humidity, soil type, and organic matter content of the target area obtained by monitoring into the calculation formula for the element mineralization release amount to calculate the element mineralization release amount. Among them, the calculation formula for the mineralization release amount of the i-th element is: , where is the potential release amount of the i-th element in the organic matter, is the temperature adjustment factor, is the soil humidity adjustment factor, is the soil type adjustment factor. Among them, the calculation formula for the potential release amount of the i-th element in the organic matter is: , where is the soil organic matter content, is the proportion of the i-th element in the organic matter, is the release rate of the i-th element in the organic matter. Among them, the calculation formula for the temperature adjustment factor is: , where is the multiple of the mineralization rate increase for every 10°C increase in temperature, is the reference temperature, T is the current soil temperature. Among them, the calculation formula for the soil humidity adjustment factor is: , where H is the current soil humidity, is the lowest mineralizable humidity, is the optimal humidity, is the over-wet inhibition value. Among them, the soil type adjustment factor Obtained through querying the soil classification database, different soil types affect the mineralization rate.

[0008] Preferably, for the construction of the element loss model, importing rainfall intensity, vegetation coverage, and soil structure parameters into the element loss model to evaluate the element loss amount includes the following specific steps: Substitute rainfall intensity, vegetation coverage, and soil structure parameters into the element loss amount calculation formula to calculate the element loss situation. Among them, the element loss amount calculation formula is: , where is the element loss coefficient, is the rainfall intensity, is the vegetation coverage rate, is the soil compaction degree, is the adsorption factor, is the content of the i-th element in the soil.

[0009] Preferably, for the construction of the plant absorption model, importing soil element content and plant growth status into the plant absorption model to evaluate the plant's absorption of elements includes the following specific steps: Substitute plant growth status and element absorption kinetic parameters into the plant absorption amount calculation formula to calculate the plant's absorption of soil elements. Among them, the plant absorption amount calculation formula is: , where is the plant's growth status at time t, is the relationship between soil element content and absorption amount. Among them, the element absorption kinetic parameter formula is: , where is the maximum absorption rate of the i-th element, is the half-saturation constant of the absorption rate. Among them, the plant's growth status calculation formula at time t is: , where is the basic absorption amount of the element by the plant in the current healthy growth stage, is the area of diseased spots on the plant leaf surface, is the surface area of the plant leaf, is the adjustment coefficient.

[0010] Preferably, for the construction of the polluting element impact model, importing the harmful element content into the polluting element impact model to evaluate the impact of harmful elements on plant absorption includes the following specific steps: S51. Substitute the content of elements in the soil that are not conducive to plant growth into the pollution evolution calculation formula to calculate the impact of unfavorable elements on plant absorption. Among them, the pollution evolution calculation formula is: , where is the content of the j-th harmful element, is the input rate of the pollutant source, is the natural attenuation rate of pollutants, where , is the source strength, is the deposition rate, , is the pollutant decay constant; S52. Substitute the pollution evolution into the inhibition function calculation formula to calculate the impact of harmful elements on plant absorption. The inhibition function calculation formula is: , where is the antagonistic sensitivity coefficient of the i-th element to the j-th harmful element, and m is the number of harmful elements.

[0011] Preferably, for the construction of the soil element content prediction model, importing the mineralization release amount, element loss amount, plant absorption amount, and pollution element impact into the soil element content prediction model to calculate the future soil element content includes the following specific steps: S61. Substitute the mineralization release amount, element loss amount, plant absorption amount, and inhibition function of the i-th element into the soil element evolution formula to calculate the soil element evolution content. The soil element evolution formula is: ; S62. Integrate the soil element evolution formula to calculate the soil element content at a future time. The future soil element content calculation formula is: , where is the content of the i-th element in the soil at time is the time interval.

[0012] Preferably, for the construction of the fertilization time prediction model, importing the future soil element content into the fertilization time prediction model to evaluate the fertilization time includes the following specific steps: S71. Obtain the demand for soil elements at different growth stages of the plant according to the plant historical growth data. The demand for soil elements at different growth stages is: , where is the minimum demand for the i-th element at the growth stage of the plant at time t, is the maximum demand for the i-th element at the growth stage of the plant at time t. Obtain the time point when the content first drops below the lower limit through the future soil element content calculation formula. The time point calculation formula when the content first drops below the lower limit is: , where N is all soil nutrient elements; S72. Substitute the time point when the content first drops below the lower limit into the fertilization time calculation formula to calculate the fertilization time. The fertilization time calculation formula is: , where is the fertilization absorption lag time.

[0013] A soil element spectral detection and analysis system based on machine learning, which is implemented based on the above-mentioned soil element spectral detection and analysis method based on machine learning, specifically includes: A data acquisition module for acquiring field soil element data, plant growth data, and environmental data; An element mineralization release module for evaluating the element mineralization release amount through soil environmental data and soil element data; An element loss module for evaluating the element loss amount through rainfall intensity, vegetation coverage, and soil structure parameters; A plant absorption module for evaluating the plant's absorption amount of elements through soil element content and plant growth status; A polluting element influence module for evaluating the influence of harmful elements on plant absorption through harmful element content; A soil element content prediction module for calculating the future soil element content through mineralization release amount, element loss amount, plant absorption amount, and polluting element influence; A fertilization time prediction module for predicting the fertilization time through the future soil element content.

[0014] An electronic device includes: a processor and a memory, wherein, a computer program that can be called by the processor is stored in the memory; The processor executes the above-mentioned soil element spectral detection and analysis method based on machine learning by calling the computer program stored in the memory.

[0015] A computer-readable storage medium, characterized in that it stores instructions, and when the instructions run on a computer, the computer is made to execute the above-mentioned soil element spectral detection and analysis method based on machine learning.

[0016] Compared with the prior art, the beneficial effects of this application are: This application obtains field soil element data, plant growth data, and environmental data, constructs an element mineralization release model, imports soil environmental data and soil element data into the element mineralization release model to evaluate the element mineralization release amount, constructs an element loss model, imports rainfall intensity, vegetation coverage, and soil structure parameters into the element loss model to evaluate the element loss amount, constructs a plant absorption model, imports soil element content and plant growth status into the plant absorption model to evaluate the plant's absorption amount of elements, constructs a polluting element impact model, imports the harmful element content into the polluting element impact model to evaluate the impact of harmful elements on plant absorption, constructs a soil element content prediction model, imports the mineralization release amount, element loss amount, plant absorption amount, and polluting element impact into the soil element content prediction model to calculate the future soil element content, constructs a fertilization time prediction model, imports the future soil element content into the fertilization time prediction model to evaluate the fertilization time. This application predicts the fertilization time by coupling the dynamic evolution of soil elements with plant growth and absorption, and combines the nutrient demand characteristics of plants in different growth states, so as to improve the fertilization efficiency of plant cultivation. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 It is a schematic diagram of the overall process of the soil element spectral detection and analysis method based on machine learning of this application; Figure 2 It is a flowchart for calculating the element mineralization release amount of this application; Figure 3 It is a flowchart for calculating the element loss amount of this application; Figure 4 It is a flowchart for calculating the evolution of soil element content of this application; Figure 5 It is a schematic diagram of the overall framework of the soil element spectral detection and analysis system based on machine learning of this application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0018] Next, the technical solutions in the embodiments of this application will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of this application. Obviously, the described embodiments are only a part of the embodiments of this application, rather than all the embodiments.

[0019] Embodiment 1

[0020] Please refer to Figures 1-4 , an embodiment provided by this application: a soil element spectral detection and analysis method based on machine learning, which includes the following specific steps: Obtain field soil element data, plant growth data, and environmental data; Construct an element mineralization release model, and import soil environmental data and soil element data into the element mineralization release model to evaluate the element mineralization release amount; Build an element loss model, import rainfall intensity, vegetation coverage, and soil structure parameters into the element loss model to evaluate the element loss amount; Build a plant absorption model, import soil element content and plant growth status into the plant absorption model to evaluate the plant's absorption amount of elements; Build a polluting element impact model, import the harmful element content into the polluting element impact model to evaluate the impact of harmful elements on plant absorption; Build a soil element content prediction model, import the mineralization release amount, element loss amount, plant absorption amount, and polluting element impact into the soil element content prediction model to calculate the future soil element content; Build a fertilization time prediction model, import the future soil element content into the fertilization time prediction model to evaluate the fertilization time.

[0021] It should be specifically noted in this embodiment that obtaining the field soil element data, plant growth data, and environmental data includes the following specific steps: S11. Real-time collect the spectral signals of soil elements through the spectrometers deployed in the field, preprocess the spectral data in combination with partial least squares regression, extract the characteristic wavelength range, identify the spectral characteristic bands related to the soil element content, and obtain the soil element content and organic matter content; It should be noted here that the spectral range covers regions such as visible light, near-infrared, and mid-infrared, captures the complete spectral characteristics of soil elements, continuously and real-time collects spectral data of the target farmland area through the spectrometers deployed in the field. The collected spectral signals are first subjected to noise reduction and normalization processing through a standard preprocessing process, including spectral smoothing and removing interference bands. The partial least squares regression algorithm is used to model the preprocessed spectral data. By comparing the corresponding relationship between the spectral signals and the element content in the measured soil samples, the characteristic wavelength range highly related to the target element is extracted from the full band, and the rapid inversion and estimation of the content of each main nutrient element in the soil are realized through the combination of spectral characteristics and the training model; S12. Real-time monitor plant growth through sensors to obtain environmental data, including obtaining soil humidity through soil humidity sensors, obtaining soil temperature through temperature sensors, and taking plant surface images through drones; S13. Obtain the basic information of plant planting, including plant variety, planting density, and plant growth stage, etc.

[0022] It should be specifically noted in this embodiment that building an element mineralization release model and importing the soil environmental data and soil element data into the element mineralization release model to evaluate the element mineralization release amount includes the following specific steps: Substitute the information of the current soil temperature, soil humidity, soil type, and organic matter content obtained by monitoring into the calculation formula for the mineralization release amount of elements. The calculation formula for the mineralization release amount of the i-th element is as follows: , where is the potential release amount of the i-th element in the organic matter, is the temperature adjustment factor, is the soil humidity adjustment factor, is the soil type adjustment factor. Among them, the calculation formula for the potential release amount of the i-th element in the organic matter is: , where is the soil organic matter content, is the proportion of the i-th element in the organic matter, is the release rate of the i-th element in the organic matter. Among them, the calculation formula for the temperature adjustment factor is: , where is the multiple of the mineralization rate increase when the temperature rises by 10°C, is the reference temperature, T is the current soil temperature measured by the temperature sensor. The temperature adjustment factor is used to reflect the influence of the current temperature relative to the reference temperature on the element mineralization rate. Among them, the calculation formula for the soil humidity adjustment factor is: , where H is the current soil humidity, is the lowest mineralizable humidity, is the optimal humidity, is the over-wet inhibition value. The humidity adjustment factor is used to characterize the non-linear influence of humidity on microbial activity and mineralization rate, and reaches the maximum value near the optimal humidity. Among them, the soil type adjustment factor is obtained by querying the soil classification database. Different soil types affect the mineralization rate. Calculate the potential release rate of the target element in the organic matter according to the organic matter content, and multiply the potential release amount by the temperature adjustment factor, humidity adjustment factor, and soil type adjustment factor to obtain the mineralization release amount of the element at the current moment; Exemplarily, in this embodiment, in the calculation formula of the temperature adjustment factor, is the multiple of the mineralization rate increase when the temperature rises by 10°C, is defaulted to 2.3, the reference temperature is defaulted to 20°C. In the calculation formula of the soil humidity adjustment factor, the lowest mineralizable humidity is 10% field water content, the optimal humidity is 60% field water content, and the over-wet inhibition value is 90% field water content. The soil type adjustment factor is set according to different soil types. It is defaulted to 1.2 for sandy soil, 1.0 for loam soil, 0.8 for clay soil, and 1.1 for peat soil.

[0023] In this embodiment, it should be specifically noted that to construct an element loss model and import rainfall intensity, vegetation coverage, and soil structure parameters into the element loss model to evaluate the element loss amount, the following specific steps are included: Substitute rainfall intensity, vegetation coverage, and soil structure parameters into the element loss amount calculation formula to calculate the element loss situation. Among them, the element loss amount calculation formula is: , where is the element loss coefficient obtained by fitting agricultural data and experiments, is the rainfall intensity, is the vegetation coverage rate, is the soil compaction degree obtained by on-site measurement, is the adsorption factor, which is the adsorption ability of the soil type to the element and is obtained by fitting agricultural data and experiments, is the content of the i-th element in the soil.

[0024] In this embodiment, it should be specifically noted that to construct a plant absorption model and import soil element content and plant growth status into the plant absorption model to evaluate the plant's absorption amount of elements, the following specific steps are included: Substitute the plant growth status and element absorption kinetic parameters into the plant absorption amount calculation formula to calculate the plant's absorption amount of soil elements. Among them, the plant absorption amount calculation formula is: , where is the plant growth status at time t, is the relationship between soil element content and absorption amount. Among them, the element absorption kinetic parameter formula is: , where is the maximum absorption rate of the i-th element, which is the absorption upper limit per unit time, is the half-saturation constant of the absorption rate. Among them, the plant growth status calculation formula at time t is: , where is the basic absorption amount of the element by the plant in the healthy state at the current growth stage, is the area of the disease spots on the plant leaf surface, is the area of the plant leaf surface, is the adjustment coefficient (1.3). Among them, the plant absorption amount increases with the increase of the element content in the soil and tends to be saturated at high contents. The plant absorption amount calculation formula includes processing the ratio of the element content to the set half-saturation constant and multiplying it by the growth status representing the current growth stage of the crop, reflecting the influence of the current plant state on the actual absorption amount of the element; Specifically, in the actual process of plant absorption, the absorption rate of plants for nutrients such as nitrogen and phosphorus does not increase linearly. When the element content is very low, the absorption rate approaches zero. When the content increases, the absorption rate increases rapidly. However, when the content continues to rise, the absorption capacity reaches the physiological limit and tends to be saturated.

[0025] Specifically, in this embodiment, to construct a pollution element impact model and import the harmful element content into the pollution element impact model to evaluate the impact of harmful elements on plant absorption, the following specific steps are included: S51. Substitute the content of elements in the soil that are unfavorable to plant growth into the pollution evolution calculation formula to calculate the impact of unfavorable elements on plant absorption. The pollution evolution calculation formula is: , where is the content of the j-th harmful element, is the input rate of pollutant sources, including industrial wastewater discharged externally, pesticide residues, and soil release, is the natural attenuation rate of pollutants, including rain dilution, microbial degradation, and phytoremediation. Among them, , is the source strength, including adjacent industrial emissions, application rates, etc., is the deposition rate, , is the pollutant decay constant, obtained by referring to the literature; S52. Substitute the pollution evolution into the inhibition function calculation formula to calculate the impact of harmful elements on plant absorption. The inhibition function calculation formula is: , where is the antagonistic sensitivity coefficient of the i-th element to the j-th harmful element, obtained by fitting experiments and literature, and m is the number of harmful elements.

[0026] Specifically, in this embodiment, to construct a soil element content prediction model and import the mineralization release amount, element loss amount, plant absorption amount, and pollution element impact into the soil element content prediction model to calculate the future soil element content, the following specific steps are included: S61. Substitute the mineralization release amount, element loss amount, plant absorption amount, and inhibition function of the i-th element into the soil element evolution formula to calculate the soil element evolution content. The soil element evolution formula is: ; S62. Integrate the soil element evolution formula to calculate the soil element content at a future time. The future soil element content calculation formula is: , where is the content of the i-th element in the soil at time is the time interval.

[0027] In this embodiment, it should be specifically noted that when constructing a fertilization time prediction model and importing future soil element contents into the fertilization time prediction model to evaluate the fertilization time, the following specific steps are included: S71. Obtain the demand for soil elements at different growth stages of plants based on historical plant growth data. Among them, the demand for soil elements at different growth stages is: , where is the minimum demand for the i-th element at the growth stage of the plant at time t, is the maximum demand for the i-th element at the growth stage of the plant at time t. The time point when the content first drops below the lower limit is obtained through the future soil element content calculation formula. Among them, the calculation formula for the time point when the content first drops below the lower limit is: , where N is all soil nutrient elements; S72. Substitute the time point when the content first drops below the lower limit into the fertilization time calculation formula to calculate the fertilization time. Among them, the fertilization time calculation formula is: , where is the fertilization absorption lag time. According to the constructed soil element evolution model, predict the target element contents at multiple future time points, compare them with the dynamic threshold of the demand of the planted crops, find the time point when the content first drops below the lower limit of the demand, and combine with the fertilizer efficiency lag period to generate a fertilization advice time window in advance. Consider the influence of pollution elements existing in the soil on the effective content of elements for correction, and avoid outputting due to sufficient content, which can be used as a decision basis for farmers, drones or intelligent fertilization devices for fertilization time.

[0028] The advantages of this embodiment compared with the prior art are as follows: This application obtains field soil element data, plant growth data and environmental data, constructs an element mineralization and release model, imports soil environmental data and soil element data into the element mineralization and release model to evaluate the element mineralization and release amount, constructs an element loss model, imports rainfall intensity, vegetation coverage and soil structure parameters into the element loss model to evaluate the element loss amount, constructs a plant absorption model, imports soil element content and plant growth status into the plant absorption model to evaluate the plant's absorption of elements, constructs a pollution element influence model, imports harmful element content into the pollution element influence model to evaluate the influence of harmful elements on plant absorption, constructs a soil element content prediction model, imports the mineralization and release amount, element loss amount, plant absorption amount and pollution element influence into the soil element content prediction model to calculate future soil element content, constructs a fertilization time prediction model, and imports future soil element content into the fertilization time prediction model to evaluate the fertilization time. By coupling the dynamic evolution of soil elements with plant growth and absorption, and combining the nutrient demand characteristics of plants in different growth states, this application predicts the fertilization time and improves the fertilization efficiency of plant planting.

[0029] Embodiment 2

[0030] As Figure 5 shown, the soil element spectral detection and analysis system based on machine learning is implemented based on the above-mentioned soil element spectral detection and analysis method based on machine learning, and specifically includes a data acquisition module, an element mineralization and release module, an element loss module, a plant absorption module, a pollution element influence module, a soil element content prediction module, and a fertilization time prediction module. The data acquisition module is used to acquire field soil element data, plant growth data, and environmental data; the element mineralization and release module is used to evaluate the element mineralization and release amount through soil environmental data and soil element data; the element loss module is used to evaluate the element loss amount through rainfall intensity, vegetation coverage, and soil structure parameters; the plant absorption module is used to evaluate the plant's absorption amount of elements through soil element content and plant growth status; the pollution element influence module is used to evaluate the influence of harmful elements on plant absorption through the content of harmful elements; the soil element content prediction module is used to calculate the future soil element content through mineralization and release amount, element loss amount, plant absorption amount, and pollution element influence; the fertilization time prediction module is used to predict the fertilization time through the future soil element content.

[0031] Embodiment 3

[0032] This embodiment provides an electronic device, including: a processor and a memory, wherein, a computer program that can be called by the processor is stored in the memory; The processor executes the above-mentioned soil element spectral detection and analysis method based on machine learning by calling the computer program stored in the memory.

[0033] This electronic device may have relatively large differences due to configuration or performance, and can include one or more processors (Central Processing Units, CPUs) and one or more memories. Among them, at least one computer program is stored in the memory, and this computer program is loaded and executed by the processor to implement the soil element spectral detection and analysis method based on machine learning provided by the above method embodiment. This electronic device can also include other components for realizing the functions of the device. For example, this electronic device can also have components such as wired or wireless network interfaces and input / output interfaces for data input and output. This embodiment will not be elaborated here.

[0034] Embodiment 4

[0035] This embodiment proposes a computer-readable storage medium, on which a rewritable computer program is stored; When the computer program runs on a computer device, the computer device is enabled to execute the above-mentioned soil element spectral detection and analysis method based on machine learning.

[0036] For example, a computer-readable storage medium can be a read-only memory (ROM), a random access memory (RAM), a compact disc read-only memory (CD-ROM), magnetic tape, a floppy disk, an optical data storage device, etc.

[0037] The above embodiments can be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the processes or functions according to the embodiments of the present application are generated in whole or in part. The computer can be a general-purpose computer, a dedicated computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via a wired network or / and a wireless network. The computer-readable storage medium can be any available medium that the computer can access or a data storage device such as a server or a data center that contains one or more collections of available media. The available medium can be a magnetic medium (e.g., a floppy disk, a hard disk, magnetic tape), an optical medium (e.g., a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state drive.

Claims

1. A method for detecting and analyzing soil element spectra based on machine learning, characterized in that, It includes the following specific steps: Obtain field soil element data, plant growth data, and environmental data; Construct an element mineralization and release model, and import the soil environmental data and soil element data into the element mineralization and release model to evaluate the element mineralization and release amount; Construct an element loss model, and import the rainfall intensity, vegetation coverage, and soil structure parameters into the element loss model to evaluate the element loss amount; Construct a plant absorption model, and import the soil element content and plant growth status into the plant absorption model to evaluate the plant's absorption amount of elements; Construct a contaminated element impact model, and import the harmful element content into the contaminated element impact model to evaluate the impact of harmful elements on plant absorption; Construct a soil element content prediction model, and import the mineralization and release amount, element loss amount, plant absorption amount, and contaminated element impact into the soil element content prediction model to calculate the future soil element content; Construct a fertilization time prediction model, and import the future soil element content into the fertilization time prediction model to evaluate the fertilization time.

2. The method for detecting and analyzing soil element spectra based on machine learning according to claim 1, wherein The obtaining of the field soil element data, plant growth data, and environmental data includes the following specific steps: Real-time collect the spectral signals of soil elements through the spectrometers deployed in the field, preprocess the spectral data in combination with partial least squares regression, extract the characteristic wavelength intervals, identify the spectral characteristic bands related to the soil element content, and obtain the soil element content and organic matter content; Real-time monitor the plant growth through sensors to obtain environmental data, take plant surface images by drones, and obtain the basic information of plant planting.

3. The method for soil element spectral detection and analysis based on machine learning according to claim 2, characterized in that The constructing of the element mineralization and release model, and importing the soil environmental data and soil element data into the element mineralization and release model to evaluate the element mineralization and release amount includes the following specific steps: Substitute the currently monitored soil temperature, soil humidity, soil type, and organic matter content information of the target area into the calculation formula of the element mineralization and release amount, multiply the potential release rate by the temperature adjustment factor, humidity adjustment factor, and soil type adjustment factor to obtain the element mineralization and release amount at the current moment; Calculate the potential release amount of the target element in the organic matter according to the organic matter content, the proportion of the element in the organic matter, and the release rate; Calculate the temperature adjustment factor according to the soil temperature based on the mineralization rate increase multiple; Calculate the humidity adjustment factor according to the soil humidity, the minimum mineralizable humidity, the optimum humidity, and the over-wet inhibition value; Determine the soil type adjustment factor according to the soil type.

4. The method for detecting and analyzing soil element spectra based on machine learning according to claim 3, wherein, The constructing of the element loss model, and importing the rainfall intensity, vegetation coverage, and soil structure parameters into the element loss model to evaluate the element loss amount includes the following specific steps: The element loss amount is determined by the rainfall intensity, vegetation coverage, soil structure parameters, and the migration characteristics of the target element. Among them, the loss amount is positively correlated with the rainfall intensity, and is inhibited by the crop coverage degree, and the soil compaction has a buffering effect on the loss. The loss amount of the target element is obtained by combining the rainfall intensity, the coverage ratio, and the soil buffer factor.

5. The method for detecting and analyzing soil element spectra based on machine learning according to claim 4, wherein The constructing of the plant absorption model, and importing the soil element content and plant growth status into the plant absorption model to evaluate the plant's absorption amount of elements includes the following specific steps: The plant absorption amount is determined according to the current content of the target element in the soil, the plant growth state, and the element absorption kinetic parameters; The plant growth state is calculated through the basic absorption amount of the element in the current growth stage of the plant in a healthy state and the area of disease spots on the plant leaf surface; The element absorption kinetic parameter is the product of the maximum absorption capacity and the current element content, divided by the sum of the current content and the set half-saturation content constant, and adjusted in combination with the exponential factor of the current growth stage of the crop to calculate the plant absorption difference in different growth periods.

6. The method for detecting and analyzing soil element spectra based on machine learning according to claim 5, characterized in that The steps for constructing a pollution element impact model and importing the harmful element content into the pollution element impact model to evaluate the impact of harmful elements on plant absorption are as follows: Substitute the content of elements in the soil that are unfavorable to plant growth into the pollution evolution calculation formula to calculate the impact of unfavorable elements on plant absorption. The pollution evolution calculation formula is as follows: , where is the content of the j-th harmful element, is the input rate of pollutant sources, is the natural attenuation rate of pollutants. Among them, , is the source strength, is the conversion or deposition rate, , is the pollutant decay constant; Substitute the pollution evolution into the calculation formula of the inhibition function to calculate the impact of harmful elements on plant absorption. The calculation formula of the inhibition function is as follows: , where is the antagonistic sensitivity coefficient of the i-th element to the j-th harmful element, is the content of the i-th element in the soil, and m is the number of harmful elements.

7. The method for detecting and analyzing soil element spectra based on machine learning according to claim 6, characterized in that The steps for constructing a soil element content prediction model and importing the mineralization release amount, element loss amount, plant absorption amount, and pollution element impact into the soil element content prediction model to calculate the future soil element content are as follows: Substitute the mineralization release amount, element loss amount, plant absorption amount, and inhibition function of the i-th element into the soil element evolution formula to calculate the soil element evolution content. Among them, the soil element evolution formula is: , where is the mineralization release amount of the i-th element at time t, is the plant absorption amount of the i-th element at time t, is the element loss amount of the i-th element at time t; Integrate the soil element evolution formula to calculate the soil element content at future times. The formula for calculating the future soil element content is as follows: , where is the content of the i-th element in the soil at time and is the time interval.

8. The method for detecting and analyzing soil element spectra based on machine learning according to claim 7, characterized in that, The steps for constructing a fertilization time prediction model and importing the future soil element content into the fertilization time prediction model to evaluate the fertilization time are as follows: Obtain the demand for soil elements at different growth stages of plants based on the historical growth data of plants. Among them, the demand for soil elements at different growth stages is as follows: , where is the minimum demand for the i-th element at the growth stage of the plant at time t, is the maximum demand for the i-th element at the growth stage of the plant at time t. Obtain the time point when the content first drops below the lower limit through the future soil element content calculation formula. Among them, the calculation formula for the time point when the content first drops below the lower limit is: , where N is all soil nutrient elements; Substitute the time point when the content first drops below the lower limit into the fertilization time calculation formula to calculate the fertilization time. The fertilization time calculation formula is as follows: , where is the fertilization absorption lag time.

9. A soil element spectral detection and analysis system based on machine learning, which is implemented based on the soil element spectral detection and analysis method based on machine learning according to any one of claims 1-8, characterized in that, Specifically, it includes: A data acquisition module for acquiring field soil element data, plant growth data, and environmental data; An element mineralization release module for evaluating the element mineralization release amount through soil environmental data and soil element data; An element loss module for evaluating the element loss amount through rainfall intensity, vegetation coverage, and soil structure parameters; A plant absorption module for evaluating the absorption amount of the element by the plant through the soil element content and the plant growth state; A pollution element impact module for evaluating the impact of harmful elements on plant absorption through the harmful element content; A soil element content prediction module for calculating the future soil element content through the mineralization release amount, element loss amount, plant absorption amount, and pollution element impact; A fertilization time prediction module for predicting the fertilization time through the future soil element content.

10. An electronic device, comprising: A processor and a memory, wherein the memory stores a computer program that can be called by the processor; It is characterized in that the processor executes the machine learning-based soil element spectral detection and analysis method according to any one of claims 1-8 by calling the computer program stored in the memory.

Images