Machine learning based rhizosphere soil health assessment method and system

Abstract

The application relates to the technical field of data processing, and discloses a rhizosphere soil health evaluation method and system based on machine learning. The method comprises the following steps: collecting multi-dimensional parameter data of a rhizosphere region, processing abnormal values by using a rhizosphere time sequence median filtering method and calculating a comprehensive index, establishing an evaluation model by using a hybrid algorithm combined with an attention mechanism to fuse static and time sequence characteristics, outputting a health score and judging a state category, and finally generating an optimized scheme through a rhizosphere factor weight self-learning algorithm, so that complete closed-loop management of dynamic evaluation and intelligent adjustment of rhizosphere soil health is realized. The application solves the problem that the prior art cannot dynamically and accurately evaluate the rhizosphere microenvironment, and improves the accuracy of rhizosphere soil health evaluation and intelligent adjustment capability.

Description

Technical Field

[0001] This application relates to the field of data processing technology, and in particular to a method and system for assessing rhizosphere soil health based on machine learning. Background Technology

[0002] In current technologies, soil health assessment primarily relies on traditional laboratory analytical methods, assessing soil health by collecting soil samples and measuring their chemical, physical, and biological properties. With the rapid development of the Internet of Things (IoT), sensor technology, and machine learning, some studies have begun to apply machine learning techniques to soil pollution assessment, particularly in predicting the content, spatial distribution, and source identification of potentially toxic elements (PTEs). Combining hyperspectral data with machine learning methods enables low-cost prediction of PTE content over large-scale areas, and machine learning algorithms integrating environmental covariates also offer superior performance to traditional geostatistical methods in spatial prediction.

[0003] However, existing technologies have significant shortcomings: traditional laboratory analysis methods are limited by high time costs, poor spatial representativeness, and lack of dynamic monitoring; existing machine learning methods mainly focus on the static content prediction and large-scale spatial distribution of PTE, and cannot accurately assess the dynamic changes of the special microenvironment of plant rhizosphere; existing algorithms mostly use a single machine learning model and lack the ability to effectively fuse and process multi-source heterogeneous data of rhizosphere soil, resulting in limited assessment accuracy.

[0004] Based on the above analysis, it is evident that existing technologies lack soil health assessment methods specifically tailored to the characteristics of the rhizosphere microenvironment. Since the rhizosphere is the most active microenvironment for plant-soil interaction, its soil parameter variation patterns differ significantly from those of general soil. Therefore, it is necessary to develop hybrid algorithms capable of simultaneously processing both static characteristics and dynamic temporal information of rhizosphere soil. Furthermore, existing technologies cannot automatically generate targeted regulation plans based on rhizosphere soil health assessment results, lacking closed-loop management capabilities from assessment to regulation. This limits the level of intelligence and practicality of soil health management. Summary of the Invention

[0005] This application provides a machine learning-based method and system for assessing rhizosphere soil health, which addresses the problem that existing technologies cannot dynamically and accurately assess the rhizosphere microenvironment, thereby improving the accuracy and intelligent regulation capabilities of rhizosphere soil health assessment.

[0006] Firstly, this application provides a machine learning-based method for assessing rhizosphere soil health, the machine learning-based method for assessing rhizosphere soil health comprising:

[0007] Step S1: Collect data on pH, electrical conductivity, temperature, humidity, nutrient concentration, and microbial activity in the rhizosphere region of the target plant as the original rhizosphere soil dataset;

[0008] Step S2: Correct outliers in the original rhizosphere soil dataset by using rhizosphere time-series median filtering, calculate the rhizosphere nutrient comprehensive index and rhizosphere microbial activity index, and obtain standardized rhizosphere soil data.

[0009] Step S3: Train the standardized rhizosphere soil data using a hybrid algorithm, and use an attention mechanism to fuse static features and temporal information to establish a rhizosphere soil health assessment algorithm.

[0010] Step S4: Output the rhizosphere soil health score through the rhizosphere soil health assessment algorithm, and determine the rhizosphere soil health status category based on the rate of change of the health score.

[0011] Step S5: Based on the rhizosphere soil health status category, the regulator dosage parameters are calculated using a rhizosphere factor weight self-learning algorithm, and the rhizosphere soil optimization scheme is output.

[0012] Secondly, this application provides a machine learning-based rhizosphere soil health assessment system, the machine learning-based rhizosphere soil health assessment system comprising:

[0013] The data acquisition module is used to collect data on pH, electrical conductivity, temperature and humidity, nutrient concentration and microbial activity in the rhizosphere region of the target plant as the raw dataset of the rhizosphere soil.

[0014] The correction module is used to correct outliers in the original rhizosphere soil dataset using the rhizosphere time-series median filtering method, calculate the rhizosphere nutrient comprehensive index and the rhizosphere microbial activity index, and obtain standardized rhizosphere soil data.

[0015] The training module is used to train the standardized rhizosphere soil data through a hybrid algorithm, and the attention mechanism integrates static features and temporal information to establish a rhizosphere soil health assessment algorithm.

[0016] The output module is used to output a rhizosphere soil health score through the rhizosphere soil health assessment algorithm, and to determine the rhizosphere soil health status category based on the rate of change of the health score.

[0017] The calculation module is used to calculate the dosage parameters of regulators based on the rhizosphere soil health status category and through a rhizosphere factor weight self-learning algorithm, and output the rhizosphere soil optimization scheme.

[0018] Thirdly, a machine learning-based rhizosphere soil health assessment device is provided, comprising: a memory and at least one processor, wherein the memory stores instructions; the at least one processor invokes the instructions in the memory to cause the machine learning-based rhizosphere soil health assessment device to perform the aforementioned machine learning-based rhizosphere soil health assessment method.

[0019] Fourthly, a computer-readable storage medium is provided, wherein instructions are stored therein, which, when executed on a computer, cause the computer to perform the aforementioned machine learning-based rhizosphere soil health assessment method.

[0020] The technical solution provided in this application collects data on pH, electrical conductivity, temperature and humidity, nutrient concentration, and microbial activity in the rhizosphere region of the target plant as the original dataset for rhizosphere soil. Compared with traditional soil assessment methods that only focus on general soil parameters, this application specifically addresses the unique characteristics of the rhizosphere microenvironment by collecting multi-dimensional data, which can more accurately reflect the true interaction between plant roots and soil. By correcting outliers using the rhizosphere time-series median filtering method and calculating the rhizosphere nutrient comprehensive index and rhizosphere microbial activity index, the problem of noise interference in rhizosphere environmental data is effectively solved. Simultaneously, a composite evaluation index system specifically suitable for rhizosphere soil is constructed, which has higher accuracy and reliability compared to existing single-parameter assessment methods. By training standardized rhizosphere soil data using a hybrid algorithm and integrating static features and time-series information through an attention mechanism, a rhizosphere soil health assessment algorithm is established. This overcomes the limitations of existing single-model processing capabilities, capturing complex nonlinear relationships between rhizosphere soil parameters and identifying time-series change patterns, significantly improving the accuracy and stability of rhizosphere soil health status prediction.

[0021] By combining a random forest module to process static feature relationships and a temporal convolutional neural network module to capture dynamic change patterns, this approach fully leverages the spatiotemporal characteristics of rhizosphere environmental data, exhibiting stronger generalization ability and adaptability compared to traditional machine learning methods. The rhizosphere soil health status category is determined based on the rate of change in health scores, achieving a shift from static assessment to dynamic early warning, enabling timely detection of early signs of rhizosphere soil health problems. Based on the rhizosphere soil health status category, a rhizosphere factor weight self-learning algorithm calculates regulator dosage parameters and outputs rhizosphere soil optimization schemes, realizing a complete closed-loop management from assessment to regulation. Compared to existing technologies that only provide assessment results, this application can automatically generate targeted regulation suggestions, significantly improving the intelligence level of rhizosphere soil management. Attached Figure Description

[0022] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0023] Figure 1 This is a schematic diagram of one embodiment of the machine learning-based rhizosphere soil health assessment method in this application.

[0024] Figure 2 This is a flowchart of the rhizosphere soil health assessment based on machine learning in the embodiments of this application;

[0025] Figure 3 The training and validation loss reduction curves are shown for the training effect of the machine learning-based rhizosphere soil health assessment algorithm in this application embodiment.

[0026] Figure 4 This is a correlation comparison chart of the prediction results of the training effect of the rhizosphere soil health assessment algorithm based on machine learning in the embodiments of this application;

[0027] Figure 5 This is a graph showing the backpropagation gradient mean of the training effect of the machine learning-based rhizosphere soil health assessment algorithm in the embodiments of this application.

[0028] Figure 6 This is a schematic diagram of one embodiment of the machine learning-based rhizosphere soil health assessment system in this application.

[0029] Figure 7 This is a schematic block diagram of the rhizosphere soil health assessment device based on machine learning in an embodiment of the present invention. Detailed Implementation

[0030] This application provides a method and system for assessing rhizosphere soil health based on machine learning. The terms "first," "second," "third," "fourth," etc. (if present) in the specification, claims, and accompanying drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments described herein can be implemented in a sequence other than that illustrated or described herein. Furthermore, the terms "comprising" or "having" and any variations thereof are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0031] For ease of understanding, the specific process of the embodiments of this application is described below. Please refer to [link / reference]. Figure 1 One embodiment of the machine learning-based rhizosphere soil health assessment method in this application includes:

[0032] Step S1: Collect data on pH, electrical conductivity, temperature, humidity, nutrient concentration, and microbial activity in the rhizosphere region of the target plant as the original rhizosphere soil dataset;

[0033] Step S2: Correct outliers in the original rhizosphere soil dataset by using rhizosphere time-series median filtering, calculate the rhizosphere nutrient comprehensive index and rhizosphere microbial activity index, and obtain standardized rhizosphere soil data.

[0034] Step S3: Train standardized rhizosphere soil data using a hybrid algorithm, and integrate static features and temporal information using an attention mechanism to establish a rhizosphere soil health assessment algorithm.

[0035] Step S4: Output the rhizosphere soil health score through the rhizosphere soil health assessment algorithm, and determine the rhizosphere soil health status category based on the rate of change of the health score.

[0036] Step S5: Based on the rhizosphere soil health status category, the regulator dosage parameters are calculated using a rhizosphere factor weight self-learning algorithm, and the rhizosphere soil optimization scheme is output.

[0037] It is understood that the executing entity of this application can be a rhizosphere soil health assessment system based on machine learning, or it can be a terminal or a server; the specific implementation is not limited here. This application's embodiments use a server as an example for illustration.

[0038] Specifically, multi-dimensional data from the rhizosphere region of target plants, including key parameters such as pH, electrical conductivity, temperature and humidity, nutrient concentration, and microbial activity, constitute the raw dataset of rhizosphere soil. During data acquisition, a sensor network is used to monitor the rhizosphere microenvironment in real time, ensuring high-frequency acquisition and accurate recording. Rhizosphere time-series median filtering is used to correct outliers in the raw data, ensuring accuracy and reliability. Based on the corrected data, a comprehensive rhizosphere nutrient index and a microbial activity index are further calculated. These comprehensive indices more comprehensively reflect soil health, resulting in standardized rhizosphere soil data. Based on the standardized data, a hybrid algorithm is used to train the data, fusing static features and time-series information using an attention mechanism. The hybrid algorithm effectively handles the complex relationship between static and time-series features, capturing nonlinear patterns using a deep learning model to establish a rhizosphere soil health assessment algorithm. This algorithm outputs the health status based on the rhizosphere soil health score and determines the rhizosphere soil health status category by analyzing the rate of change of the health score. Based on the scores, the system can classify soil health status into healthy, moderately healthy, or unhealthy, providing a basis for subsequent soil regulation. Based on the health status category, the system further analyzes key influencing factors of the soil through a rhizosphere factor weight self-learning algorithm, automatically calculating regulator dosage parameters. This algorithm can output corresponding regulator dosage schemes based on changes in factors such as soil pH deviation, nutrient concentration, and microbial activity, aiming to optimize rhizosphere soil health. This process achieves closed-loop management from assessment to regulation, not only enabling real-time monitoring of rhizosphere soil health but also providing targeted and precise soil optimization solutions, greatly improving the intelligence and accuracy of soil management.

[0039] In one specific embodiment, step S1 further includes:

[0040] A rhizosphere sensor network is deployed in the rhizosphere region of the target plant to simultaneously monitor multiple parameters of the rhizosphere microenvironment and obtain rhizosphere environment monitoring data.

[0041] Rhizosphere environmental monitoring data were collected at preset time intervals: pH data was collected every 30 minutes, electrical conductivity data was collected every 30 minutes, temperature and humidity data was collected every 60 minutes, nutrient concentration data was collected every 120 minutes, and microbial activity data was collected every 120 minutes, thus obtaining rhizosphere soil time-series data.

[0042] The rhizosphere soil time-series data was transmitted via a wireless communication module. A timestamp, sensor identifier, and location information were added to each data record to obtain the original rhizosphere soil dataset.

[0043] Data integrity was verified based on the original rhizosphere soil dataset, and missing data was marked to obtain the original rhizosphere soil dataset.

[0044] Specifically, a rhizosphere sensor network is deployed in the rhizosphere region of the target plant to monitor multiple key parameters in the rhizosphere microenvironment in real time, including pH, conductivity, temperature and humidity, nutrient concentration, and microbial activity. The sensor network synchronously collects this data at preset time intervals to ensure high-frequency and comprehensive data collection. pH and conductivity data are collected every 30 minutes, temperature and humidity data every 60 minutes, nutrient concentration data every 120 minutes, and microbial activity data every 120 minutes, thus forming a time-series rhizosphere soil dataset. To ensure smooth and stable data transmission, all collected data is transmitted in real time via a wireless communication module. Each data record is automatically added with a timestamp, sensor identifier, and location information, ensuring data integrity and traceability, generating a raw rhizosphere soil dataset. After the raw rhizosphere soil dataset is generated, the system verifies the integrity of the data to ensure no missing data or abnormal records. Missing or incomplete data is marked for special processing in subsequent data analysis, resulting in a complete and reliable raw rhizosphere soil dataset. In this way, the entire data collection process ensured the accuracy, timeliness, and high completeness of the information, providing a solid data foundation for subsequent rhizosphere soil health assessment.

[0045] For example, in the deployment of a rhizosphere sensor network, sensors are precisely positioned within the rhizosphere region of the target plant to simultaneously monitor multiple key parameters of the rhizosphere soil. For instance, pH sensors collect data on soil acidity and alkalinity changes in real time, recording data every 30 minutes to ensure sufficient monitoring of dynamic changes in the soil environment. Conductivity sensors also collect data at the same time intervals to reflect the dissolved salt content of the soil, helping to analyze soil fertility. Temperature and humidity sensors collect data every 60 minutes to capture changes in soil and air temperature and humidity, aiding in the study of the rhizosphere's climatic adaptability. Nutrient concentration and microbial activity data are collected every 120 minutes to ensure accurate recording of soil nutrient and microbial activity levels over extended periods. This data is transmitted via wireless communication modules, accompanied by timestamps, sensor identifiers, and location information, ensuring that each record accurately corresponds to a specific time and location, avoiding data corruption or loss.

[0046] For example, during data integrity verification, the system checks the collected raw rhizosphere soil dataset to ensure that each data item is valid. If some data is missing at certain times, the system automatically marks the missing data as "missing" or "abnormal," allowing for appropriate compensation or correction measures in subsequent data processing. For instance, if pH data for a certain moment cannot be successfully collected, the system will record it using a marking system, and this data gap can be filled later using interpolation methods or other means to ensure data consistency and integrity. In this way, through rigorous verification and processing of data integrity, the system ensures that the obtained dataset accurately reflects the health status of the rhizosphere soil, providing accurate basic data for subsequent assessment and optimization planning.

[0047] In one specific embodiment, step S2 further includes:

[0048] The pH, electrical conductivity, temperature and humidity, nutrient concentration and microbial activity data in the original rhizosphere soil dataset were tested for range. Values ​​that exceeded the preset range of the rhizosphere environment were marked as outliers, and outlier-marked data were obtained.

[0049] Rhizosphere time-series median filtering is performed on outlier-marked data, and outliers are replaced and corrected by the median of adjacent time point data to obtain rhizosphere soil corrected data.

[0050] The nitrogen, phosphorus, potassium and organic acid concentrations in the rhizosphere soil correction data were calculated using a weighted summation formula to obtain the rhizosphere nutrient comprehensive index.

[0051] The urease activity, phosphatase activity, and sucrase activity in the rhizosphere soil correction data were weighted according to the enzyme activity weighting coefficient to obtain the rhizosphere microbial activity index and rhizosphere soil standardized data.

[0052] Specifically, the pH, electrical conductivity, temperature and humidity, nutrient concentration, and microbial activity data in the original rhizosphere soil dataset are subjected to range testing. Values ​​exceeding the preset range of the rhizosphere environment are marked as outliers, thus obtaining outlier-labeled data. During the outlier labeling process, the system compares each parameter against a set threshold. If a data item exceeds the normal range, it is marked as an outlier, ensuring the data quality and validity of subsequent processing. Based on these outlier-labeled data, rhizosphere time-series median filtering is performed. Outliers are replaced by the median of adjacent time points, thereby correcting the outliers in the data. The formula for median filtering is: ,in, This is the corrected data at the current moment. This consists of raw data from k consecutive time points. This processing method effectively reduces the impact of noise and preserves the overall trend of the data, making the data smoother and more stable, resulting in corrected rhizosphere soil data. The nitrogen, phosphorus, potassium, and organic acid concentrations in the corrected rhizosphere soil data are calculated using a weighted summation formula to obtain the rhizosphere nutrient comprehensive index. Assume the weights of each nutrient concentration are as follows: Therefore, the formula for calculating the rhizosphere nutrient comprehensive index is: ,in, These are the concentrations of nitrogen, phosphorus, and potassium, respectively. The urease, phosphatase, and sucrase activities in the rhizosphere soil correction data are also weighted according to preset enzyme activity weighting coefficients to obtain the rhizosphere microbial activity index. The weighted calculation formula for the enzyme activity index is: ,in, These are urease activity, phosphatase activity, and sucrase activity. These are the weighting coefficients for the activities of each enzyme. Through these processing steps, standardized rhizosphere soil data were obtained, providing reliable and accurate data support for subsequent health assessments.

[0053] Taking a rhizosphere soil health assessment at a farm as an example, a sensor network is deployed in the rhizosphere regions of different crops to collect real-time data on pH, conductivity, temperature, humidity, nutrient concentration, and microbial activity. Every 30 minutes, the sensors record pH and conductivity data; every 60 minutes, temperature and humidity data; and every 120 minutes, nutrient concentration and microbial activity data. The collected data is transmitted to a server via a wireless communication module, and includes timestamps, sensor identifiers, and location information. During data processing, an anomaly is detected in a pH measurement, exceeding the preset soil pH range (e.g., 3.0 to 9.0). This anomaly is flagged and corrected using a time-series median filtering method, replacing the anomaly with the median of adjacent time points to obtain the corrected data. Next, the system calculates a weighted sum of nitrogen, phosphorus, and potassium concentrations, assuming weights of 0.3, 0.3, and 0.4, respectively, to derive a comprehensive rhizosphere nutrient index. For microbial activity, the system multiplies the activity data of urease, phosphatase, and sucrase by their respective weighting coefficients (e.g., 0.4, 0.3, 0.3) to derive the rhizosphere microbial activity index. This processed data is used to generate a rhizosphere soil health assessment report and provides a basis for subsequent soil conditioning programs.

[0054] In one specific embodiment, step S3 further includes:

[0055] Standardized rhizosphere soil data is input into a random forest module for nonlinear feature relationship learning, and the complex correlation between rhizosphere soil parameters is modeled to obtain rhizosphere static feature weights.

[0056] Standardized rhizosphere soil data is input into a temporal convolutional neural network module for temporal change pattern recognition. The temporal series variation patterns of rhizosphere soil parameters are extracted and processed to obtain rhizosphere temporal feature vectors.

[0057] Based on the rhizosphere static feature weights and rhizosphere temporal feature vectors, an attention mechanism is used to perform fusion calculation. The static features and temporal information are weighted and merged according to dynamic weights to obtain the comprehensive characteristics of rhizosphere soil.

[0058] The rhizosphere soil health assessment algorithm is obtained by iteratively optimizing and training the comprehensive characteristics of rhizosphere soil using the gradient descent method and adaptively adjusting the parameters of the hybrid algorithm.

[0059] Specifically, standardized rhizosphere soil data is input into a random forest module for nonlinear feature relationship learning. The model automatically identifies and establishes nonlinear relationships between various soil parameters by analyzing their complex correlations. These relationships are used to calculate the weights of static rhizosphere features. Through this process, the model can understand the relative importance of different soil parameters in the overall health assessment and provide accurate feature weights for subsequent analyses. For example, assuming the weights of soil pH, temperature, humidity, and nutrient concentration are... The comprehensive features of static features can be expressed as: ,in, It is a static characteristic of the rhizosphere. Represents pH value. Represents temperature and humidity. Represents nutrient concentration. These are the weights of each feature. These standardized data are input into a temporal convolutional neural network module to identify temporal change patterns. By extracting the time-series variation patterns of soil parameters, the model can identify the trends and patterns of soil state changes over time, thereby generating a rhizosphere temporal feature vector. For example, the calculation of the temporal feature vector can be expressed as: ,in, It is a time-series feature vector. These are soil parameter data for the previous moment, the current moment, and the next moment, respectively. This represents the function for extracting temporal patterns in a temporal convolutional neural network. Based on the static feature weights and temporal feature vectors of the rhizosphere, an attention mechanism is used for fusion calculation. The model weights and merges the static features and temporal information according to dynamic weights. In this way, the static characteristics and dynamic changes of the rhizosphere soil are effectively fused to obtain a comprehensive representation of the rhizosphere soil features. The comprehensive rhizosphere soil features are iteratively optimized and trained using the gradient descent method. The model continuously adjusts the parameters of the hybrid algorithm for adaptive optimization, enabling the health assessment algorithm to accurately predict soil health status.

[0060] Specifically, in the attention mechanism fusion calculation part, the model combines rhizosphere static features and temporal features, using dynamic weights to weight and merge these two parts of information. In essence, the core of the attention mechanism is to dynamically adjust the weights based on the importance of each feature in the current soil health assessment. Assuming the weights of the rhizosphere static features are... The weights of the root time series features are The integrated characteristics after fusion It can be expressed as the following weighted sum: ,in, Indicates the static characteristics of the rhizosphere. Indicating the rhizosphere temporal characteristics, and These are the dynamic weighting coefficients in the attention mechanism, representing the importance of static and temporal features in the current health assessment. By calculating these two weighting coefficients, the model can adaptively adjust the contribution of features in the assessment based on the characteristics of the actual data. Specifically, and The model will be continuously updated during the training process, enabling it to better capture the changing patterns of soil characteristics and obtain a comprehensive representation of rhizosphere soil features. Figure 2 The figure illustrates the rhizosphere soil health assessment process based on machine learning.

[0061] For example, when standardized rhizosphere soil data is input into the Random Forest module for nonlinear feature relationship learning, the system analyzes multiple parameters such as pH, electrical conductivity, temperature and humidity, nutrient concentration, and microbial activity. During this process, the Random Forest algorithm discovers complex nonlinear relationships between different soil parameters through training data. For instance, there may be a hidden correlation between electrical conductivity and nutrient concentration. These relationships help the model assign appropriate feature weights to each parameter, thereby providing accurate static features for subsequent evaluation.

[0062] For example, when combining rhizosphere static feature weights and rhizosphere temporal feature vectors through an attention mechanism for fusion calculation, the model adaptively allocates weights based on the dynamic changes and importance of the data. For instance, when soil nutrient concentration has a greater impact on health assessment at a certain stage, the model automatically increases the weight of this feature, ensuring that soil health assessment can make more accurate judgments based on feature changes at different stages.

[0063] For example, when iteratively optimizing the comprehensive characteristics of rhizosphere soil using gradient descent, the model adjusts the parameters of the hybrid algorithm through backpropagation to optimize the rhizosphere soil health assessment algorithm. For instance, if the assessment results at a certain stage deviate significantly, gradient descent will adjust the model parameters based on error feedback, enabling the system to more accurately predict soil health status in subsequent training, resulting in an intelligent soil health assessment tool.

[0064] In one specific embodiment, the process of performing iterative optimization training on the comprehensive characteristics of rhizosphere soil using the gradient descent method and adaptively adjusting the parameters of the hybrid algorithm to obtain the rhizosphere soil health assessment algorithm can specifically include the following steps:

[0065] The comprehensive characteristics of rhizosphere soil are matched with the rhizosphere soil health labeling data, and a corresponding health score label is assigned to each group of rhizosphere soil samples to obtain the rhizosphere soil training dataset.

[0066] The loss function value is calculated based on the rhizosphere soil training dataset. The error between the prediction result and the true label is quantified and processed to obtain the rhizosphere soil health prediction error.

[0067] The rhizosphere soil health prediction error is calculated by backpropagation algorithm. The gradient values ​​of the parameters of each layer in the hybrid algorithm are differentiated to obtain the gradient vector of rhizosphere soil parameters.

[0068] The parameter update operation is performed based on the gradient vector of rhizosphere soil parameters. The weight parameters of the hybrid algorithm are iteratively adjusted according to the learning rate to obtain the rhizosphere soil health assessment algorithm.

[0069] Specifically, the system matches the comprehensive features of rhizosphere soil with rhizosphere soil health labeling data, associating each group of rhizosphere soil samples with a corresponding health score label to form a rhizosphere soil training dataset. These labels represent the health status of each sample. Through this step, the model obtains a training dataset, which includes multiple features and corresponding health scores. Then, based on the rhizosphere soil training dataset, the system calculates a loss function value to measure the error between the prediction result and the true label. Through quantification, the error is converted into a numerical value, yielding the rhizosphere soil health prediction error. The loss function commonly used is the mean squared error (MSE) or the cross-entropy loss function, and the specific calculation formula is as follows: ,in, It is the value of the loss function. It is the first The true label of each sample It is the health score predicted by the model. This refers to the sample size. The rhizosphere soil health prediction error is calculated using a backpropagation algorithm. The model uses the chain rule to differentiate the parameters of each layer in the hybrid algorithm, calculating the gradient value of each parameter, resulting in the gradient vector of the rhizosphere soil parameters. These gradient values ​​represent the contribution of each parameter to the error, helping to determine which parameters need adjustment and in what direction. Through backpropagation, the model can capture the relationship between features and errors and optimize the algorithm's performance based on this information. Based on the rhizosphere soil parameter gradient vector, the system performs parameter update operations, iteratively adjusting the weight parameters of the hybrid algorithm using gradient descent. The update operation follows a set learning rate to ensure the model gradually converges to the optimal solution. The learning rate controls the step size of each update; too large a rate may cause oscillations, while too small a rate may lead to slow convergence. Through continuous iterative updates, an optimized rhizosphere soil health assessment algorithm is obtained, which can accurately assess soil health and has strong generalization ability. (Reference) Figures 3-5 The image shows a comparison of the training effects of machine learning-based rhizosphere soil health assessment algorithms. Figure 3 The diagram shows the training and validation loss descent curves. Figure 4 The correlation comparison of the prediction results is shown. Figure 5 The diagram illustrates the backpropagation gradient mean method, where RF-only uses only random forests, ignoring temporal dynamics, and TCN-only uses only temporal networks, ignoring static correlations. The method presented in this paper is a hybrid machine learning framework that models the nonlinear correlation of static parameters through random forests, captures dynamic changes through temporal convolutional networks, and achieves adaptive feature fusion by combining attention mechanisms. It also jointly optimizes the health score prediction task through gradient descent.

[0070] For example, when matching the comprehensive characteristics of rhizosphere soil with rhizosphere soil health labeling data, the system combines the various characteristics of each soil sample (such as pH value, temperature and humidity, nutrient concentration, etc.) with the corresponding health score labels to form a training dataset.

[0071] For example, a soil sample might have characteristics such as a pH of 6.5, a temperature and humidity of 70%, and high nutrient concentration. These data will be paired with the sample's health score (e.g., 0.85, indicating a healthy state) to form a complete training sample. The system evaluates the model's prediction accuracy by calculating a loss function. If the actual health score of a soil sample is 0.85, while the model predicts a score of 0.75, the system calculates the prediction error, reflecting the magnitude of the model's error. The smaller the loss function value, the more accurate the model's prediction. The model uses the backpropagation algorithm to calculate gradients, calculating the gradient of each parameter based on the loss function value.

[0072] For example, in a hybrid algorithm, if a certain feature (such as temperature and humidity) has a significant impact on the prediction result, the gradient calculation will yield a large gradient value for that feature parameter, indicating that this feature contributes significantly to the error and requires more substantial adjustment. The system performs parameter update operations based on the gradient vector. Assuming a learning rate of 0.01, the model will adjust the weight parameters according to this learning rate, gradually reducing the error. After multiple iterations, the model can eventually optimize the parameters and accurately predict soil health status.

[0073] In one specific embodiment, step S4 further includes:

[0074] Real-time collected rhizosphere sensor data is input into the rhizosphere soil health assessment algorithm for prediction and calculation, and the current rhizosphere soil status is quantitatively assessed to obtain a rhizosphere soil health score.

[0075] Based on the rhizosphere soil health score, a health status judgment threshold is set, and the health score is classified according to a preset range. When the health score is greater than or equal to 0.8, it is judged as a healthy state; when the health score is greater than or equal to 0.4 and less than 0.8, it is judged as a moderately healthy state; when the health score is less than 0.4, it is judged as an unhealthy state, thus obtaining the rhizosphere soil health status classification result.

[0076] The difference in rhizosphere soil health scores within a continuous time window is calculated, and the temporal trend of health scores is quantitatively analyzed to obtain the rate of change in health scores.

[0077] Based on the rate of change of health score, a trend judgment operation is performed to predict and analyze the development direction of rhizosphere soil health status, and the rhizosphere soil health status category is obtained.

[0078] Specifically, real-time collected rhizosphere sensor data is input into a rhizosphere soil health assessment algorithm for predictive calculation. The model analyzes various real-time soil data to obtain a current rhizosphere soil health score, which is used to quantitatively assess the soil's health status. Based on the obtained health score, the system sets a health status judgment threshold and classifies the health score according to a preset range. For example, when the health score is greater than or equal to 0.8, the system classifies the soil as healthy; when the health score is between 0.4 and 0.8, it is classified as moderately healthy; and when the health score is less than 0.4, it is classified as unhealthy. This classification result helps to quickly determine the overall health status of the soil, thus providing a basis for agricultural management decisions. Next, the system calculates the difference between the rhizosphere soil health scores within a continuous time window to quantify the temporal trend of the health score. This analysis helps to observe the changes in soil health scores over time, revealing fluctuations in soil health and deriving the rate of change of the health score. This rate of change is used to further measure the speed of improvement or deterioration of soil health status. Based on the calculated rate of change of the health score, the system performs a trend judgment operation to predict and analyze the development direction of the rhizosphere soil health status. Through this prediction, the system can predict the changing trends of soil health in a timely manner and provide early warnings of future soil health status, thus providing more precise control plans and decision support for agricultural management.

[0079] Taking an agricultural plantation as an example, real-time rhizosphere sensor data is input into a soil health assessment algorithm for predictive calculations. By monitoring parameters such as soil pH, temperature, humidity, and nutrient concentration through sensors, the algorithm calculates a soil health score in real time. Assuming the current score is 0.75, and based on a set health status threshold, a score of 0.75 falls within the moderately healthy range, the soil is therefore classified as moderately healthy. Next, the system calculates the difference in health scores over a consecutive week, finding a change rate of +0.05, indicating a slow improvement in soil health. Based on this rate of change, the system predicts that soil health may further improve in the coming days, adjusting the early warning system and preparing corresponding optimization measures, such as increasing fertilizer or adjusting irrigation plans. Through this analysis, agricultural managers can make timely decisions based on the real-time dynamics of soil health, effectively managing soil health and ensuring optimal crop growth conditions.

[0080] In one specific embodiment, step S5 further includes:

[0081] Based on the identification of key influencing factors of rhizosphere soil health status categories, the deviation of pH value, nutrient content and microbial activity were quantitatively analyzed to obtain rhizosphere factor deviation parameters.

[0082] The rhizosphere factor deviation parameter is input into the rhizosphere factor weight self-learning algorithm for weight calculation. The different rhizosphere environmental factors are dynamically weighted according to their degree of influence on soil health to obtain the rhizosphere factor weight coefficient.

[0083] The dosage of regulators was calculated based on the weighting coefficients of rhizosphere factors. The dosage of pH regulators was calculated by multiplying the pH deviation value by the rhizosphere soil volume. The dosage of fertilizers was calculated by multiplying the nutrient deviation value by the plant biomass. The dosage of biological agents was calculated by multiplying the microbial activity deviation value by the soil organic matter content. The regulator dosage parameters were obtained.

[0084] By correlating the regulator dosage parameters with the expected regulation effect, and integrating the implementation time and precautions of the regulation measures, an optimized rhizosphere soil scheme was obtained.

[0085] Specifically, based on the rhizosphere soil health status categories, the system identifies and analyzes the degree of deviation of key factors affecting soil health, such as pH deviation, nutrient content deviation, and microbial activity deviation. Through quantitative analysis, the system obtains deviation parameters for each factor, which reflect the gap between the current rhizosphere soil and the ideal health state. The system inputs these rhizosphere factor deviation parameters into a rhizosphere factor weight self-learning algorithm for weight calculation. The algorithm dynamically allocates weights according to the degree of influence of each factor on soil health. Through this process, the system can automatically adjust the weights of different factors according to the actual soil condition, maximizing the impact on health assessment and ensuring the effectiveness of soil improvement measures. Based on the calculated rhizosphere factor weight coefficients, the system performs regulator dosage calculation. For pH regulators, the system calculates the required dosage based on the product of pH deviation and soil volume; for fertilizer dosage, the system calculates the required amount based on the product of nutrient deviation and plant biomass; for bio-agents, the system calculates the required amount based on the product of microbial activity deviation and soil organic matter content. All calculation results are summarized into regulator dosage parameters to ensure that the dosage of each regulator matches the actual needs of the soil. The system then performs correlation analysis between the regulator dosage parameters and the expected regulatory effects to ensure that the regulatory measures achieve optimal results. Based on the analysis results, the system integrates the implementation time and precautions of the regulatory measures, and finally generates a rhizosphere soil optimization plan, providing an effective soil improvement strategy for agricultural management. Through this process, soil health management becomes more precise and intelligent, providing continuous support for healthy crop growth.

[0086] For example, when identifying the degree of deviation of key influencing factors based on the category of rhizosphere soil health status, the system first identifies that the current rhizosphere soil pH is 5.2, lower than the ideal range of 6.0-7.0, the nutrient content of nitrogen, phosphorus, and potassium is low, and microbial activity is also low. These deviations are quantified to obtain deviation parameters, for example, pH deviation is -0.8, nutrient deviation is -10%, and microbial activity deviation is -15%. Then, the system inputs these deviation parameters into the rhizosphere factor weight self-learning algorithm. According to the degree of influence of each factor on soil health, the algorithm assigns a dynamic weight to each factor. Assuming that the influence weight of pH on soil health is 0.5, the influence weight of nutrient content is 0.3, and the influence weight of microbial activity is 0.2. Next, the system calculates the required amount of regulator based on these weight coefficients. For example, for pH regulators, the system determines the required amount by calculating the product of the pH deviation value and the soil volume. If the soil volume is 100 cubic meters, the required regulator dosage is 0.8 (pH deviation) × 100 (volume), resulting in a required regulator dosage of 80 units. For fertilizers, assuming a nutrient deviation of -10% and a plant biomass of 1500 kg, the system will calculate a fertilizer application rate of 150 kg. For microbial inoculants, the system calculates the required dosage by multiplying the microbial activity deviation by the soil organic matter content; if the soil organic matter content is 3%, the calculated result is 45 units of inoculant. Ultimately, all this data will generate a regulator dosage plan, which the system will compare with the expected regulatory effect to ensure that these measures achieve the desired soil health goals.

[0087] Taking a greenhouse plantation as an example, the real-time monitored rhizosphere soil pH was 5.2, lower than the target range of 6.0-7.0, and the nutrient concentration was also lower than the ideal standard, indicating low microbial activity. The system, through quantitative analysis, determined that the pH deviation was -0.8, the nutrient concentration deviation was -12%, and the microbial activity deviation was -18%. These deviations were input into a rhizosphere factor weighting self-learning algorithm. The algorithm automatically calculated the weight based on the degree of influence of each factor on soil health, assuming a weight of 0.4 for pH, 0.35 for nutrient concentration, and 0.25 for microbial activity. Subsequently, the system calculated the required adjustment dosage based on these weights, combined with soil volume and other parameters. For example, with a pH deviation of -0.8 and a soil volume of 200 cubic meters, the system calculated a required pH adjustment dosage of 160 units; due to the low nutrient concentration, the system calculated a need for 150 units of fertilizer; and due to the large deviation in microbial activity, a need for 50 units of biological inoculant. After all the adjustment doses are integrated into the system, they are compared and analyzed with the expected adjustment effects to ensure that the adjustment doses meet the actual needs of the soil. Finally, a detailed soil optimization plan is generated, providing managers with specific fertilization and adjustment plans to ensure that the crop growth conditions in the greenhouse are optimally improved.

[0088] The above describes the machine learning-based rhizosphere soil health assessment method in the embodiments of this application. The following describes the machine learning-based rhizosphere soil health assessment system in the embodiments of this application. Please refer to [link / reference]. Figure 6 One embodiment of the machine learning-based rhizosphere soil health assessment system in this application includes:

[0089] The data acquisition module is used to collect data on pH, electrical conductivity, temperature and humidity, nutrient concentration and microbial activity in the rhizosphere region of the target plant as the raw dataset of the rhizosphere soil.

[0090] The correction module is used to correct outliers in the original rhizosphere soil dataset using the rhizosphere time-series median filtering method, calculate the rhizosphere nutrient comprehensive index and the rhizosphere microbial activity index, and obtain standardized rhizosphere soil data.

[0091] The training module is used to train the standardized rhizosphere soil data through a hybrid algorithm, and the attention mechanism integrates static features and temporal information to establish a rhizosphere soil health assessment algorithm.

[0092] The output module is used to output a rhizosphere soil health score through the rhizosphere soil health assessment algorithm, and to determine the rhizosphere soil health status category based on the rate of change of the health score.

[0093] The calculation module is used to calculate the dosage parameters of regulators based on the rhizosphere soil health status category and through a rhizosphere factor weight self-learning algorithm, and output the rhizosphere soil optimization scheme.

[0094] above Figure 6 The rhizosphere soil health assessment system based on machine learning in this embodiment of the invention will be described in detail from the perspective of modular functional entities. The rhizosphere soil health assessment device based on machine learning in this embodiment of the invention will be described in detail from the perspective of hardware processing.

[0095] Reference Figure 7 This invention also provides a machine learning-based rhizosphere soil health assessment device, which can be a server, and its internal structure can be as follows: Figure 7As shown, the machine learning-based rhizosphere soil health assessment device includes a processor, memory, display screen, input device, network interface, and database connected via a system bus. The processor, designed as a computer, provides computational and control capabilities. The memory of the machine learning-based rhizosphere soil health assessment device includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores the operating system, computer programs, and database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The database of the machine learning-based rhizosphere soil health assessment device stores the data corresponding to this embodiment. The network interface of the machine learning-based rhizosphere soil health assessment device is used for communication with external terminals via a network connection. When the computer program is executed by the processor, it implements the above-described method.

[0096] Those skilled in the art will understand that Figure 7 The structure shown is merely a block diagram of a portion of the structure related to the present invention and does not constitute a limitation on the machine learning-based rhizosphere soil health assessment device to which the present invention is applied.

[0097] The present invention also provides a computer-readable storage medium, which may be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium, wherein the computer-readable storage medium stores instructions that, when executed on a computer, cause the computer to perform the steps of the machine learning-based rhizosphere soil health assessment method.

[0098] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0099] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a machine learning-based rhizosphere soil health assessment device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0100] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do 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 machine learning-based method for assessing rhizosphere soil health, characterized in that, The method includes: Step S1: Collect data on pH, electrical conductivity, temperature, humidity, nutrient concentration, and microbial activity in the rhizosphere region of the target plant as the original rhizosphere soil dataset; Step S2: Correct outliers in the original rhizosphere soil dataset by using rhizosphere time-series median filtering, calculate the rhizosphere nutrient comprehensive index and rhizosphere microbial activity index, and obtain standardized rhizosphere soil data. Step S3 involves training the standardized rhizosphere soil data using a hybrid algorithm, fusing static features and temporal information through an attention mechanism, and establishing a rhizosphere soil health assessment algorithm, including: The standardized rhizosphere soil data is input into a random forest module for nonlinear feature relationship learning, modeling the complex correlations between rhizosphere soil parameters to obtain rhizosphere static feature weights. The standardized rhizosphere soil data is then input into a temporal convolutional neural network module for temporal change pattern recognition, extracting the time-series variation patterns of rhizosphere soil parameters to obtain rhizosphere temporal feature vectors. Based on the rhizosphere static feature weights and the rhizosphere temporal feature vectors, an attention mechanism is used for fusion calculation, weighting and merging the static features and temporal information according to dynamic weights to obtain comprehensive rhizosphere soil features. Finally, the comprehensive rhizosphere soil features are iteratively optimized and trained using gradient descent, with adaptive adjustments made to the hybrid algorithm parameters to obtain the rhizosphere soil health assessment algorithm. Step S4: Output the rhizosphere soil health score through the rhizosphere soil health assessment algorithm, and determine the rhizosphere soil health status category based on the rate of change of the health score. Step S5: Based on the rhizosphere soil health status category, calculate the regulator dosage parameters using a rhizosphere factor weight self-learning algorithm and output a rhizosphere soil optimization scheme. This includes: identifying the deviation degree of key influencing factors based on the rhizosphere soil health status category; quantitatively analyzing pH deviation, nutrient content deviation, and microbial activity deviation to obtain rhizosphere factor deviation parameters; inputting the rhizosphere factor deviation parameters into the rhizosphere factor weight self-learning algorithm for weight calculation; dynamically allocating weights to different rhizosphere environmental factors according to their impact on soil health to obtain rhizosphere factor weight coefficients; calculating regulator dosage based on the rhizosphere factor weight coefficients; calculating the pH regulator dosage by multiplying the pH deviation value by the rhizosphere soil volume; calculating the fertilizer dosage by multiplying the nutrient deviation value by the plant biomass; and calculating the bio-agent dosage by multiplying the microbial activity deviation value by the soil organic matter content to obtain regulator dosage parameters; and performing correlation analysis between the regulator dosage parameters and the expected regulation effect, integrating the implementation time and precautions of the regulation measures to obtain the rhizosphere soil optimization scheme.

2. The machine learning-based rhizosphere soil health assessment method according to claim 1, characterized in that, Step S1 further includes: A rhizosphere sensor network is deployed in the rhizosphere region of the target plant to simultaneously monitor multiple parameters of the rhizosphere microenvironment and obtain rhizosphere environment monitoring data. The rhizosphere environment monitoring data were collected at preset time intervals: pH data was collected every 30 minutes, electrical conductivity data was collected every 30 minutes, temperature and humidity data was collected every 60 minutes, nutrient concentration data was collected every 120 minutes, and microbial activity data was collected every 120 minutes to obtain rhizosphere soil time-series data. The rhizosphere soil time-series data is transmitted via a wireless communication module, and a timestamp, sensor identifier, and location information are added to each data record to obtain the original rhizosphere soil dataset. Based on the original rhizosphere soil dataset, data integrity was verified, and missing data was marked to obtain the original rhizosphere soil dataset.

3. The machine learning-based rhizosphere soil health assessment method according to claim 1, characterized in that, Step S2 further includes: The pH, electrical conductivity, temperature and humidity, nutrient concentration and microbial activity data in the original rhizosphere soil dataset are tested for range. Values ​​that exceed the preset range of the rhizosphere environment are marked as outliers, and outlier marked data are obtained. Based on the outlier-marked data, rhizosphere time-series median filtering is performed, and outliers are replaced and corrected by the median of adjacent time point data to obtain rhizosphere soil corrected data. The nitrogen, phosphorus, potassium, and organic acid concentrations in the rhizosphere soil correction data are calculated using a weighted summation formula to obtain the rhizosphere nutrient comprehensive index. The urease activity, phosphatase activity, and sucrase activity in the rhizosphere soil correction data were weighted according to the enzyme activity weighting coefficient to obtain the rhizosphere microbial activity index and rhizosphere soil standardized data.

4. The machine learning-based rhizosphere soil health assessment method according to claim 1, characterized in that, The process of iteratively optimizing and training the comprehensive characteristics of the rhizosphere soil using the gradient descent method, and adaptively adjusting the parameters of the hybrid algorithm, yields a rhizosphere soil health assessment algorithm, including: The comprehensive characteristics of the rhizosphere soil are matched with the rhizosphere soil health labeling data, and a corresponding health score label is assigned to each group of rhizosphere soil samples to obtain the rhizosphere soil training dataset. Based on the rhizosphere soil training dataset, the loss function value is calculated, and the error between the prediction result and the true label is quantitatively calculated to obtain the rhizosphere soil health prediction error. The rhizosphere soil health prediction error is calculated by backpropagation algorithm, and the gradient values ​​of the parameters of each layer in the hybrid algorithm are differentiated to obtain the rhizosphere soil parameter gradient vector. Based on the gradient vector of the rhizosphere soil parameters, a parameter update operation is performed, and the weight parameters of the hybrid algorithm are iteratively adjusted according to the learning rate to obtain the rhizosphere soil health assessment algorithm.

5. The machine learning-based rhizosphere soil health assessment method according to claim 1, characterized in that, Step S4 further includes: The rhizosphere sensor data collected in real time is input into the rhizosphere soil health assessment algorithm for prediction and calculation, and the current rhizosphere soil status is quantitatively assessed to obtain a rhizosphere soil health score. Based on the rhizosphere soil health score, a health status judgment threshold is set, and the health score is classified according to a preset range. When the health score is greater than or equal to 0.8, it is judged as a healthy state; when the health score is greater than or equal to 0.4 and less than 0.8, it is judged as a moderately healthy state; when the health score is less than 0.4, it is judged as an unhealthy state, thus obtaining the rhizosphere soil health status classification result. The difference in the rhizosphere soil health score within a continuous time window is calculated, and the temporal trend of the health score is quantitatively analyzed to obtain the rate of change of the health score. Based on the rate of change of the health score, a trend judgment operation is performed to predict and analyze the development direction of the rhizosphere soil health status, and the rhizosphere soil health status category is obtained.

6. A machine learning-based rhizosphere soil health assessment system, characterized in that, For implementing the machine learning-based rhizosphere soil health assessment method as described in any one of claims 1-5, the machine learning-based rhizosphere soil health assessment system comprises: The data acquisition module is used to collect data on pH, electrical conductivity, temperature and humidity, nutrient concentration and microbial activity in the rhizosphere region of the target plant as the raw dataset of the rhizosphere soil. The correction module is used to correct outliers in the original rhizosphere soil dataset using the rhizosphere time-series median filtering method, calculate the rhizosphere nutrient comprehensive index and the rhizosphere microbial activity index, and obtain standardized rhizosphere soil data. The training module is used to train the standardized rhizosphere soil data through a hybrid algorithm, and the attention mechanism integrates static features and temporal information to establish a rhizosphere soil health assessment algorithm. The output module is used to output a rhizosphere soil health score through the rhizosphere soil health assessment algorithm, and to determine the rhizosphere soil health status category based on the rate of change of the health score. The calculation module is used to calculate the dosage parameters of regulators based on the rhizosphere soil health status category and through a rhizosphere factor weight self-learning algorithm, and output the rhizosphere soil optimization scheme.

7. A rhizosphere soil health assessment device based on machine learning, characterized in that, It includes a memory and a processor, the memory storing a computer program that can run on the processor, the processor executing the computer program to implement the machine learning-based rhizosphere soil health assessment method according to any one of claims 1 to 5.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is run by a processor, it causes the processor to perform the machine learning-based rhizosphere soil health assessment method as described in any one of claims 1 to 5.

Images