Method and system for evaluating and analyzing grass and livestock balance in agricultural and pastoral compound area

Abstract

The invention discloses a method and a system for evaluating and analyzing grass and livestock balance in an agricultural and pastoral compound area, and particularly relates to the field of grass and livestock management. According to the method, target evaluation parameters of grassland resource and livestock demand evaluation are determined and monitored, data are verified, compared and analyzed in real time, the real-time performance of the data is evaluated, and classified management is performed, so that the accuracy and timeliness of the data are ensured. And through dynamic adjustment of management decisions, non-real-time data is further analyzed, and evaluation of grassland and livestock supply and demand balance is continuously optimized. According to the method, excessive grazing and degeneration of the grassland are prevented, sustainable utilization of grassland resources and harmonious development of animal husbandry production are achieved, and scientificity and effectiveness of ecological environment management and animal husbandry production are improved.

Description

Technical Field

[0001] The present invention relates to the field of grass-livestock management, and in particular to a grass-livestock balance assessment and analysis method and system in an agro-pastoral complex area. Background Art

[0002] The assessment and analysis of the grass-livestock balance in agro-pastoral complex areas refers to the assessment of the balance between grassland and livestock supply and demand through a comprehensive analysis of grassland resources (such as grass yield, grass species diversity) and livestock needs (such as food intake, livestock numbers) in mixed agricultural and pastoral areas. This analysis aims to ensure that grassland resources can sustainably meet the food needs of livestock and prevent overgrazing or degradation of grasslands, thereby achieving harmonious development of the ecological environment and livestock production. However, if the assessment of grassland and livestock needs relies on outdated data, it may not reflect the current actual situation, resulting in inaccurate data on which the grassland and livestock needs assessment relies. At the same time, if inaccurate data overestimates grassland production, managers may believe that the grassland can support more livestock, thereby increasing grazing intensity. This will lead to overgrazing of grasslands, reduction of grassland vegetation, destruction of soil structure, and ultimately grassland degradation and reduced productivity. Summary of the invention

[0003] The purpose of the present invention is to provide a method and system for evaluating and analyzing the grass-livestock balance in an agro-pastoral complex area, so as to overcome the deficiencies in the background technology.

[0004] In order to achieve the above object, the present invention provides the following technical solution: a method for evaluating and analyzing the balance between grass and livestock in an agricultural and pastoral complex area, comprising the following steps:

[0005] S1: Determine the target assessment parameters for judging grassland resources and livestock demand assessment, and monitor and obtain grassland resource status and livestock activity data within a fixed time period based on the target assessment parameters;

[0006] S2: Verify the grassland resource status and livestock activity data obtained in real time, compare and analyze them with historical data, and divide the grassland into data abnormality monitoring areas and data normal monitoring areas according to the analysis results;

[0007] S3: In the data anomaly monitoring area, the real-time performance of grassland resource status and livestock activity data is evaluated based on the deviation of the collection frequency of grassland resource status and livestock activity data and the fluctuation range of data trends;

[0008] S4; Based on the evaluation results, the real-time data of grassland resource status and livestock activity are divided into real-time data and non-real-time data, and corresponding management decisions are made;

[0009] S5: Further analyze non-real-time data on grassland resource status and livestock activities, provide feedback on the accuracy of target assessment parameters for grassland resource and livestock demand assessments, and dynamically adjust management decisions based on the feedback results.

[0010] Preferably, in S2, the grassland resource status and livestock activity data obtained in real time are verified and compared with the historical data for analysis, specifically:

[0011] The real-time grassland resource status and livestock activity data are compared and analyzed with historical data through the isolation forest algorithm, and the grassland is divided into data abnormality monitoring areas and data normal monitoring areas according to the analysis results;

[0012] Extract grassland resources and livestock activity data for the past T time period from the database;

[0013] Merge real-time and historical data to form a complete data set covering all key parameters;

[0014] The key parameters selected for analysis were: grassland production, grass species diversity, vegetation cover, soil moisture, livestock numbers, food intake, health status;

[0015] Use historical data as training data set to train the isolation forest model;

[0016] Through the Isolation Forest algorithm, the model is trained to identify normal and abnormal data patterns;

[0017] Use real-time data as the detection dataset and perform detection using the trained isolation forest model.

[0018] Preferably, the model assigns an anomaly score to each data point, compares the obtained anomaly score with the anomaly score threshold, and classifies it as normal data if the anomaly score is less than or equal to the anomaly score threshold; if the anomaly score is greater than the anomaly score threshold, classifies it as anomaly data;

[0019] According to the prediction results of the model, abnormal data points are marked, and the grassland is divided into data anomaly monitoring areas according to the geographical location of the abnormal data points;

[0020] According to the prediction results of the model, normal data points are marked, and the grassland is divided into normal data monitoring areas according to the geographical locations of the normal data points.

[0021] Preferably, in S3, the real-time performance of the grassland resource status and livestock activity data is evaluated according to the collection frequency deviation and data trend fluctuation range of the grassland resource status and livestock activity data, specifically:

[0022] In the data anomaly monitoring area, the collection frequency deviation index is generated according to the grassland resource status and the collection frequency deviation of livestock activity data. The method for obtaining the collection frequency deviation index is as follows:

[0023] Real-time acquisition of grassland resource status and livestock activity data within a time period W, with a collection frequency FK, and actual collection interval data X obtained based on the collection frequency s and the reference collection interval F, calculate the actual deviation D of the actual collection interval at time s relative to the reference collection interval s , the expression is: The forecast deviation value EWMA at time s is calculated by exponentially weighted moving average method s , the expression is: EWMA s =α*X s +(1-α)EWMA s-1 ; In the formula, α is the smoothing parameter, between 0 and 1, EWMA s is the predicted deviation value at time s, and the acquisition frequency deviation index is calculated. The expression is: ER=|D s -EWMA s |; Where ER is the acquisition frequency deviation index.

[0024] Preferably, in S3, in the data anomaly monitoring area, a trend fluctuation index is generated according to the data trend fluctuation amplitude of grassland resource conditions and livestock activity data, and the method for obtaining the trend fluctuation index is:

[0025] Determine grassland resource status data: grassland yield, vegetation coverage, soil moisture, grass species diversity, and livestock activity data: livestock number, food intake, and health status; construct several time series m(s) from the acquired data in chronological order, where s is time, select wavelet basis functions to decompose each time data, and perform continuous wavelet transform on it. The expression is: Among them, W(a,b) is the wavelet coefficient, which represents the transformation result on scale a and displacement b, a is the scale parameter, b is the displacement parameter, ψ * is the complex conjugate of the wavelet basis function; the wavelet coefficients are the detail coefficient Dj and the approximation coefficient Aj, the expression is: W(a,b)=Dj+Aj; where Dj=∑ k m(s)ψ j,k (s), k is the scale number; the standard deviation of the detail coefficient is calculated as the fluctuation amplitude of each scale, the fluctuation amplitudes of all scales are summed up, and the weighted average of each time series of grassland resource status and livestock activity data is calculated to generate the trend fluctuation index.

[0026] Preferably, the acquisition frequency deviation index and the trend fluctuation index are converted into a first eigenvector, and the first eigenvector is used as the input of the machine learning model. The machine learning model uses each group of first eigenvectors to predict the real-time value labels of grassland resource conditions and livestock activity data as the prediction target, and takes minimizing the sum of prediction errors of all real-time value labels of grassland resource conditions and livestock activity data as the training target. The machine learning model is trained until the sum of prediction errors converges and the model training is stopped. The real-time value of grassland resource conditions and livestock activity data is determined according to the model output results, wherein the machine learning model is a polynomial regression model.

[0027] Preferably, in S4, the real-time data of grassland resource status and livestock activity are divided into real-time data and non-real-time data;

[0028] The real-time value of the obtained grassland resource status and livestock activity data is compared with the real-time value reference threshold. If the real-time value of the grassland resource status and livestock activity data is greater than or equal to the real-time value reference threshold, it means that the real-time of the grassland resource status and livestock activity data is high, and it is classified as real-time data; if the real-time value of the grassland resource status and livestock activity data is less than the real-time value reference threshold, it means that the real-time of the grassland resource status and livestock activity data is low, and it is classified as non-real-time data.

[0029] Preferably, in S5, the non-real-time data of grassland resource status and livestock activities are further analyzed, and the accuracy of the target evaluation parameters of grassland resource and livestock demand evaluation is evaluated and fed back, specifically:

[0030] When grassland resource status and livestock activity data are classified as non-real-time data, that is, the real-time values ​​of grassland resource status and livestock activity data generated within a period of time are less than the real-time value reference threshold, the real-time values ​​less than the real-time value reference threshold within a period of time after the management decision is adjusted are collected, and the corresponding data set is established, the mean and standard deviation of the data set are calculated, and after analysis, the accuracy of the target evaluation parameters of grassland resource and livestock demand assessment is evaluated and fed back.

[0031] Preferably, if the mean of the real-time values ​​in the data set is greater than or equal to the reference threshold of the mean of the real-time values, and the standard deviation of the real-time values ​​is less than the reference threshold of the standard deviation of the real-time values, it means that the real-time performance of the data is high and the fluctuation is small. At this time, a data accuracy signal is generated to maintain the current management strategy;

[0032] If the mean of the real-time values ​​is greater than or equal to the reference threshold of the mean of the real-time values, and the standard deviation of the real-time values ​​is greater than or equal to the reference threshold of the standard deviation of the real-time values, it means that the real-time performance of the data is high but the fluctuation is large. In this case, an inaccurate data signal is generated to further refine the management strategy and reduce data fluctuation.

[0033] If the mean of the real-time values ​​is less than the reference threshold of the mean of the real-time values, and the standard deviation of the real-time values ​​is greater than or equal to the reference threshold of the standard deviation of the real-time values, it means that the real-time performance of the data is low and the fluctuation is large. At this time, an inaccurate data signal is generated, and management decisions are adjusted to improve the frequency and stability of data collection;

[0034] If the mean of the real-time values ​​is less than the reference threshold of the mean of the real-time values, and the standard deviation of the real-time values ​​is less than the reference threshold of the standard deviation of the real-time values, it means that the real-time performance of the data is low but the fluctuation is small. At this time, an inaccurate data signal is generated, and management decisions are adjusted to gradually improve the real-time performance of the data.

[0035] The present invention also provides a grass-livestock balance assessment and analysis system for an agro-pastoral complex area, comprising a data acquisition module, a comparative analysis module, a real-time assessment module, a decision-making module and a feedback adjustment module;

[0036] Data collection module: determine the target evaluation parameters for judging grassland resources and livestock demand assessment, and monitor and obtain grassland resource status and livestock activity data within a fixed time period based on the target evaluation parameters;

[0037] Comparative analysis module: Verify the grassland resource status and livestock activity data obtained in real time, compare and analyze them with historical data, and divide the grassland into data abnormality monitoring areas and data normal monitoring areas according to the analysis results;

[0038] Real-time evaluation module: In the data anomaly monitoring area, the real-time performance of grassland resource status and livestock activity data is evaluated based on the deviation of the collection frequency of grassland resource status and livestock activity data and the fluctuation range of data trend;

[0039] Decision-making module: Based on the evaluation results, the grassland resource status and livestock activity data are divided into real-time data and non-real-time data, and corresponding management decisions are made;

[0040] Feedback adjustment module: further analyzes non-real-time data on grassland resource status and livestock activities, provides feedback on the accuracy of target assessment parameters for grassland resource and livestock demand assessment, and dynamically adjusts management decisions based on feedback results.

[0041] In the above technical solution, the technical effects and advantages provided by the present invention are:

[0042] 1. The present invention effectively solves the problem of inaccurate evaluation caused by reliance on outdated data by determining evaluation parameters, real-time verification data, real-time evaluation data, classifying management data, and dynamically adjusting management decisions. By applying advanced algorithms such as the isolation forest algorithm and wavelet transform, the grassland resource status and livestock activity data are accurately analyzed, and the supply and demand balance between grassland and livestock can be dynamically and accurately evaluated, thereby preventing overgrazing and degradation of grassland and ensuring the sustainable use of grassland resources.

[0043] 2. The present invention not only improves the real-time and accuracy of data, ensures the scientificity and effectiveness of management decisions, but also continuously optimizes the assessment process of grassland resources and livestock needs by dynamically adjusting management strategies. Ultimately, the harmonious development of ecological environment and animal husbandry production is achieved, the overall level of grassland resource management and animal husbandry production is improved, and solid technical support is provided for the sustainable utilization of grassland resources and efficient management of animal husbandry. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0045] Figure 1 The present invention is a flow chart of the method.

[0046] Figure 2 It is a system module diagram of the present invention. DETAILED DESCRIPTION

[0047] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0048] Example 1, please refer to Figure 1 As shown, the method for evaluating and analyzing the balance between grass and livestock in an agricultural and pastoral complex area described in this embodiment includes the following steps:

[0049] S1: Determine the target assessment parameters for judging grassland resources and livestock demand assessment, and monitor and obtain grassland resource status and livestock activity data within a fixed time period based on the target assessment parameters;

[0050] S2: Verify the grassland resource status and livestock activity data obtained in real time, compare and analyze them with historical data, and divide the grassland into data abnormality monitoring areas and data normal monitoring areas according to the analysis results;

[0051] S3: In the data anomaly monitoring area, the real-time performance of grassland resource status and livestock activity data is evaluated based on the deviation of the collection frequency of grassland resource status and livestock activity data and the fluctuation range of data trends;

[0052] S4; Based on the evaluation results, the real-time data of grassland resource status and livestock activity are divided into real-time data and non-real-time data, and corresponding management decisions are made;

[0053] S5: Further analyze non-real-time data on grassland resource status and livestock activities, provide feedback on the accuracy of target assessment parameters for grassland resource and livestock demand assessments, and dynamically adjust management decisions based on the feedback results.

[0054] Among them, in S1, the target evaluation parameters for judging grassland resources and livestock demand evaluation are determined, and the grassland resource status and livestock activity data within a fixed time period are monitored and acquired according to the target evaluation parameters.

[0055] Grassland resource assessment parameters include: Grassland yield: the weight of hay per unit area. Grass species diversity: the number of grass species per unit area and their uniformity of distribution. Vegetation coverage: the percentage of grassland covered by vegetation and the greenness index (such as NDVI). Soil moisture: the water content in the soil. Climate conditions: precipitation, temperature, sunshine hours, etc. Soil nutrients: nitrogen, phosphorus, potassium content and organic matter content.

[0056] Parameters for assessing livestock needs include: Number of livestock: the number and growth rate of different types of livestock. Food intake: the daily intake of grass per livestock. Health status: body weight, health index (such as body condition score), disease incidence. Reproduction rate: the number of newborn livestock each year and the success rate of reproduction. Production performance: milk production, meat quality production, etc. Behavioral patterns: foraging time, frequency and location, activity monitoring.

[0057] Grassland resource monitoring includes: Remote sensing technology: Use satellite remote sensing or drone photography to monitor vegetation coverage and grassland yield. Ground sensors: Install soil moisture sensors and weather stations to monitor soil moisture and climate conditions. Field sampling: Collect grassland samples regularly at different locations to measure grass yield and grass species diversity.

[0058] Livestock demand monitoring includes: GPS collars: GPS collars are equipped for livestock to monitor their behavior patterns and range of activities in real time. Health monitoring equipment: Wireless body temperature sensors, health bracelets, etc. are used to monitor livestock's body temperature, heart rate and other health indicators. Intake records: The food intake of each livestock is recorded through an automatic feeding system. Weight monitoring: The weight of livestock is regularly weighed using electronic scales to record growth rate and health status. Reproduction records: Record the reproduction of each livestock, and count the number of newborns and the success rate of reproduction.

[0059] Set a fixed collection cycle (such as daily, weekly, monthly) to collect various parameters of grassland and livestock. Use automated equipment and sensors to collect data in real time and upload it to the database. Regular field surveys and manual measurements are carried out to ensure the accuracy and comprehensiveness of the data.

[0060] Establish a central database to store all collected data. Categorize and store data according to different parameters of grassland resources and livestock needs to facilitate subsequent analysis. Back up data regularly to prevent data loss.

[0061] S2: Verify the grassland resource status and livestock activity data obtained in real time, compare and analyze them with historical data, and divide the grassland into data abnormality monitoring areas and data normal monitoring areas based on the analysis results.

[0062] Real-time data verification includes: checking the integrity of real-time data to ensure that there is no missing or incomplete data. Ensuring the consistency of real-time data with historical data in terms of format and unit.

[0063] Select key parameters for comparison, such as grassland yield, grass species diversity, vegetation coverage, soil moisture, livestock population, food intake, health status, etc. Use statistical methods to calculate the mean, standard deviation, and rate of change of real-time data and historical data. Use time series analysis methods to analyze the trend differences between real-time data and historical data. Apply anomaly detection algorithms such as Isolation Forest and DBSCAN to identify abnormal patterns in the data.

[0064] The real-time grassland resource status and livestock activity data are compared and analyzed with historical data through the isolation forest algorithm, and the grassland is divided into data abnormality monitoring area and data normal monitoring area according to the analysis results. Specifically:

[0065] The grassland resources and livestock activity data for the past T time period are extracted from the database.

[0066] Merge real-time and historical data to form a complete data set covering all key parameters.

[0067] Select key parameters for analysis such as grassland production, grass species diversity, vegetation cover, soil moisture, livestock numbers, food intake, health status.

[0068] Standardize the data to ensure that the numerical ranges of different parameters are consistent and improve the performance of the algorithm.

[0069] Use historical data as training data set to train the isolation forest model.

[0070] Through the Isolation Forest algorithm, the model is trained to identify normal and abnormal data patterns.

[0071] Use real-time data as the detection dataset and perform detection using the trained isolation forest model.

[0072] The model assigns an anomaly score to each data point and compares the obtained anomaly score with the anomaly score threshold. If the anomaly score is less than or equal to the anomaly score threshold, it is classified as normal data. If the anomaly score is greater than the anomaly score threshold, it is classified as anomaly data.

[0073] According to the prediction results of the model, the abnormal data points are marked. According to the geographical location of the abnormal data points, the grassland is divided into data anomaly monitoring areas.

[0074] According to the prediction results of the model, normal data points are marked. According to the geographical location of normal data points, the grassland is divided into normal data monitoring areas.

[0075] Store the divided regional data in the database and establish corresponding indexes for easy query and management. Update regional data in real time and adjust regional divisions and labels according to new monitoring data. Use GIS system or visualization tools to display data abnormal monitoring areas and data normal monitoring areas on the map. Generate a report containing the abnormal area and normal area division results to provide decision support for managers.

[0076] S3: In the data anomaly monitoring area, the real-time performance of grassland resource status and livestock activity data is evaluated based on the deviation of their collection frequency and the fluctuation range of data trends.

[0077] In the data anomaly monitoring area, the collection frequency deviation index is generated according to the grassland resource status and the collection frequency deviation of livestock activity data. The method for obtaining the collection frequency deviation index is as follows:

[0078] Real-time acquisition of grassland resource status and livestock activity data within a time period W, with a collection frequency FK, and actual collection interval data X obtained based on the collection frequency s and the reference collection interval F, calculate the actual deviation D of the actual collection interval at time s relative to the reference collection intervals , the expression is: The forecast deviation value EWMA at time s is calculated by exponentially weighted moving average method s , the expression is: EWMA s =α*X s +(1-α)EWMA s-1 ; In the formula, α is the smoothing parameter, between 0 and 1, EWMA s is the predicted deviation value at time s, and the acquisition frequency deviation index is calculated. The expression is: ER=|D s -EWMA s |; Where ER is the acquisition frequency deviation index.

[0079] The larger the collection frequency deviation index, the greater the deviation between the collection frequency of grassland resource status and livestock activity data and the expected standard collection frequency. This means that the real-time performance of the data is worse, that is, the data collection is not frequent or timely enough to accurately reflect the current actual situation, which may lead to inaccurate management and decision-making basis. To ensure the timeliness and accuracy of the data, the collection frequency deviation index should be kept at a low level as much as possible.

[0080] In the data anomaly monitoring area, the trend fluctuation index is generated according to the data trend fluctuation amplitude of grassland resource conditions and livestock activity data. The method for obtaining the trend fluctuation index is as follows:

[0081] Determine grassland resource status data: grassland yield, vegetation coverage, soil moisture, grass species diversity, and livestock activity data: livestock number, food intake, and health status; construct several time series m(s) from the acquired data in chronological order, where s is time, select wavelet basis functions to decompose each time data, and perform continuous wavelet transform on it. The expression is: Among them, W(a,b) is the wavelet coefficient, which represents the transformation result on scale a and displacement b. a is the scale parameter, which controls the width of the wavelet function, b is the displacement parameter, which controls the position of the wavelet function, and ψ * is the complex conjugate of the wavelet basis function; the wavelet coefficients are the detail coefficient Dj and the approximation coefficient Aj, the expression is: W(a,b)=Dj+Aj; where Dj=∑ k m(s)ψ j,k (s), k is the scale label; the approximation coefficient is used to capture the main trend of the data, the detail coefficient is used to capture the local fluctuation of the data, the standard deviation of the detail coefficient is calculated as the fluctuation amplitude of each scale, the fluctuation amplitude of all scales is summed up, and the weighted average calculation of each time series of grassland resource status and livestock activity data is performed to generate the trend fluctuation index.

[0082] The larger the trend fluctuation index, the greater the fluctuation of grassland resource status and livestock activity data, which means the data is less stable and less real-time. This means that the data has large changes between different time points and cannot accurately and continuously reflect the actual situation, which may lead to a decrease in the accuracy of management and decision-making basis. To ensure the reliability and real-time nature of the data, the trend fluctuation index should be kept at a low level as much as possible.

[0083] The acquisition frequency deviation index and the trend fluctuation index are converted into the first eigenvector, and the first eigenvector is used as the input of the machine learning model. The machine learning model predicts the real-time value labels of grassland resource conditions and livestock activity data for each group of first eigenvectors as the prediction target, and minimizes the sum of prediction errors of all real-time value labels of grassland resource conditions and livestock activity data as the training target. The machine learning model is trained until the sum of prediction errors converges and the model training is stopped. The real-time value of grassland resource conditions and livestock activity data is determined according to the model output results, wherein the machine learning model is a polynomial regression model.

[0084] The method for obtaining the real-time value of grassland resource status and livestock activity data is as follows: from the first eigenvector training data of the trained machine learning model, the corresponding function expression is obtained: LP=f(ER, EL); wherein f is the output function of the model, ER is the acquisition frequency deviation index, EL is the trend fluctuation index, and LP is the real-time value of grassland resource status and livestock activity data.

[0085] S4; Based on the evaluation results, the grassland resource status and livestock activity data are divided into real-time data and non-real-time data, and corresponding management decisions are made.

[0086] The real-time value of the obtained grassland resource status and livestock activity data is compared with the real-time value reference threshold. If the real-time value of the grassland resource status and livestock activity data is greater than or equal to the real-time value reference threshold, it means that the real-time of the grassland resource status and livestock activity data is high, the fluctuation range is small, the data is stable, and it can accurately reflect the actual situation, and it is classified as real-time data; if the real-time value of the grassland resource status and livestock activity data is less than the real-time value reference threshold, it means that the real-time of the grassland resource status and livestock activity data is low, the fluctuation range is large, the data is unstable, and it cannot continuously and accurately reflect the actual situation, and it is classified as non-real-time data.

[0087] Management decisions for real-time data include: Continue existing monitoring: Maintain the current frequency and method of data collection to ensure the continuity and real-time nature of the data. Data application: Use real-time data for real-time analysis and decision-making, such as regulating grassland grazing intensity and adjusting livestock feeding strategies. Feedback mechanism: Establish a rapid feedback mechanism to handle data anomalies in a timely manner and ensure the effectiveness of management measures.

[0088] Management decisions for non-real-time data include: Increase the frequency of data collection: Increase the frequency of data collection, reduce the interval between data collection, and ensure the real-time and continuity of data. Improve the collection method: Use more advanced monitoring technology and equipment, such as remote sensing technology, drone monitoring, IoT sensors, etc., to improve the accuracy and timeliness of data collection. Data completion: For missing or abnormal data, complete and correct data to ensure the integrity and reliability of the data. Regular evaluation: Regularly evaluate the data collection and processing process, continuously improve, and ensure data quality.

[0089] S5: Further analyze non-real-time data on grassland resource status and livestock activities, provide feedback on the accuracy of target assessment parameters for grassland resource and livestock demand assessments, and dynamically adjust management decisions based on the feedback results.

[0090] When grassland resource status and livestock activity data are classified as non-real-time data, that is, the real-time values ​​of grassland resource status and livestock activity data generated within a period of time are less than the real-time value reference threshold, the real-time values ​​less than the real-time value reference threshold within a period of time after the management decision is adjusted are collected, and the corresponding data set is established, the mean and standard deviation of the data set are calculated, and after analysis, the accuracy of the target evaluation parameters of grassland resource and livestock demand assessment is evaluated and fed back.

[0091] If the mean of the real-time values ​​in the data set is greater than or equal to the reference threshold of the mean of the real-time values, and the standard deviation of the real-time values ​​is less than the reference threshold of the standard deviation of the real-time values, it means that the real-time performance of the data is high and the fluctuation is small. At this time, the data is accurate. The signal is generated, the management decision adjustment is effective, and the current management strategy can be maintained;

[0092] If the mean of the real-time values ​​is greater than or equal to the reference threshold of the mean of the real-time values, and the standard deviation of the real-time values ​​is greater than or equal to the reference threshold of the standard deviation of the real-time values, it means that the real-time performance of the data is high but the fluctuation is large. In this case, an inaccurate data signal is generated, and the management strategy needs to be further refined to reduce data fluctuations.

[0093] If the mean of the real-time values ​​is less than the reference threshold of the mean of the real-time values, and the standard deviation of the real-time values ​​is greater than or equal to the reference threshold of the standard deviation of the real-time values, it means that the real-time performance of the data is low and fluctuates greatly. At this time, an inaccurate data signal is generated, and management decisions need to be significantly adjusted to improve the frequency and stability of data collection;

[0094] If the mean of the real-time values ​​is less than the reference threshold of the mean of the real-time values, and the standard deviation of the real-time values ​​is less than the reference threshold of the standard deviation of the real-time values, it means that the real-time performance of the data is low but the fluctuation is small. At this time, an inaccurate data signal is generated, and management decisions need to be adjusted appropriately to gradually improve the real-time performance of the data.

[0095] Dynamically adjust management decisions on grassland resources and livestock needs based on the results of feedback analysis: Increase collection frequency: Increase the frequency of data collection and shorten the collection interval. Improve collection technology: Use more advanced technology and equipment for data collection, such as drones, IoT sensors, etc. Data correction: Correct historical data to ensure data accuracy and consistency. Strengthen monitoring: Focus on monitoring non-real-time data areas to identify and solve problems in a timely manner.

[0096] Implement corresponding management decision adjustments based on feedback results. Continue to monitor adjusted data, regularly evaluate the real-time nature of data, and further adjust management decisions based on evaluation results.

[0097] In this embodiment, by determining the target evaluation parameters for judging grassland resources and livestock demand assessment, the grassland resource status and livestock activity data within a fixed time period are monitored and obtained according to the target evaluation parameters; the grassland resource status and livestock activity data obtained in real time are verified, compared and analyzed with historical data, and the grassland is divided into a data abnormality monitoring area and a data normal monitoring area according to the analysis results; in the data abnormality monitoring area, the real-time performance of the grassland resource status and livestock activity data is evaluated according to the collection frequency deviation of the grassland resource status and livestock activity data and the data trend fluctuation amplitude; according to the evaluation results, the real-time performance of the grassland resource status and livestock activity data is divided into real-time data and non-real-time data, and corresponding management decisions are made; the non-real-time data of the grassland resource status and livestock activity are further analyzed, the accuracy of the target evaluation parameters of the grassland resource and livestock demand assessment is evaluated and fed back, and the management decision is dynamically adjusted according to the feedback results. Not only the accuracy and real-time performance of the data are improved, the scientificity and effectiveness of the management decision are ensured, but also the data abnormality problem can be discovered and solved in time, the grassland resource and livestock demand assessment process is optimized, and the sustainable utilization of grassland resources and the efficient management of livestock production are realized.

[0098] Example 2, please refer to Figure 2 As shown, the grass-livestock balance assessment and analysis system for the agricultural-pastoral complex area described in this embodiment includes a data acquisition module, a comparison and analysis module, a real-time assessment module, a decision-making module and a feedback adjustment module;

[0099] Data collection module: determine the target evaluation parameters for judging grassland resources and livestock demand assessment, and monitor and obtain grassland resource status and livestock activity data within a fixed time period based on the target evaluation parameters;

[0100] Comparative analysis module: Verify the grassland resource status and livestock activity data obtained in real time, compare and analyze them with historical data, and divide the grassland into data abnormality monitoring areas and data normal monitoring areas according to the analysis results;

[0101] Real-time evaluation module: In the data anomaly monitoring area, the real-time performance of grassland resource status and livestock activity data is evaluated based on the deviation of the collection frequency of grassland resource status and livestock activity data and the fluctuation range of data trend;

[0102] Decision-making module: Based on the evaluation results, the grassland resource status and livestock activity data are divided into real-time data and non-real-time data, and corresponding management decisions are made;

[0103] Feedback adjustment module: further analyzes non-real-time data on grassland resource status and livestock activities, provides feedback on the accuracy of target assessment parameters for grassland resource and livestock demand assessment, and dynamically adjusts management decisions based on feedback results.

[0104] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0105] The above embodiments can be implemented in whole or in part by software, hardware, firmware or any other combination. When implemented by 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 process or function described in the embodiment of the present application is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from one website site, computer, server or data center to another website site, computer, server or data center by wired or wireless (e.g., infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that contains one or more available media sets. The available medium can be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state hard disk.

[0106] It should be understood that the term "and / or" in this article is only a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. A and B can be singular or plural. In addition, the character " / " in this article generally indicates that the associated objects before and after are in an "or" relationship, but it may also indicate an "and / or" relationship. Please refer to the context for specific understanding.

[0107] It should be understood that in the various embodiments of the present application, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0108] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0109] The above description is only a specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application.

Claims

1. A method for evaluating and analyzing the balance between grass and livestock in agro-pastoral complex areas. Features: The steps include: S1: Determine the target assessment parameters for judging grassland resources and livestock demand assessment, and monitor and obtain grassland resource status and livestock activity data within a fixed time period based on the target assessment parameters; S2: Verify the grassland resource status and livestock activity data obtained in real time, compare and analyze them with historical data, and divide the grassland into data abnormality monitoring areas and data normal monitoring areas according to the analysis results; S3: In the data anomaly monitoring area, the real-time performance of grassland resource status and livestock activity data is evaluated based on the deviation of the collection frequency of grassland resource status and livestock activity data and the fluctuation range of data trends; S4; Based on the evaluation results, the real-time data of grassland resource status and livestock activity are divided into real-time data and non-real-time data, and corresponding management decisions are made; S5: Further analyze non-real-time data on grassland resource status and livestock activities, provide feedback on the accuracy of target assessment parameters for grassland resource and livestock demand assessments, and dynamically adjust management decisions based on the feedback results.

2. According to the method for evaluating and analyzing the balance between grass and livestock in an agro-pastoral complex area according to claim 1, Features: In S2, the grassland resource status and livestock activity data obtained in real time are verified and compared with the historical data for analysis, specifically: The real-time grassland resource status and livestock activity data are compared and analyzed with historical data through the isolation forest algorithm, and the grassland is divided into data abnormality monitoring areas and data normal monitoring areas according to the analysis results; Extract grassland resources and livestock activity data for the past T time period from the database; Merge real-time and historical data to form a complete data set covering all key parameters; The key parameters selected for analysis were: grassland production, grass species diversity, vegetation cover, soil moisture, livestock numbers, food intake, health status; Use historical data as training data set to train the isolation forest model; Through the Isolation Forest algorithm, the model is trained to identify normal and abnormal data patterns; Use real-time data as the detection dataset and perform detection using the trained isolation forest model.

3. The method for evaluating and analyzing the balance between grass and livestock in an agro-pastoral complex area according to claim 2, Features: The model assigns an anomaly score to each data point and compares the obtained anomaly score with the anomaly score threshold. If the anomaly score is less than or equal to the anomaly score threshold, it is classified as normal data; if the anomaly score is greater than the anomaly score threshold, it is classified as anomaly data. According to the prediction results of the model, abnormal data points are marked, and the grassland is divided into data anomaly monitoring areas according to the geographical location of the abnormal data points; According to the prediction results of the model, normal data points are marked, and the grassland is divided into normal data monitoring areas according to the geographical locations of the normal data points.

4. The method for evaluating and analyzing the balance between grass and livestock in an agro-pastoral complex area according to claim 1, Features: In S3, the real-time performance of grassland resource status and livestock activity data is evaluated based on the deviation of the collection frequency of grassland resource status and livestock activity data and the fluctuation range of data trend, specifically: In the data anomaly monitoring area, the collection frequency deviation index is generated according to the grassland resource status and the collection frequency deviation of livestock activity data. The method for obtaining the collection frequency deviation index is as follows: Real-time acquisition of grassland resource status and livestock activity data within a time period W, with a collection frequency FK, and actual collection interval data X obtained based on the collection frequency s and the reference collection interval F, calculate the actual deviation D of the actual collection interval at time s relative to the reference collection interval s , the expression is: The forecast deviation value EWMA at time s is calculated by exponentially weighted moving average method s , the expression is: EWMA s =α*X s +(1-α)EWMA s-1 ; In the formula, α is the smoothing parameter, between 0 and 1, EWMA s is the predicted deviation value at time s, and the acquisition frequency deviation index is calculated. The expression is: ER=|D s -EWMA s |; Where ER is the acquisition frequency deviation index.

5. The method for evaluating and analyzing the balance between grass and livestock in an agro-pastoral complex area according to claim 4, Features: In S3, within the data anomaly monitoring area, a trend fluctuation index is generated according to the data trend fluctuation amplitude of grassland resource conditions and livestock activity data. The method for obtaining the trend fluctuation index is: Determine grassland resource status data: grassland yield, vegetation coverage, soil moisture, grass species diversity, and livestock activity data: livestock number, food intake, and health status; construct several time series m(s) from the acquired data in chronological order, where s is time, select wavelet basis functions to decompose each time data, and perform continuous wavelet transform on it. The expression is: Among them, W(a,b) is the wavelet coefficient, which represents the transformation result on scale a and displacement b, a is the scale parameter, b is the displacement parameter, ψ * is the complex conjugate of the wavelet basis function; the wavelet coefficients are the detail coefficient Dj and the approximation coefficient Aj, the expression is: W(a,b)=Dj+Aj; where Dj=∑ k m(s)ψ j,k (s), k is the scale number; the standard deviation of the detail coefficient is calculated as the fluctuation amplitude of each scale, the fluctuation amplitudes of all scales are summed up, and the weighted average of each time series of grassland resource status and livestock activity data is calculated to generate the trend fluctuation index.

6. A method for evaluating and analyzing the balance between grass and livestock in an agro-pastoral complex area according to claim 5, Features: The acquisition frequency deviation index and the trend fluctuation index are converted into the first eigenvector, and the first eigenvector is used as the input of the machine learning model. The machine learning model predicts the real-time value labels of grassland resource conditions and livestock activity data for each group of first eigenvectors as the prediction target, and minimizes the sum of prediction errors of all real-time value labels of grassland resource conditions and livestock activity data as the training target. The machine learning model is trained until the sum of prediction errors converges and the model training is stopped. The real-time value of grassland resource conditions and livestock activity data is determined according to the model output results, wherein the machine learning model is a polynomial regression model.

7. The method for evaluating and analyzing the balance between grass and livestock in an agro-pastoral complex area according to claim 6, Features: In S4, the real-time data of grassland resource status and livestock activities are divided into real-time data and non-real-time data; The real-time value of the obtained grassland resource status and livestock activity data is compared with the real-time value reference threshold. If the real-time value of the grassland resource status and livestock activity data is greater than or equal to the real-time value reference threshold, it means that the real-time of the grassland resource status and livestock activity data is high, and it is classified as real-time data; if the real-time value of the grassland resource status and livestock activity data is less than the real-time value reference threshold, it means that the real-time of the grassland resource status and livestock activity data is low, and it is classified as non-real-time data.

8. The method for evaluating and analyzing the balance between grass and livestock in an agro-pastoral complex area according to claim 1, Features: In S5, the non-real-time data of grassland resource status and livestock activities are further analyzed to provide feedback on the accuracy of the target assessment parameters for grassland resource and livestock demand assessment, specifically: When grassland resource status and livestock activity data are classified as non-real-time data, that is, the real-time values ​​of grassland resource status and livestock activity data generated within a period of time are less than the real-time value reference threshold, the real-time values ​​less than the real-time value reference threshold within a period of time after the management decision is adjusted are collected, and the corresponding data set is established, the mean and standard deviation of the data set are calculated, and after analysis, the accuracy of the target evaluation parameters of grassland resource and livestock demand assessment is evaluated and fed back.

9. The method for evaluating and analyzing the balance between grass and livestock in an agro-pastoral complex area according to claim 8, Features: If the mean of the real-time values ​​in the data set is greater than or equal to the reference threshold of the mean of the real-time values, and the standard deviation of the real-time values ​​is less than the reference threshold of the standard deviation of the real-time values, it means that the real-time performance of the data is high and the fluctuation is small. At this time, a data accuracy signal is generated and the current management strategy is maintained; If the mean of the real-time values ​​is greater than or equal to the reference threshold of the mean of the real-time values, and the standard deviation of the real-time values ​​is greater than or equal to the reference threshold of the standard deviation of the real-time values, it means that the real-time performance of the data is high but the fluctuation is large. In this case, an inaccurate data signal is generated to further refine the management strategy and reduce data fluctuation. If the mean of the real-time values ​​is less than the reference threshold of the mean of the real-time values, and the standard deviation of the real-time values ​​is greater than or equal to the reference threshold of the standard deviation of the real-time values, it means that the real-time performance of the data is low and the fluctuation is large. At this time, an inaccurate data signal is generated, and management decisions are adjusted to improve the frequency and stability of data collection; If the mean of the real-time values ​​is less than the reference threshold of the mean of the real-time values, and the standard deviation of the real-time values ​​is less than the reference threshold of the standard deviation of the real-time values, it means that the real-time performance of the data is low but the fluctuation is small. At this time, an inaccurate data signal is generated, and management decisions are adjusted to gradually improve the real-time performance of the data.

10. A system for evaluating and analyzing the balance between grass and livestock in an agro-pastoral complex area, used to implement the method for evaluating and analyzing the balance between grass and livestock in an agro-pastoral complex area as claimed in any one of claims 1 to 9. Features: It includes data collection module, comparative analysis module, real-time evaluation module, decision-making module and feedback adjustment module; Data collection module: determine the target evaluation parameters for judging grassland resources and livestock demand assessment, and monitor and obtain grassland resource status and livestock activity data within a fixed time period based on the target evaluation parameters; Comparative analysis module: Verify the grassland resource status and livestock activity data obtained in real time, compare and analyze them with historical data, and divide the grassland into data abnormality monitoring areas and data normal monitoring areas according to the analysis results; Real-time evaluation module: In the data anomaly monitoring area, the real-time performance of grassland resource status and livestock activity data is evaluated based on the deviation of the collection frequency of grassland resource status and livestock activity data and the fluctuation range of data trend; Decision-making module: Based on the evaluation results, the grassland resource status and livestock activity data are divided into real-time data and non-real-time data, and corresponding management decisions are made; Feedback adjustment module: further analyzes non-real-time data on grassland resource status and livestock activities, provides feedback on the accuracy of target assessment parameters for grassland resource and livestock demand assessment, and dynamically adjusts management decisions based on feedback results.

Images