Agricultural economic heat map real-time updating method based on characteristic agricultural product circulation

Abstract

This invention relates to the field of agricultural economic data processing technology, and discloses a method for real-time updating of agricultural economic heat maps based on the circulation of specialty agricultural products. The method includes: collecting real-time agricultural product circulation data across the entire circulation chain of specialty agricultural products, performing preprocessing and anomaly detection to obtain a standardized circulation dataset; determining the value of each core evaluation indicator in the agricultural economic heat map, thereby determining the comprehensive economic heat value; determining the data change intensity of the entire circulation chain of specialty agricultural products and the update frequency of the agricultural economic heat map based on the comprehensive economic heat value; updating the agricultural economic heat map based on the update frequency, and converting the comprehensive economic heat value into color levels and mapping it to the map. This invention solves the problems of information asymmetry, data fragmentation, and low utilization rate in the circulation process of specialty agricultural products, improves the data foundation quality and evaluation dimension completeness of the agricultural economic heat map, and enhances the real-time performance, accuracy, and practicality of agricultural economic operation monitoring.

Description

Technical Field

[0001] This invention relates to the field of agricultural economic data processing technology, and in particular to a method for real-time updating of agricultural economic heat maps based on the circulation of specialty agricultural products. Background Technology

[0002] With the continuous advancement of agricultural modernization in my country, the agricultural product circulation system is becoming increasingly complex and networked, placing higher demands on the real-time perception and dynamic regulation of the agricultural economic operation. Specialty agricultural products such as honey, edible fungi, citrus fruits, traditional Chinese medicine, and tea, due to their strong regional characteristics, short shelf life, and high value sensitivity, involve multiple stages in their circulation process, including harvesting, storage, transportation, and trading. Storage and cold chain transportation infrastructure are relatively weak, and the requirements for preservation and anti-corrosion are higher. Furthermore, the data sources at each stage are scattered, heterogeneous in format, and updated at inconsistent frequencies, making it difficult to form a unified and accurate depiction of the overall circulation situation. Therefore, specialty agricultural products have higher demands for circulation efficiency, and the problem of information asymmetry is more prominent, seriously hindering the high-quality development of specialty agriculture.

[0003] While some existing agricultural information systems can collect localized data or display static heat maps of specialty agriculture, they generally suffer from the following shortcomings: First, data collection relies on a single channel and lacks the ability to integrate multi-source data covering the entire chain from production to logistics to sales. Second, data processing methods are crude, with insufficient identification of outliers, affecting the accuracy of subsequent analysis. Third, the heat map update mechanism is rigid, typically using fixed time intervals for refreshing, failing to dynamically adjust the update frequency according to actual economic activity, resulting in resource waste or information lag. Fourth, heat maps only reflect single-dimensional indicators (such as transaction volume) and lack a comprehensive evaluation of multi-dimensional economic characteristics such as circulation scale, efficiency, benefits, and losses, making it difficult to truly reflect the level of regional agricultural economic activity.

[0004] Therefore, there is an urgent need for an agricultural economic monitoring method that can integrate multi-source heterogeneous data, intelligently identify anomalies, dynamically assess economic activity, and adaptively update heat maps accordingly, in order to improve the timeliness, scientific rigor, and visualization of the supervision of the circulation of specialty agricultural products. Summary of the Invention

[0005] The purpose of this invention is to provide a real-time updating method for agricultural economic heat maps based on the circulation of specialty agricultural products, aiming to solve the problems of data fragmentation, unreliable data quality, rigid heat map updating mechanism, and single evaluation dimension in traditional solutions.

[0006] This invention provides a method for real-time updating of agricultural economic heat maps based on the circulation of specialty agricultural products, including:

[0007] A multi-source data acquisition terminal network is constructed to collect real-time agricultural product circulation data across the entire circulation chain of specialty agricultural products. The real-time agricultural product circulation data includes basic circulation data, IoT sensing data, logistics tracking data, market transaction data, and environmental auxiliary data.

[0008] The real-time agricultural product circulation data is preprocessed and anomaly detected to obtain a standardized circulation dataset;

[0009] The core evaluation index system of the agricultural economic heat map is determined. The value of each core evaluation index is determined based on the standardized circulation dataset. The comprehensive economic heat value is determined based on the core evaluation index values. The core evaluation indicators include circulation scale indicators, circulation efficiency indicators, circulation benefit indicators and circulation loss indicators.

[0010] The intensity of data change in the entire circulation chain of specialty agricultural products is determined based on the comprehensive economic heat value, and the update frequency of the agricultural economic heat map is determined based on the intensity of data change. The intensity of data change is determined based on the magnitude, rate, and continuity of change of the comprehensive economic heat value.

[0011] An agricultural economic heat map of specialty agricultural products is constructed, the agricultural economic heat map is updated based on the update frequency, and the comprehensive economic heat value is converted into a color level and mapped onto the agricultural economic heat map.

[0012] Preferably, a multi-source data acquisition terminal network is constructed, including: a harvest monitoring terminal at the place of origin, used to obtain the harvest quantity, harvest time and category of specialty agricultural products;

[0013] Internet of Things (IoT) sensor networks are used to acquire data on temperature, humidity, and storage environment in the production areas of specialty agricultural products;

[0014] Logistics tracking terminals are used to obtain information on the transportation routes, transportation time, and cargo status of specialty agricultural products.

[0015] The sales market trading terminal is used to obtain the transaction amount, transaction volume, and transaction price of specialty agricultural products;

[0016] Meteorological and transportation data interfaces are used to acquire meteorological and transportation data.

[0017] Preferably, the real-time agricultural product circulation data is preprocessed and anomaly detected to obtain a standardized circulation dataset, specifically as follows:

[0018] The preprocessing includes standardization, spatiotemporal alignment, cleaning, normalization, and temporal smoothing.

[0019] An anomaly detection model is constructed based on the improved isolated forest algorithm to perform a first anomaly detection on the real-time agricultural product circulation data; a second anomaly detection is performed on the real-time agricultural product circulation data based on the environmental auxiliary data.

[0020] One of the anomaly detection steps includes:

[0021] A time-series data matrix is ​​constructed from the real-time agricultural product circulation data in chronological order. Each row of the time-series data matrix represents multi-dimensional data of a collection period, and each column represents a type of circulation data.

[0022] Based on the timeliness characteristics of real-time agricultural product circulation data, time-series weights are assigned to data at different time points, and the impact of recent data on anomaly detection is determined through weighted processing.

[0023] An improved isolated forest model is constructed. Based on the improved isolated forest model, all samples are traversed, and the path length of each sample in each decision tree is calculated. The average of the path lengths of all decision trees is taken as the mean path length of the sample. The anomaly score of the sample is calculated based on the mean path length.

[0024] A dynamic anomaly threshold is set, the sample score is compared with the dynamic anomaly threshold, anomaly samples are identified based on the comparison result, and the anomaly samples are removed.

[0025] Preferably, a secondary anomaly detection is performed on the real-time agricultural product circulation data based on the environmental auxiliary data, specifically as follows:

[0026] The environmental auxiliary data includes meteorological data, traffic data, and market activity data;

[0027] An environment-circulation correlation database is constructed based on the aforementioned environmental auxiliary data. The environment-circulation correlation database includes several rules and the fluctuation range of circulation data for each rule for the characteristic agricultural products.

[0028] Based on the real-time agricultural product circulation data of the current collection period, environmental auxiliary data of the corresponding time and region are extracted simultaneously. The rules corresponding to the current environmental auxiliary data and the fluctuation range of circulation data of specialty agricultural products are determined by retrieval and matching.

[0029] The fluctuation range of real-time agricultural product circulation data in the current collection period is compared with that of the retrieved circulation data. If real-time agricultural product circulation data exceeds the fluctuation range of circulation data, the corresponding data is determined to be abnormal and removed.

[0030] Preferably, the comprehensive economic heat index value is determined based on the core evaluation index values, specifically as follows:

[0031] The subjective weights of each core evaluation indicator are determined by combining the analytic hierarchy process with expert experience. Then, the objective weights of each core evaluation indicator are calculated based on the information entropy of the standardized circulation dataset using the entropy weight method. Finally, the comprehensive weight of each core evaluation indicator is obtained by weighted summation.

[0032] The comprehensive economic heat index value is determined based on the aforementioned comprehensive weights and core evaluation index values.

[0033] The formula for calculating the comprehensive economic heat index is as follows:

[0034] ;

[0035] In the above formula, EI represents the comprehensive economic heat index value, wi represents the comprehensive weight of the i-th core evaluation indicator, and xi represents the value of the i-th core evaluation indicator.

[0036] Preferably, the intensity of data change across the entire agricultural product distribution chain is determined based on the comprehensive economic heat index value, specifically as follows:

[0037] The magnitude, rate, and continuity of change of the comprehensive economic heat index value are determined based on the comprehensive economic heat index value.

[0038] The intensity of data change across the entire agricultural product circulation chain is determined based on the magnitude, rate, and continuity of the changes in the comprehensive economic heat index value.

[0039] The intensity of the data change is determined according to the following formula:

[0040] ;

[0041] In the above formula, CI represents the intensity of data change, and EI represents the intensity of change. t EI represents the comprehensive economic heat index value during the t-collection period. t-1 This represents the comprehensive economic heat index value during the t-1 data collection period, tt t-1 The time interval between adjacent acquisition cycles is represented by C, which represents the value of the continuity of change. α, β, and γ represent the weighting coefficients of the values ​​of change amplitude, change rate, and change continuity, respectively, and α+β+γ=1.

[0042] The values ​​of the change amplitude, change rate, and change continuity are calculated according to the following formula:

[0043] ;

[0044] ;

[0045] ;

[0046] In the above formula, d1 represents the sign of the change direction of the t-1 acquisition period relative to the t-2 acquisition period, d2 represents the sign of the change direction of the t acquisition period relative to the t-1 acquisition period, and 0 indicates that the change direction is zero.

[0047] Preferably, determining the update frequency of the agricultural economic heat map based on the intensity of data change includes:

[0048] A first change intensity threshold and a second change intensity threshold are preset, wherein the first change intensity threshold is less than the second change intensity threshold;

[0049] The update frequency of the agricultural economic heat map is set according to the relationship between the intensity of data change and the first intensity of change threshold and the second intensity of change threshold.

[0050] If the intensity of the data change is less than or equal to the first intensity of change threshold, then the update frequency of the agricultural economic heat map is set to a low update frequency.

[0051] If the intensity of the data change is greater than the first intensity of change threshold, and the intensity of the data change is less than or equal to the second intensity of change threshold, then the update frequency of the agricultural economic heat map is set to a medium update frequency.

[0052] If the intensity of the data change is greater than the second intensity of change threshold, then the update frequency of the agricultural economic heat map is set to a high update frequency.

[0053] Preferably, the construction of an agricultural economic heat map of specialty agricultural products is specifically as follows: a hierarchical agricultural economic heat map visualization framework is constructed based on a geographic information system. The framework includes a basic geographic layer, a hierarchical display layer of circulation indicators, an anomaly warning layer, and a detailed interactive layer.

[0054] The basic geographic layer provides basic geographic information such as administrative divisions, transportation networks, and the distribution of production / sales areas; the circulation indicator layer allows users to switch between heat map displays of different core evaluation indicators; the anomaly warning layer marks areas where the comprehensive economic heat value exceeds the normal range in real time, using a flashing highlight effect for warning; and the details interaction layer allows users to click on a heat map area to view the circulation details of agricultural products in the corresponding area.

[0055] Preferably, when updating the agricultural economic heat map based on the update frequency, the method further includes:

[0056] If the agricultural product circulation period is a low-peak period, determine whether the update frequency of the agricultural economic heat map is a medium update frequency or a high update frequency. If so, update the agricultural economic heat map according to the medium update frequency or the high update frequency; otherwise, update the agricultural economic heat map according to the basic update frequency.

[0057] If the agricultural product circulation period is the peak circulation period, the agricultural economic heat map will be updated according to the set update frequency of the agricultural economic heat map.

[0058] The basic update frequency is lower than the low update frequency.

[0059] Preferably, when converting the comprehensive economic heat value into a color level and mapping it to the agricultural economic heat map, the color level is determined based on the comprehensive economic heat value, specifically as follows:

[0060] A first heat value, a second heat value, and a third heat value are preset, and the first heat value, the second heat value, and the third heat value increase sequentially;

[0061] Color levels are set according to the relationship between the comprehensive economic heat value and the first heat value, the second heat value, and the third heat value;

[0062] If the comprehensive economic heat value is less than the first heat value, then the color level is set to green.

[0063] If the comprehensive economic heat value is greater than or equal to the first heat value, and the comprehensive economic heat value is less than the second heat value, then the color level is set to yellow.

[0064] If the comprehensive economic heat value is greater than or equal to the second heat value, and the comprehensive economic heat value is less than the third heat value, then the color level is set to orange.

[0065] If the comprehensive economic heat value is greater than or equal to the third heat value, then the color level is set to red, with green, yellow, orange and red levels increasing in that order.

[0066] Compared with existing technologies, the beneficial effects of this invention are as follows: This invention constructs a multi-source data acquisition terminal network covering the entire circulation chain of specialty agricultural products, simultaneously acquiring multi-dimensional real-time information, providing comprehensive and three-dimensional data support for agricultural economic heat maps, effectively solving the problems of data fragmentation and information "islands" in the agricultural product circulation process, and significantly improving the monitoring efficiency of specialty agricultural product circulation; it introduces a dual anomaly detection mechanism, combining time-series weights and external environmental constraints, to accurately identify and remove abnormal data, significantly improving the quality and reliability of standardized circulation datasets; and it establishes a comprehensive economic heat value calculation model based on a core evaluation index system, determining index weights through a subjective and objective weight fusion method, making the heat map... Color mapping is more scientific and has greater economic explanatory power. By quantifying the magnitude, rate, and continuity of changes in comprehensive economic heat values, the intensity of data changes is calculated, and high, medium, and low update frequencies are adaptively set accordingly. This avoids invalid updates during low-activity periods and ensures real-time response during high-fluctuation periods, significantly improving system resource utilization efficiency and information timeliness. A basic update frequency safety net mechanism is introduced during off-peak periods to further save energy and reduce consumption. During peak periods, updates are strictly performed according to dynamic frequencies to ensure timely decision support. The resulting color-level heat map intuitively reflects the regional agricultural economic heat gradient, providing a visualized, quantifiable, and traceable intelligent support tool for agricultural enterprise scheduling, industry supervision, and farmer decision-making. Attached Figure Description

[0067] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0068] Figure 1 This is a flowchart illustrating a method for real-time updating of agricultural economic heat maps based on the circulation of specialty agricultural products, as proposed in this invention. Detailed Implementation

[0069] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of this application.

[0070] like Figure 1 As shown, this invention provides a method for real-time updating of agricultural economic heat maps based on the circulation of specialty agricultural products, including:

[0071] A multi-source data acquisition terminal network is constructed to collect real-time agricultural product circulation data across the entire circulation chain of specialty agricultural products. The real-time agricultural product circulation data includes basic circulation data, IoT sensing data, logistics tracking data, market transaction data, and environmental auxiliary data.

[0072] The real-time agricultural product circulation data is preprocessed and anomaly detected to obtain a standardized circulation dataset;

[0073] The core evaluation index system of the agricultural economic heat map is determined. The value of each core evaluation index is determined based on the standardized circulation dataset. The comprehensive economic heat value is determined based on the core evaluation index values. The core evaluation indicators include circulation scale indicators, circulation efficiency indicators, circulation benefit indicators and circulation loss indicators.

[0074] The intensity of data change in the entire agricultural product circulation chain is determined based on the comprehensive economic heat value, and the update frequency of the agricultural economic heat map is determined based on the intensity of data change. The intensity of data change is determined based on the magnitude, rate, and continuity of change of the comprehensive economic heat value.

[0075] An agricultural economic heat map of specialty agricultural products is constructed, the agricultural economic heat map is updated based on the update frequency, and the comprehensive economic heat value is converted into a color level and mapped onto the agricultural economic heat map.

[0076] This invention, by constructing a multi-source data acquisition terminal network, achieves comprehensive real-time data capture of the entire agricultural product supply chain from production to consumption, breaking through the limitations and lag of traditional data acquisition and providing a massive and dynamic data foundation for accurate heat map creation. Preprocessing and anomaly detection of real-time agricultural product circulation data effectively eliminates noisy data and outliers, ensuring the quality and reliability of the standardized circulation dataset and laying a solid foundation for the accurate calculation of subsequent core evaluation indicators. The established core evaluation indicator system covers key dimensions such as circulation scale, efficiency, benefits, and losses, reflecting the economic situation of agricultural product circulation from multiple perspectives and in depth. The comprehensive economic heat value determined based on this system quantifies and integrates complex circulation conditions. The update frequency of the agricultural economic heat map is dynamically adjusted according to the intensity of changes in the comprehensive economic heat value data, changing the previous fixed-period update mode. This allows the heat map to respond quickly when data fluctuations are severe and reduce unnecessary computational resource consumption when data is stable, achieving an optimal balance between update efficiency and resource costs. Ultimately, the comprehensive economic heat value is converted into color levels and mapped onto a heat map, presenting the distribution and dynamic changes of economic heat in agricultural product circulation in an intuitive and visual way. This provides scientific, accurate, and timely decision support for agricultural production planning, market regulation, logistics optimization, and policy formulation, helping to improve the efficiency of agricultural product circulation, reduce circulation losses, and improve the overall efficiency of agricultural economic operation.

[0077] Specifically, basic circulation data refers to data reflecting the most fundamental business activities of specialty agricultural products during the circulation process. It is typically generated at the place of origin or primary processing stage and serves as the starting point for subsequent circulation analysis. This type of data is characterized by strong structure, moderate update frequency, and direct correlation with product identity. Taking honey as an example, basic circulation data may include single batch harvest quantity (e.g., 500 kg), harvest date (e.g., April 12, 2025), nectar source plant type (e.g., acacia honey), bee farm number, and initial inspection pass rate; for tea, it includes picking grade (e.g., one bud and one leaf), fresh leaf weight, processing batch number, and initial processing completion time.

[0078] IoT-based sensing data consists of environmental and status parameters collected in real time by sensor devices deployed in production sites, transport vehicles, or transit warehouses. This data is used to monitor the physical conditions of agricultural products during distribution, ensuring quality and aiding in anomaly detection. For example, during honey storage and transportation, temperature and humidity sensors can upload real-time data on warehouse temperature (e.g., 22°C) and relative humidity (e.g., 60%RH). During tea transportation, vibration sensors record the degree of bumps and jolts, while gas sensors detect the introduction of off-odors. This data can be used to determine if suitable storage thresholds have been exceeded.

[0079] Logistics tracking data refers to the spatiotemporal trajectory and status information of agricultural products during transportation and distribution, obtained through GPS, BeiDou positioning, electronic waybill systems, or logistics platform interfaces, reflecting the circulation path and timeliness. For example, a batch of tea is shipped from a tea factory in Anxi, Fujian, and transported to a wholesale market in Shanghai via refrigerated trucks. The system records its departure time, the highway sections it passes through, the estimated arrival time, the actual receipt time, intermediate stops, and the duration of each stop. If the transportation time significantly exceeds the historical average, it may indicate a decrease in efficiency or abnormal delays.

[0080] Market transaction data, sourced from wholesale markets, e-commerce platforms, community group-buying platforms, or retail outlets in the sales areas, reflects the actual transaction volume of agricultural products at the consumer end and serves as a core basis for measuring circulation efficiency. For example, a certain brand of honey sold 300 bottles on a major fresh food e-commerce platform on a given day, with an average price of 85 yuan per bottle and a return rate of 2%; during the same period, in a wholesale agricultural market in Hangzhou, the daily transaction volume of bulk tea was 1.2 tons, with an average price of 220 yuan per kilogram. This type of data is directly used to calculate regional transaction activity and revenue levels.

[0081] Environmental auxiliary data does not come directly from the agricultural products themselves or their trading activities, but rather from macro- or regional background information provided by external public or industry information systems. This information is used to explain or correct the rationality of circulation data. For example, if a major tea-producing area experiences continuous heavy rain (meteorological data), it may lead to harvest delays or transportation disruptions. If circulation volume drops sharply in this situation, it should not be simply interpreted as a market downturn, but rather a secondary anomaly assessment should be made in conjunction with the weather conditions. Similarly, if a section of highway is closed due to an accident (traffic data), it can explain why the logistics time for a certain batch of honey has increased abnormally.

[0082] Specifically, preprocessing refers to cleaning, transforming, and structuring raw real-time agricultural product circulation data from multiple heterogeneous terminals to eliminate format differences, fill in missing data, unify spatiotemporal benchmarks, and improve data consistency, providing high-quality input for subsequent anomaly detection and indicator calculation. This process includes, but is not limited to, sub-steps such as standardization, spatiotemporal alignment, data cleaning, normalization, and time-series smoothing. For example, honey harvesting volume in a certain area is reported in "jin" (a unit of weight), while the logistics system records the load in "kilograms," requiring unification to international standard units; tea transaction timestamps come from e-commerce platforms in different time zones, needing conversion to a unified local time; and if a sensor experiences a brief power outage resulting in missing temperature and humidity data, this can be filled by interpolation of preceding and following periods or by averaging neighboring nodes. Furthermore, high-frequency fluctuating logistics location points can be smoothed using a sliding window averaging method to avoid false path fluctuations.

[0083] Anomaly detection is based on the inherent statistical regularities of the data. An improved Isolation Forest algorithm is used to perform unsupervised anomaly identification on preprocessed multidimensional circulation data. This method constructs a weighted time-series data matrix, assigning higher weights to recent data, thereby more sensitively capturing deviations in the current circulation state, and removing abnormal samples based on dynamic thresholds.

[0084] Secondary anomaly detection involves verifying the contextual reasonableness of initially screened data by combining external environmental auxiliary data (such as weather, traffic, and market activities). By pre-constructing an "environment-circulation correlation database," current environmental conditions are matched with historical experience rules to determine whether circulation data is within a reasonable fluctuation range, thereby identifying statistically normal but logically illogical situations. For example, during the "Double 11" e-commerce promotion, a certain brand of honey typically sees a 300% increase in online sales. If a region's sales only increase slightly by 10% on a given day, although this doesn't trigger an initial anomaly detection, the "promotion" market activity rules suggest that the data may have been missed or the system may not have synchronized, thus marking it as an anomaly. Similarly, after a typhoon warning is issued and main roads are closed, if large quantities of tea are still being transported normally, a secondary verification based on traffic control rules is necessary.

[0085] Specifically, the circulation scale indicator measures the overall volume of specialty agricultural products circulating within a specific region or time period, reflecting market activity and the breadth of supply chain coverage. This indicator is typically calculated based on standardized circulation data that centrally relates to quantity, frequency, or coverage, and is an important dimension for assessing the fundamental vitality of a region's agricultural economy.

[0086] The circulation efficiency index reflects the time cost, route optimization, and resource utilization of agricultural products throughout the entire process from production to sales, demonstrating the responsiveness and organizational capacity of the circulation system. This index relies on logistics tracking data, timestamp alignment data, and transportation network information for calculation.

[0087] The circulation efficiency indicator measures the economic value return generated by agricultural products during the circulation process, reflecting market acceptance and the profitability of the industrial chain. This indicator is mainly based on market transaction data, combined with cost or price information for quantification, and is key to judging the "gold content" of a region's agricultural economy.

[0088] The circulation loss index is used to quantify the quality or quantity loss of agricultural products throughout the entire supply chain from harvest to sale, reflecting the supply chain's preservation capabilities and management level. This index is calculated by combining IoT sensing data, logistics status, and final delivery and acceptance results.

[0089] The comprehensive economic heat index is a single quantitative indicator generated by scientifically weighting and integrating the above four core evaluation indicators. It is used to uniformly represent the overall economic activity of a region's specialty agricultural product circulation within a specific time period. The weights are determined using a combination of subjective and objective methods: subjective weights are set using the Analytic Hierarchy Process (AHP) combined with expert experience, while objective weights are calculated using the entropy weighting method based on data information entropy. Finally, a comprehensive weight is obtained through weighted summation, ensuring that the evaluation conforms to industry logic and reflects data distribution characteristics.

[0090] Specifically, the intensity of data change is a composite indicator used to quantify the dynamic fluctuations of agricultural economic activity over time. Its core purpose is to measure the "degree of drastic change" in the current economic climate compared to historical levels. This intensity is determined by three sub-dimensions:

[0091] Variation range: Reflects the absolute difference in the comprehensive economic heat index between adjacent weeks;

[0092] Rate of change: Reflects how quickly the heat value changes per unit of time;

[0093] Continuity of change: Determine whether the direction of change in popularity is consistent (such as continuous rise or fall) to avoid misjudging the trend due to a single accidental fluctuation.

[0094] The three factors are weighted and integrated to form a unified "data change intensity" value, which serves as the direct basis for subsequent decisions on the heatmap update frequency.

[0095] Update frequency refers to the time interval between re-rendering and publishing of the agricultural economic heatmap (e.g., every 5 minutes, hourly, daily). This invention abandons the fixed refresh mode and instead divides the update frequency into high, medium, and low levels based on the intensity of data change, achieving automatic switching through two preset thresholds:

[0096] Specifically, constructing an agricultural economic heat map refers to the process of spatially visualizing the comprehensive economic heat value of specialty agricultural products (such as honey, tea, and traditional Chinese medicine) across various administrative regions or distribution nodes, based on a Geographic Information System (GIS). This heat map is not simply a color-filled map, but rather an interactive visualization framework with a multi-layered structure, typically including a basic geographic information layer, a hierarchical indicator display layer, an anomaly warning layer, and a detailed interactive layer to support multi-dimensional analysis and decision-making. For example, on a national map, Anxi in Fujian Province has a high comprehensive economic heat value due to its active spring tea market, efficient logistics, and significant premium pricing; the corresponding area is displayed in red on the heat map. Conversely, a non-tea-producing area in Northwest China has a low heat value due to the lack of related distribution activities, and is displayed in green. Users can switch layers to view the distribution of "distribution efficiency" or "loss rate" separately, and can also click on red areas to bring up details such as the tea shipment volume, average price, and main destination cities for the past 24 hours.

[0097] Color-coded grading refers to dividing continuous comprehensive economic activity values ​​into several discrete levels according to preset grading thresholds and assigning corresponding color labels (such as green → yellow → orange → red) to intuitively reflect the gradient differences in regional economic activity. This mapping rule has clear business implications: green represents low activity (cold circulation zone), red represents high activity (hot circulation zone), and intermediate colors indicate transitional states.

[0098] In some embodiments of this application, a multi-source data acquisition terminal network is constructed, including: a harvest monitoring terminal at the production site, used to acquire the harvest quantity, harvest time, and category of specialty agricultural products; an Internet of Things sensor network, used to acquire temperature, humidity, and storage environment data of the production site of specialty agricultural products; a logistics tracking terminal, used to acquire the transportation route, transportation time, and cargo status of specialty agricultural products; a market transaction terminal at the sales destination, used to acquire the transaction amount, transaction volume, and transaction price of specialty agricultural products; and data interfaces from meteorological and transportation departments, used to acquire meteorological and transportation data.

[0099] Specifically, this solution constructs a multi-source data acquisition terminal network to comprehensively and in real-time acquire relevant data on specialty agricultural products (such as honey, tea, vegetables, and traditional Chinese medicine) from their production source to market sales. By integrating data from different sources, accurate and timely information support can be provided for subsequent analysis of the circulation status of agricultural products, thereby enabling effective assessment and monitoring of the agricultural economic climate.

[0100] First, the harvest monitoring terminal at the production site is used to collect harvesting information on specialty agricultural products at the production site, including but not limited to key data such as harvest volume, harvest time, and specific varieties. This data is crucial for understanding the yield and seasonal variations of agricultural products, and can help identify the optimal harvesting time and yield fluctuations.

[0101] Secondly, the Internet of Things (IoT) sensor network is responsible for collecting environmental parameters from the production areas of specialty agricultural products, such as temperature, humidity, and storage conditions. This type of data helps assess the environmental conditions of agricultural products during their growth stages and can further analyze the impact of environmental factors on the quality of agricultural products.

[0102] Furthermore, the role of logistics tracking terminals is to record detailed information about specialty agricultural products throughout the entire transportation process, including the transportation route, required time, and the status of the goods. This information is crucial for monitoring logistics efficiency and ensuring product quality and safety, and it also facilitates the timely detection and resolution of potential problems during transportation.

[0103] Next, the sales terminals in the destination markets aim to obtain detailed transaction information for specialty agricultural products, including transaction value, volume, and price. This data is crucial for measuring market demand, product popularity, and market price trends, and is essential for developing marketing strategies and adjusting production plans.

[0104] Finally, the data interfaces for meteorological and transportation departments are mainly used to access external meteorological and transportation data. This data can help predict the impact of weather changes on agricultural production, and also provide a reference for optimizing transportation routes and arranging logistics scheduling, thereby reducing losses caused by severe weather or traffic congestion.

[0105] By establishing such a multi-source data collection terminal network, it is possible to accurately track and deeply analyze the entire process of agricultural product circulation, providing strong support for improving the efficiency of agricultural product circulation and increasing agricultural economic benefits.

[0106] In some embodiments of this application, the real-time agricultural product circulation data is preprocessed and anomaly detected to obtain a standardized circulation dataset. Specifically, the preprocessing includes standardization, spatiotemporal alignment, cleaning, normalization, and temporal smoothing. An anomaly detection model is constructed based on an improved isolated forest algorithm to perform a first anomaly detection on the real-time agricultural product circulation data. A second anomaly detection is performed on the real-time agricultural product circulation data based on the environmental auxiliary data.

[0107] The anomaly detection process includes: constructing a time-series data matrix from the real-time agricultural product circulation data in chronological order, where each row of the time-series data matrix represents multi-dimensional data from a collection period, and each column represents a type of circulation data; assigning time-series weights to data at different time points based on the timeliness characteristics of the real-time agricultural product circulation data, and determining the impact of recent data on anomaly detection through weighted processing; constructing an improved isolated forest model, and based on the improved isolated forest model, traversing all samples, calculating the path length of each sample in each decision tree, and taking the average of all decision tree path lengths as the mean path length of the sample; calculating the anomaly score of the sample based on the mean path length; setting a dynamic anomaly threshold, comparing the sample score with the dynamic anomaly threshold, determining the anomaly samples based on the comparison results, and removing the anomaly samples.

[0108] Specifically, this application improves the quality and ensures the credibility of real-time agricultural product circulation data collected from multiple terminals. Through systematic preprocessing and a dual anomaly detection mechanism, noise, errors, or unreasonable data are eliminated, ultimately forming a standardized circulation dataset with a unified structure, reliable content, and suitability for subsequent analysis, laying a data foundation for the accurate construction of agricultural economic heat maps.

[0109] Specifically, the improved Isolation Forest algorithm in this application enhances the model's ability to identify time-series anomalies (such as sudden increases / decreases in circulation volume) by introducing a temporal weight factor to assign different weights to data at different time points. The dynamic anomaly threshold is adaptively adjusted based on the statistical characteristics of historical circulation data (such as mean and standard deviation) and real-time scenarios (such as raising the threshold during peak seasons and lowering it during off-seasons), avoiding misjudgments or omissions caused by fixed thresholds. During the collection and transmission of agricultural product circulation data, it is susceptible to interference from various factors, resulting in anomalies and noisy data: First, terminal equipment failure (such as sensor malfunction or GPS signal interruption) may lead to data collection distortion (such as constant temperature and humidity data or missing transportation location information); second, network transmission fluctuations may cause data packet loss, duplication, or numerical drift (such as sudden peaks in circulation volume data or negative transaction amounts); third, human operational errors (such as incorrect harvest volume entry or deviations in sales time entry) may introduce invalid data. If these anomaly data are not identified and removed, they will seriously affect the accuracy of subsequent agricultural economic indicator calculations, leading to a distortion of the comprehensive agricultural economic index. Consequently, the heat map will fail to accurately reflect the regional agricultural economic situation, misleading industry management and decision-making. Therefore, anomaly detection is a crucial step in ensuring data quality and improving the accuracy of heat map drawing.

[0110] The following are the specific steps for improving the Isolation Forest algorithm for anomaly detection:

[0111] Data processing preparation: Select time-series data of real-time agricultural product circulation dataset after data cleaning (such as circulation volume, circulation efficiency and other indicators data of multiple consecutive collection periods), construct a time-series data matrix in chronological order, with each row representing multi-dimensional data of a collection period and each column representing a circulation indicator.

[0112] Introducing a time-series weighting factor: Based on the timeliness characteristics of agricultural product circulation data, time-series weights are assigned to data at different time points. The weighting coefficient for recent data (such as the last 24 hours) is set to 0.7, and the weighting coefficient for older data (such as data from 24 hours ago) is set to 0.3. Through weighted processing, the influence of recent data on anomaly detection is strengthened.

[0113] Construct an improved isolated forest model: Initialize the number of decision trees in the forest (set to 100). For each decision tree, randomly select a subset of samples from the weighted time series data matrix (sampling ratio set to 0.7). Within each subset of samples, randomly select a feature dimension (such as flow rate or transportation time) and randomly select a split point within the value range of the feature. Recursively split the samples until each leaf node contains only one sample.

[0114] Calculate the outlier score: Traverse all samples and calculate the path length of each sample in each decision tree (i.e., the number of splits from the root node to the leaf node). Take the average of the path lengths of all decision trees as the mean path length of the sample. Calculate the outlier score of the sample based on the mean path length. The closer the outlier score is to 1, the higher the probability that the sample is an outlier. The closer it is to 0, the higher the probability that it is normal data.

[0115] Dynamic threshold setting and anomaly removal: Combining knowledge of agricultural product circulation (such as the fluctuation range of circulation volume during peak production and sales seasons and the normal logistics loss rate range) and historical data statistical characteristics (such as the mean ± 3 times the standard deviation of data in the past 3 months), dynamic anomaly thresholds are set; for example, the anomaly threshold for circulation volume during peak production and sales seasons is increased by 20%, and the anomaly threshold for logistics efficiency during winter rain and snow is decreased by 15%; the anomaly scores of the samples are compared with the dynamic thresholds, and samples that exceed the thresholds are judged as outliers and removed, finally obtaining a high-quality preprocessed dataset.

[0116] In some embodiments of this application, secondary anomaly detection is performed on the real-time agricultural product circulation data based on the environmental auxiliary data. Specifically, the environmental auxiliary data includes meteorological data, traffic data, and market activity data; an environment-circulation association database is constructed based on the environmental auxiliary data, which includes several rules and the fluctuation range of circulation data for each characteristic agricultural product; based on the real-time agricultural product circulation data of the current collection period, environmental auxiliary data of the corresponding time and region is extracted synchronously, and the rules corresponding to the current environmental auxiliary data and the fluctuation range of circulation data for the characteristic agricultural product are determined by retrieval and matching; the real-time agricultural product circulation data of the current collection period is compared with the retrieved fluctuation range of circulation data, and if there is real-time agricultural product circulation data that exceeds the fluctuation range of circulation data, the corresponding data is determined to be abnormal and removed.

[0117] Specifically, this solution, building upon anomaly detection based on statistical data characteristics, further incorporates external environmental context information to verify the rationality of real-time agricultural product circulation data. By introducing auxiliary environmental data such as meteorological, traffic, and market activity data, a secondary anomaly detection mechanism driven by business rules is constructed to identify data that appears statistically normal but is clearly inconsistent with actual environmental conditions, thereby improving the logical consistency and business credibility of the standardized circulation dataset.

[0118] Specifically, the anomaly detection method based on environmental auxiliary data can detect anomalies based on the strong correlation between environmental auxiliary data and agricultural product circulation data. The core logic is that agricultural product circulation data (such as circulation volume, transportation losses, and storage status) fluctuates regularly with changes in environmental conditions (weather, transportation, and policies). When the circulation data does not conform to the normal fluctuation range under the corresponding environmental conditions, it is judged as an anomaly. The specific steps are as follows:

[0119] Environmental auxiliary data classification and feature extraction: First, the collected environmental auxiliary data is classified into core influencing factors and secondary influencing factors. The core influencing factors include meteorological data (temperature, precipitation, wind force, cold wave / rainstorm disaster warnings, etc.), traffic data (traffic congestion index, road closure information, logistics hub operation status), and market activities (production and sales subsidy policies, market access rules). Key features of each factor are extracted (such as "daily precipitation ≥ 50 mm" and "wind force ≥ level 6" for meteorological data, "congestion index ≥ 8" for traffic data, and "peak production and sales season subsidy launch" for market activity data).

[0120] Constructing an environment-circulation correlation rule base: Based on knowledge of the agricultural product circulation field and historical data statistics, a rule base for the normal fluctuation range of circulation data under different combinations of environmental conditions is established. The rule base is structured around "combination of environmental conditions → normal range of circulation indicators," with example rules as follows:

[0121] Rule 1: When the meteorological conditions are "high temperature in summer (temperature ≥ 35℃) + no rainfall" and the traffic conditions are "normal (congestion index < 5)" → the normal range for agricultural product circulation loss rate is 3%-5% and the normal range for circulation time is 8-12 hours;

[0122] Rule 2: When the meteorological conditions are "rainstorm warning (daily precipitation ≥ 100 mm)" and the traffic conditions are "severe congestion (congestion index ≥ 8)", the normal range for agricultural product circulation volume is 60%-80% of the average daily value, and the normal range for transportation time is 1.5-2 times the average daily value.

[0123] Real-time environmental condition matching and interval retrieval: For the standardized circulation data of the current collection period, the corresponding time and region of the environmental auxiliary data are extracted simultaneously. The current combination of environmental conditions (such as "winter cold wave + highway closure + non-holiday") is determined by feature matching. Based on the combination, the normal fluctuation range of the corresponding circulation indicator is retrieved in the association rule base. If there are multiple similar combinations of environmental conditions, the normal range of each combination is merged by weighted average method (the weight is set according to the similarity of environmental conditions).

[0124] Anomaly detection in circulation data: The actual values ​​of each circulation indicator in the current period (such as circulation volume, loss rate, and transportation time) are compared with the retrieved normal fluctuation range. If the actual value of an indicator exceeds the range, the data is initially determined to be a potential anomaly. For scenarios involving multiple indicators (such as a sudden decrease in circulation volume and a sudden increase in transportation time), if multiple related indicators exceed the normal range, the confidence level of the anomaly detection can be improved.

[0125] Taking tea as an example: On April 5, 2025, a certain tea-producing area reported a daily harvest of 50 tons. The system simultaneously retrieved environmental data for that day: the local meteorological observatory issued a red alert for heavy rain, and the highway leading to the main sales area was closed due to a landslide. The environment-circulation correlation database contains the rule: "Red alert for heavy rain + closure of main roads → reasonable range for daily tea harvest is 0–5 tons." Since 50 tons far exceeds this range, even though this value is within the normal range of historical annual data, the system still judges it as abnormal data and removes it.

[0126] In some embodiments of this application, the comprehensive economic heat value is determined based on the core evaluation index values, specifically by: determining the subjective weights of each core evaluation index using the analytic hierarchy process combined with expert experience; calculating the objective weights of each core evaluation index based on the information entropy of a standardized circulation dataset using the entropy weight method; and obtaining the comprehensive weight of each core evaluation index by weighted summation; and determining the comprehensive economic heat value based on the comprehensive weights and the core evaluation index values.

[0127] The formula for calculating the comprehensive economic heat index is as follows:

[0128] ;

[0129] In the above formula, EI represents the comprehensive economic heat index value, wi represents the comprehensive weight of the i-th core evaluation indicator, and xi represents the value of the i-th core evaluation indicator.

[0130] Specifically, the core evaluation indicators include circulation scale indicators (such as the total circulation volume of agricultural products in the region and the circulation proportion of each category), circulation efficiency indicators (such as the time spent in the circulation process and the logistics turnover rate), circulation benefit indicators (such as the circulation transaction volume and the profit per unit product), and circulation loss indicators (such as the loss rate and the amount of loss in each stage).

[0131] Specifically, the steps for determining the subjective weights of each indicator based on the Analytic Hierarchy Process (AHP) are as follows:

[0132] The core of the Analytic Hierarchy Process (AHP) is to break down the problem of determining the weights of multiple indicators into hierarchical layers, combine expert experience to make pairwise comparisons, and finally quantify the subjective weights. Specific steps include:

[0133] Construct a hierarchical structure model: clarify the three-level structure of the target layer, the criteria layer, and the indicator layer; the target layer is to "determine the weight of the core evaluation indicators of the agricultural economic heat map"; the criteria layer consists of four primary indicators (circulation scale indicator, circulation efficiency indicator, circulation benefit indicator, and circulation loss indicator); the indicator layer consists of secondary indicators under each criteria layer (such as the total regional circulation volume of agricultural products and the circulation proportion of product categories under the circulation scale indicator).

[0134] Invite experts to form an evaluation team: Select 10-15 experts in the fields of agricultural economics, logistics management, and agricultural product circulation practice to ensure that the expert team covers both theoretical research and actual operation scenarios, and to guarantee the comprehensiveness of the judgment.

[0135] Constructing the judgment matrix: For the criterion layer (4 primary indicators), experts, based on their domain experience, compare the importance of any two indicators pairwise, using a 1-9 scale (1 indicates that the two indicators are equally important, 3 indicates that the former is slightly more important than the latter, 5 indicates that the former is significantly more important than the latter, 7 indicates that the former is strongly more important than the latter, 9 indicates that the former is extremely more important than the latter, 2, 4, 6, and 8 are the median values ​​of the above adjacent judgments, and the reciprocal indicates that the latter is more important than the former) to give the judgment value, and construct the criterion layer judgment matrix A (4×4 order); Similarly, for the secondary indicators of the indicator layer under each criterion layer, construct the indicator layer judgment matrices A1, A2, A3, and A4 respectively (e.g., there are 2 secondary indicators under the circulation scale indicator, and A1 is a 2×2 order matrix).

[0136] Example criterion-level judgment matrix A (using four primary indicators as an example):

[0137] A=[[a11,a12,a13,a14],[a21,a22,a23,a24],[a31,a32,a33,a34],[a41,a42,a43,a44]]

[0138] Where a_ij represents the importance scale value of the i-th primary indicator relative to the j-th primary indicator as determined by experts, and a_ii=1, a_ij=1 / a_ji.

[0139] Consistency check: Since expert subjective judgment may be biased, a consistency check is required to ensure the rationality of the judgment matrix.

[0140] (1) Calculate the maximum eigenvalue λ_max of the judgment matrix: The maximum eigenvalue λ_max of the judgment matrix is ​​obtained by solving the characteristic equation Aω=λω (ω is the eigenvector);

[0141] (2) Calculate the consistency index CI: CI=(λ_max-n) / (n-1), where n is the order of the judgment matrix (criteria layer n=4, index layer n is the number of secondary indicators under each criterion layer).

[0142] (3) Determine the average random consistency index RI: Based on the order n of the judgment matrix, query the standard RI value (e.g., RI=0 when n=2, RI=0.58 when n=3, RI=0.90 when n=4, and RI=1.12 when n=5).

[0143] (4) Calculate the consistency ratio CR: CR = CI / RI; if CR < 0.1, it means that the consistency of the judgment matrix is ​​good and the expert judgment is effective; if CR ≥ 0.1, experts need to be invited to readjust the scale value of the judgment matrix until the consistency requirements are met.

[0144] Calculating subjective weights: For the judgment matrix that passes the consistency test, the eigenvectors are calculated using the sum-product method. After normalizing the eigenvectors, the subjective weights of each indicator are obtained.

[0145] (1) Normalize the elements of each column of the judgment matrix: b_ij=a_ij / ∑(k=1 to n)a_kj (i,j=1 to n);

[0146] (2) Sum the normalized matrix by row: ω_i'=∑(j=1 to n)b_ij (i=1 to n);

[0147] (3) Normalize the summation result: ω_i=ω_i' / ∑(i=1 to n)ω_i', to obtain the subjective weights of each first-level indicator of the criteria layer W1=[w11,w12,w13,w14] (corresponding to circulation scale, circulation efficiency, circulation benefits, and circulation losses), and the subjective weights of each second-level indicator of the indicator layer relative to the corresponding criteria layer (such as the weights of the second-level indicators under circulation scale W11=[w111,w112]).

[0148] The total subjective weight of the secondary indicators is obtained by multiplying the subjective weight of the secondary indicators relative to the criteria layer with the subjective weight of the primary indicators of the criteria layer. For example, the total subjective weight of the total regional agricultural product circulation volume = w11 × w111, the total subjective weight of the category circulation ratio = w11 × w112, and so on, to obtain the subjective weight vector W_sub = [w_sub1, w_sub2, ..., w_subm] of all core evaluation indicators (m is the total number of core evaluation indicators).

[0149] Specifically, the steps for determining the objective weights of each indicator based on the entropy weight method are as follows:

[0150] The core of the entropy weighting method is to determine the dispersion of indicators based on the information entropy of the preprocessed dataset. The greater the dispersion (the smaller the information entropy), the greater the influence of the indicator on the overall evaluation, and the higher its weight. Specific steps are as follows:

[0151] Construct the indicator data matrix: Select the core evaluation indicator data from multiple collection periods (e.g., the last 30 days, 288 five-minute collection periods per day) in the preprocessed dataset (high-quality, anomaly-free circulating data) and construct the indicator data matrix X (n×m order), where n is the number of data samples (total number of collection periods), m is the total number of core evaluation indicators, and x_ij represents the value of the j-th indicator in the i-th sample.

[0152] Data normalization: Due to the different dimensions of the core evaluation indicators (e.g., total circulation volume is in tons, loss rate is in %), normalization is required to eliminate the influence of dimensions.

[0153] For positive indicators (the larger the value, the better, such as total circulation and trading volume): x_ij'=(x_ij-min(x_j)) / (max(x_j)-min(x_j))

[0154] For negative indicators (the smaller the value, the better, such as time spent in the circulation process and loss rate): x_ij'=(max(x_j)-x_ij) / (max(x_j)-min(x_j))

[0155] Where x_j represents all sample values ​​of the j-th indicator, and x_ij' represents the normalized data, with a value range of [0,1], resulting in the normalized indicator data matrix X' (n×m order).

[0156] Calculate the information entropy of each indicator: The formula for calculating information entropy is:

[0157] H_j=-k×∑(i=1 to n)(p_ij×lnp_ij), where k=1 / lnn, p_ij=x_ij' / ∑(i=1 to n)x_ij' (if x_ij'=0, define p_ij×lnp_ij=0)

[0158] H_j is the information entropy of the j-th indicator, with a value range of [0,1]. The smaller H_j is, the greater the dispersion of the sample data for that indicator, the more effective information it contains, and the greater its contribution to the overall evaluation. ln represents the natural logarithm operation.

[0159] Calculate the difference coefficient for each indicator: The difference coefficient reflects the distinguishing ability of the indicator. The calculation formula is: g_j=1-H_j; the larger g_j is, the stronger the distinguishing ability of the indicator, and the higher the weight should be assigned.

[0160] Calculate the objective weights: Normalize the difference coefficients to obtain the objective weight vector W_obj=[w_obj1,w_obj2,...,w_objm] for each core evaluation indicator. The calculation formula is as follows:

[0161] w_objj=g_j / ∑(j=1 to m)g_j

[0162] Specifically, the steps to obtain the comprehensive weight of each indicator based on weighted summation are as follows:

[0163] The comprehensive weighting combines subjective weighting (expert experience) and objective weighting (data characteristics), and is calculated using a weighted summation formula. The core principle is to adjust the proportion of each through weighting coefficients. Specific steps are as follows:

[0164] Determine the weighting ratios of subjective and objective factors: dynamically adjust the weighting ratios based on the real-time scenario of agricultural product circulation. Let the subjective weighting ratio be α and the objective weighting ratio be β, satisfying α+β=1.

[0165] Example scenario proportion settings:

[0166] (1) Normal scenario (no special production and sales fluctuations, no extreme environment): α=0.4, β=0.6 (balancing expert experience and data characteristics);

[0167] (2) Peak season scenario: α=0.5, β=0.5 (increase the weight of experts’ experience in judging the scale and efficiency indicators of circulation);

[0168] (3) Extreme weather scenarios (heavy rain, cold wave, etc.): α=0.5, β=0.5 (increase the weight of experts' experience in judging the circulation loss index);

[0169] (4) New regional promotion scenario (lack of historical data): α=0.6, β=0.4 (relying on expert experience to make up for insufficient data).

[0170] Weighted summation calculation of overall weight: For each core evaluation indicator, its overall weight is calculated using the following formula:

[0171] w_j=α×w_subj+β×w_objj

[0172] Where w_j is the comprehensive weight of the j-th core evaluation indicator, w_subj is the subjective weight of the j-th indicator, and w_objj is the objective weight of the j-th indicator.

[0173] Comprehensive weight normalization verification: Calculate the sum of the comprehensive weights of all indicators. If the sum deviates from 1 (error ≤ 0.01), normalization is performed to ensure that the sum of weights is 1. Finally, the comprehensive weight vector W=[w1,w2,...,wm] of each core evaluation indicator is obtained, which is used for the subsequent calculation of the comprehensive agricultural economic index.

[0174] In some embodiments of this application, the intensity of data change in the entire agricultural product circulation chain is determined based on the comprehensive economic heat value, specifically by: determining the magnitude, rate, and continuity of change of the comprehensive economic heat value; and determining the intensity of data change in the entire agricultural product circulation chain based on the magnitude, rate, and continuity of change of the comprehensive economic heat value.

[0175] The intensity of the data change is determined according to the following formula:

[0176] ;

[0177] In the above formula, CI represents the intensity of data change, and EI represents the intensity of change. t EI represents the comprehensive economic heat index value during the t-collection period. t-1 This represents the comprehensive economic heat index value during the t-1 data collection period, tt t-1 The time interval between adjacent acquisition cycles is represented by C, which represents the value of the continuity of change. α, β, and γ represent the weighting coefficients of the values ​​of change amplitude, change rate, and change continuity, respectively, and α+β+γ=1.

[0178] The values ​​of the change amplitude, change rate, and change continuity are calculated according to the following formula:

[0179] ;

[0180] ;

[0181] ;

[0182] In the above formula, d1 represents the sign of the change direction of the t-1 acquisition period relative to the t-2 acquisition period, d2 represents the sign of the change direction of the t acquisition period relative to the t-1 acquisition period, and 0 indicates that the change direction is zero.

[0183] Specifically, if EI t -EI t-1 If EI > 0, denote the direction of change as positive (sign +1); if EI t -EI t-1 If EI < 0, the direction of change is recorded as negative (sign -1); if EI t =EI t-1 Let the direction of change be zero (symbol 0). When the direction of change of two consecutive adjacent periods is the same and not zero, it indicates that the exponential change has good continuity, and C is 1; when the direction of change is inconsistent, or the direction of change of any period is zero (no change in the exponent), the continuity of change is weak, and C is 0.5.

[0184] Specifically, regarding the range and constraints of α, β, and γ, α, β, and γ are all non-negative real numbers with a range of [0,1]; they satisfy the normalization condition: α+β+γ=1; and the three represent the relative importance of the magnitude of change, the rate of change, and the continuity of change in the comprehensive evaluation.

[0185] In this invention, α, β, and γ can be determined using any one or a combination of the following methods, all of which are reasonable implementation methods:

[0186] a) Expert experience method

[0187] The settings are determined by experts in agricultural economics, logistics, or data science based on the characteristics of the specialty agricultural products. For example, for price-sensitive products (such as high-end tea), a higher α setting is used to focus on short-term, sharp fluctuations; for time-sensitive distribution (such as cold chain logistics for fresh honey), a higher β setting is used to focus on the speed of change; and for products with obvious seasonality, a higher γ setting is used to emphasize trend continuity.

[0188] b) Historical data backtesting optimization method

[0189] Based on historical circulation data, the matching degree between CI and actual business events (such as price fluctuations and logistics disruptions) under different combinations of α, β, and γ is simulated, and the weight combination that maximizes the accuracy of early warning or minimizes the false alarm rate is selected.

[0190] c) Dynamic adaptive adjustment

[0191] During system operation, multiple preset weight templates are automatically switched based on the current circulation period (peak / off-peak) or regional characteristics (main production area / consumption area). For example:

[0192] Peak period: α=0.5, β=0.3, γ=0.2 (emphasizing abrupt changes); Low period: α=0.2, β=0.2, γ=0.6 (emphasizing trend stability).

[0193] By introducing three dimensions—variation magnitude, rate of change, and continuity—and fusing them with configurable weighting coefficients α, β, and γ, this step achieves a refined characterization of the dynamic features of agricultural economic activity. This mechanism retains sensitivity to sudden fluctuations while also considering the assessment of trend stability, providing a reliable basis for intelligent decision-making regarding the subsequent heatmap update frequency.

[0194] In some embodiments of this application, determining the update frequency of the agricultural economic heat map based on the intensity of data change includes: pre-setting a first intensity of change threshold and a second intensity of change threshold, wherein the first intensity of change threshold is less than the second intensity of change threshold; setting the update frequency of the agricultural economic heat map based on the relationship between the intensity of data change and the first intensity of change threshold and the second intensity of change threshold; if the intensity of data change is less than or equal to the first intensity of change threshold, then setting the update frequency of the agricultural economic heat map to a low update frequency; if the intensity of data change is greater than the first intensity of change threshold and the intensity of data change is less than or equal to the second intensity of change threshold, then setting the update frequency of the agricultural economic heat map to a medium update frequency; if the intensity of data change is greater than the second intensity of change threshold, then setting the update frequency of the agricultural economic heat map to a high update frequency.

[0195] Specifically, the core of this solution lies in establishing a heat map update mechanism based on dynamic grading of data change intensity to replace the traditional fixed-cycle refresh mode. The system pre-sets two thresholds: the first change intensity threshold (T1) and the second change intensity threshold (T2), and T1 < T2 is satisfied. These two thresholds form three intervals, corresponding to three update frequencies: low, medium, and high: When the currently calculated data change intensity CI ≤ T1: It indicates that the agricultural product circulation state is stable without significant fluctuations, and a low update frequency (such as once a day) is adopted to save computing and network resources; when T1 < CI ≤ T2: It indicates that the circulation state has moderate changes, with local activities or adjustments, and a medium update frequency (such as once every 4 hours) is adopted to balance timeliness and efficiency; when CI > T2: It indicates that the circulation heat has changed violently (such as concentrated listing, price fluctuations, logistics interruption, etc.), which requires high attention, and a high update frequency (such as once every 30 minutes or real-time push) is adopted to ensure that decision-making information is available in time. This mechanism realizes "updating on demand" - high-frequency response when data changes violently, and reducing the system load during the stable period, improving the overall operation efficiency and user experience.

[0196] Taking the main tea-producing areas as an example:

[0197] Scenario 1 (low frequency): During the winter production break period, the comprehensive economic heat value has been stable at around 40 for several consecutive days, and the calculated CI = 1.2. If T1 = 3.0 and T2 = 8.0 are set, then CI ≤ T1, and the system adopts a low update frequency (such as updating the heat map once at 2 am every day);

[0198] In some embodiments of the present application, an agricultural economic heat map of characteristic agricultural products is constructed. Specifically, a visualization framework for an agricultural economic hierarchical heat map is constructed based on a geographic information system. The framework includes a basic geographic layer, a hierarchical display layer of circulation indicators, an abnormal warning layer, and a detail interaction layer; the basic geographic layer is used to provide basic geographic information such as administrative region division, traffic road network, and origin / destination distribution; the hierarchical display layer of circulation indicators is used to support users to switch the heat map display of different core evaluation indicators; the abnormal warning layer is used to mark in real time the areas where the comprehensive economic heat value exceeds the normal range, and a flashing and highlighting effect is used for warning; the detail interaction layer is used to support users to click on the heat map area to view the details of the agricultural product circulation in the corresponding area.

[0199] Taking tea as an example: After opening the system, users first see a national map displayed in the basic geographic layer, marking major producing areas such as Anxi in Fujian and Longjing in Zhejiang, as well as major sales destinations such as Beijing and Shanghai, with the highway network overlaid. In the circulation indicator layer, when users select the "circulation efficiency" indicator, the map immediately displays the average selling price and premium level of tea in various regions using color levels (green to red). Anxi is displayed in orange due to its brand effect, while ordinary producing areas are displayed in yellow. At the same time, the anomaly warning layer detects that the daily heat value of a certain Yunnan tea region suddenly increases to 95 (exceeding the third heat value of 90). The system automatically overlays a red flashing border on this area, indicating that there may be price anomalies or data reporting errors. When users click on the flashing area, a details interaction window pops up, displaying: "Daily harvest volume 12 tons (+300%), average price 420 yuan / kg (+180%), mainly sold to Guangdong, logistics transit time 1.2 days", to help users determine whether it is a real market behavior.

[0200] In some embodiments of this application, constructing an agricultural economic heat map of specialty agricultural products specifically involves: building a hierarchical agricultural economic heat map visualization framework based on a geographic information system.

[0201] Specifically, the agricultural economic heat map for specialty agricultural products can be constructed based on historical circulation data or initial circulation data. During construction, a mature GIS engine is used as the underlying framework, providing basic geographic information loading and management capabilities such as administrative region division, geographic coordinate mapping, road network and point marking, which is the core foundational technology for layered visualization. Layer overlay and layered management techniques are employed to independently encapsulate and overlay the basic geographic layer, circulation indicator layer, anomaly warning layer, and detailed interactive layer. This supports individual rendering, hiding / showing, and layer adjustment of each layer, achieving an orderly presentation of multi-dimensional information.

[0202] In some embodiments of this application, when updating the agricultural economic heat map based on the update frequency, the method further includes: obtaining the agricultural product circulation period; if the agricultural product circulation period is a low-peak circulation period, determining whether the set update frequency of the agricultural economic heat map is a medium update frequency or a high update frequency; if so, updating the agricultural economic heat map according to the medium update frequency or the high update frequency; otherwise, updating the agricultural economic heat map according to the basic update frequency; if the agricultural product circulation period is a high-peak circulation period, updating the agricultural economic heat map according to the set update frequency of the agricultural economic heat map; wherein, the basic update frequency is lower than the low update frequency.

[0203] Specifically, the system first identifies the current agricultural product circulation period, which can be determined through historical production and sales patterns, phenological characteristics, policy announcements, or actual circulation volume thresholds. If the current period is a low-peak circulation period (such as the off-season or off-season), the set update frequency is verified: if the original setting was a medium or high update frequency (i.e., high data change intensity), the high-frequency update is retained to cope with possible abnormal fluctuations during the low-peak period (such as sudden promotions or temporary logistics interruptions); if the original setting was a low update frequency or lower, it is downgraded to the basic update frequency (slower than the low update frequency, such as once a week) to maximize the conservation of system resources; if the current period is a peak circulation period (such as the spring tea picking season, honey harvesting season, or e-commerce promotion period), the update frequency (low / medium / high) set according to the data change intensity is directly adopted without downgrading, ensuring the real-time information during critical periods. The "basic update frequency" is the minimum refresh rate preset by the system, with a time interval longer than the low update frequency, and is only used for minimal maintenance under long-term stable conditions. This mechanism achieves the dual optimization goals of "full response during peak periods and intelligent cost-saving during off-peak periods".

[0204] Take honey as an example:

[0205] Scenario 1: Low circulation period + stable data

[0206] It is currently December, which is the off-season for honey harvesting (low peak season). The overall economic heat value has remained stable for several consecutive days, with low data volatility. The system's original update frequency was set to "low update frequency" (e.g., once a day). According to this step, the system will be further downgraded to "basic update frequency" (e.g., once every 3 days) to reduce server load.

[0207] Scenario 2: Low circulation period + sudden data change

[0208] Similarly, during the off-peak period in December, a certain region experienced a surge in daily sales due to temporary subsidies from e-commerce platforms, resulting in high data volatility. The system was originally set to "high update frequency" (every 30 minutes). Although it was during the off-peak period, because the frequency was set to medium or high, the system still updated the heat map at a high frequency and triggered anomaly alerts to avoid missing special events.

[0209] In some embodiments of this application, when converting the comprehensive economic heat value into a color level and mapping it to the agricultural economic heat map, the color level is determined based on the comprehensive economic heat value, specifically as follows: a first heat value, a second heat value, and a third heat value are pre-set, with the first heat value, the second heat value, and the third heat value increasing sequentially; the color level is set according to the relationship between the comprehensive economic heat value and the first heat value, the second heat value, and the third heat value; if the comprehensive economic heat value is less than the first heat value, the color level is set to green; if the comprehensive economic heat value is greater than or equal to the first heat value and the comprehensive economic heat value is less than the second heat value, the color level is set to yellow; if the comprehensive economic heat value is greater than or equal to the second heat value and the comprehensive economic heat value is less than the third heat value, the color level is set to orange; if the comprehensive economic heat value is greater than or equal to the third heat value, the color level is set to red, wherein the green level, yellow level, orange level, and red level increase sequentially.

[0210] Specifically, this solution aims to convert the calculated comprehensive economic heat value (EI) into color levels that users can intuitively understand, and map it to the corresponding geographical area of ​​the agricultural economic heat map, so as to realize the spatial visualization of the activity level of agricultural product circulation.

[0211] The specific implementation method is as follows:

[0212] Three incremental thresholds—the first heat value (H1), the second heat value (H2), and the third heat value (H3)—are pre-defined, forming four consecutive intervals. Based on the interval in which the comprehensive economic heat value (EI) falls, a corresponding color level is assigned: Green level: EI

[0213] Taking tea as an example, assuming the system presets H1=50, H2=70, and H3=90:

[0214] ​A county in Fujian province had a daily comprehensive economic heat index of 45 (<50), displayed in green, indicating no harvesting or trading activity and the off-season. A city in Zhejiang province had a heat index of 62 (50≤62<70), displayed in yellow, reflecting the beginning of spring tea harvesting and the gradual market recovery. A production area in Yunnan province had a heat index of 82 (70≤82<90), displayed in orange, indicating a large influx of new tea, active trading, and increased logistical pressure. A wholesale market in Guangdong province, due to a rush to buy "pre-Qingming tea," had a heat index of 95 (≥90), displayed in red, indicating soaring prices or tight inventory, and was automatically marked as a high-risk area by the monitoring platform. Users can easily judge the circulation situation in various regions by color, without needing to look up specific numerical values.

[0215] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program goods. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program goods embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0216] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. An agricultural economic heat map real-time updating method based on characteristic agricultural product circulation, characterized in that, include: A multi-source data acquisition terminal network is constructed to collect real-time agricultural product circulation data across the entire circulation chain of specialty agricultural products. The real-time agricultural product circulation data includes basic circulation data, IoT sensing data, logistics tracking data, market transaction data, and environmental auxiliary data. The real-time agricultural product circulation data is preprocessed and anomaly detected to obtain a standardized circulation dataset; The core evaluation index system of the agricultural economic heat map is determined. The value of each core evaluation index is determined based on the standardized circulation dataset. The comprehensive economic heat value is determined based on the core evaluation index values. The core evaluation indicators include circulation scale indicators, circulation efficiency indicators, circulation benefit indicators and circulation loss indicators. The intensity of data change in the entire agricultural product circulation chain is determined based on the comprehensive economic heat value, and the update frequency of the agricultural economic heat map is determined based on the intensity of data change. The intensity of data change is determined based on the magnitude, rate, and continuity of change of the comprehensive economic heat value. An agricultural economic heat map of specialty agricultural products is constructed, the agricultural economic heat map is updated based on the update frequency, and the comprehensive economic heat value is converted into a color level and mapped onto the agricultural economic heat map. 2.The method of claim 1, wherein, Construct a multi-source data acquisition terminal network, including: a harvest monitoring terminal at the production site, used to obtain the harvest quantity, harvest time and category of specialty agricultural products; Internet of Things (IoT) sensor networks are used to acquire data on temperature, humidity, and storage environment in the production areas of specialty agricultural products; Logistics tracking terminals are used to obtain information on the transportation routes, transportation time, and cargo status of specialty agricultural products. The sales market trading terminal is used to obtain the transaction amount, transaction volume, and transaction price of specialty agricultural products; Meteorological and transportation data interfaces are used to acquire meteorological and transportation data. 3.The method of claim 1, wherein, The real-time agricultural product circulation data is preprocessed and anomaly detected to obtain a standardized circulation dataset, specifically: The preprocessing includes standardization, spatiotemporal alignment, cleaning, normalization, and temporal smoothing. An anomaly detection model is constructed based on the improved isolated forest algorithm to perform anomaly detection on the real-time agricultural product circulation data. A secondary anomaly detection is performed on the real-time agricultural product circulation data based on the aforementioned environmental auxiliary data. One of the anomaly detection steps includes: A time-series data matrix is ​​constructed from the real-time agricultural product circulation data in chronological order. Each row of the time-series data matrix represents multi-dimensional data of a collection period, and each column represents a type of circulation data. Based on the timeliness characteristics of real-time agricultural product circulation data, time-series weights are assigned to data at different time points, and the impact of recent data on anomaly detection is determined through weighted processing. An improved isolated forest model is constructed. Based on the improved isolated forest model, all samples are traversed, and the path length of each sample in each decision tree is calculated. The average of the path lengths of all decision trees is taken as the mean path length of the sample. The anomaly score of the sample is calculated based on the mean path length. A dynamic anomaly threshold is set, the anomaly score is compared with the dynamic anomaly threshold, anomaly samples are identified based on the comparison result, and the anomaly samples are removed. The improved isolated forest algorithm introduces a time-series weight factor, assigning different weights to data at different time points, thereby enhancing the model's ability to identify time-series abnormal data. 4.The method of claim 3, wherein, A secondary anomaly detection is performed on the real-time agricultural product circulation data based on the aforementioned environmental auxiliary data, specifically as follows: The environmental auxiliary data includes meteorological data, traffic data, and market activity data; An environment-circulation correlation database is constructed based on the aforementioned environmental auxiliary data. The environment-circulation correlation database includes several rules and the fluctuation range of circulation data for each rule for the characteristic agricultural products. Based on the real-time agricultural product circulation data of the current collection period, environmental auxiliary data of the corresponding time and region are extracted simultaneously. The rules corresponding to the current environmental auxiliary data and the fluctuation range of circulation data of specialty agricultural products are determined by retrieval and matching. The fluctuation range of real-time agricultural product circulation data in the current collection period is compared with that of the retrieved circulation data. If real-time agricultural product circulation data exceeds the fluctuation range of circulation data, the corresponding data is determined to be abnormal and removed. 5.The method of claim 1, wherein the method further comprises: receiving a request for a specific agricultural product from a user; and providing the user with information on the specific agricultural product. The comprehensive economic heat index value is determined based on the aforementioned core evaluation index values, specifically as follows: The subjective weights of each core evaluation indicator are determined by combining the analytic hierarchy process with expert experience. Then, the objective weights of each core evaluation indicator are calculated based on the information entropy of the standardized circulation dataset using the entropy weight method. Finally, the comprehensive weight of each core evaluation indicator is obtained by weighted summation. The comprehensive economic heat index value is determined based on the aforementioned comprehensive weights and core evaluation index values. The formula for calculating the comprehensive economic heat index is as follows: ； In the above formula, EI represents the comprehensive economic heat index value, wi represents the comprehensive weight of the i-th core evaluation indicator, and xi represents the value of the i-th core evaluation indicator. 6.The method of claim 1, wherein the method further comprises: receiving a request for a specific agricultural product from a user; and providing the user with information on the specific agricultural product. The intensity of data change across the entire agricultural product distribution chain is determined based on the aforementioned comprehensive economic heat index value, specifically as follows: The magnitude, rate, and continuity of change of the comprehensive economic heat index value are determined based on the comprehensive economic heat index value. The intensity of data change across the entire agricultural product circulation chain is determined based on the magnitude, rate, and continuity of the changes in the comprehensive economic heat index value. The intensity of the data change is determined according to the following formula: ； In the formula, CI represents the data change intensity, EI t represents the comprehensive economic heat value of the t collection period, EI t-1 represents the comprehensive economic heat value of the t-1 collection period, t-t t-1 represents the time interval of adjacent collection periods, C represents the change continuity value, and α, β, and γ represent the weight coefficients of the change amplitude, change rate, and change continuity value respectively, and α+β+γ=1. The values ​​of the change amplitude, change rate, and change continuity are calculated according to the following formula: ； ； ； In the above formula, d1 represents the sign of the change direction of the t-1 acquisition period relative to the t-2 acquisition period, d2 represents the sign of the change direction of the t acquisition period relative to the t-1 acquisition period, and 0 indicates that the change direction is zero. 7.The method of claim 1, wherein the method further comprises: receiving a request for a specific agricultural product from a user; and providing the user with information on the specific agricultural product. The update frequency of the agricultural economic heat map is determined based on the intensity of the data changes, including: A first change intensity threshold and a second change intensity threshold are preset, wherein the first change intensity threshold is less than the second change intensity threshold; The update frequency of the agricultural economic heat map is set according to the relationship between the intensity of data change and the first intensity of change threshold and the second intensity of change threshold. If the intensity of the data change is less than or equal to the first intensity of change threshold, then the update frequency of the agricultural economic heat map is set to a low update frequency. If the intensity of the data change is greater than the first intensity of change threshold, and the intensity of the data change is less than or equal to the second intensity of change threshold, then the update frequency of the agricultural economic heat map is set to a medium update frequency. If the intensity of the data change is greater than the second intensity of change threshold, then the update frequency of the agricultural economic heat map is set to a high update frequency. 8.The method of claim 7, wherein the method further comprises: receiving a request for a specific agricultural product from a user; and providing the user with information on the specific agricultural product. Constructing an agricultural economic heat map of specialty agricultural products involves: building a hierarchical agricultural economic heat map visualization framework based on a geographic information system. The framework includes a basic geographic layer, a hierarchical display layer of circulation indicators, an anomaly warning layer, and a detailed interactive layer. The basic geographic layer is used to provide basic geographic information on administrative divisions, transportation networks, and the distribution of production / sales locations. The circulation indicator layer is used to support users in switching between heatmap displays of different core evaluation indicators; the anomaly warning layer is used to mark areas where the comprehensive economic heat value exceeds the normal range in real time, and uses a flashing highlighting effect to issue warnings. The details interaction layer allows users to click on a heatmap area to view the corresponding agricultural product circulation details. 9.The method of claim 7, wherein the method further comprises: receiving a request for a specific agricultural product from a user; and providing the user with information on the specific agricultural product. When updating the agricultural economic heatmap based on the update frequency, the method further includes: If the agricultural product circulation period is a low-peak period, determine whether the update frequency of the agricultural economic heat map is a medium update frequency or a high update frequency. If so, update the agricultural economic heat map according to the medium update frequency or the high update frequency; otherwise, update the agricultural economic heat map according to the basic update frequency. If the agricultural product circulation period is the peak circulation period, the agricultural economic heat map will be updated according to the set update frequency of the agricultural economic heat map. The basic update frequency is lower than the low update frequency. 10.The method of claim 1, wherein the method further comprises: receiving a request for a specific agricultural product from a user; and providing the user with information on the specific agricultural product. When converting the comprehensive economic heat value into a color level and mapping it to the agricultural economic heat map, the color level is determined based on the comprehensive economic heat value, specifically as follows: A first heat value, a second heat value, and a third heat value are preset, and the first heat value, the second heat value, and the third heat value increase sequentially; Color levels are set according to the relationship between the comprehensive economic heat value and the first heat value, the second heat value, and the third heat value; If the comprehensive economic heat value is less than the first heat value, then the color level is set to green. If the comprehensive economic heat value is greater than or equal to the first heat value, and the comprehensive economic heat value is less than the second heat value, then the color level is set to yellow. If the comprehensive economic heat value is greater than or equal to the second heat value, and the comprehensive economic heat value is less than the third heat value, then the color level is set to orange. If the comprehensive economic heat value is greater than or equal to the third heat value, then the color level is set to red, with green, yellow, orange and red levels increasing in that order.

Images