Agricultural product-oriented data information tracing method and system

By analyzing the extreme point trends and commonalities of agricultural products under key growth conditions, the inaccuracy of agricultural product traceability methods is solved, and a more accurate and credible traceability information transmission is achieved.

CN119067679AActive Publication Date: 2024-12-03黑龙江省农业科学院农业遥感与信息研究所

Patent Information

Application Number
CN202411094040.7
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-09
Publication Date
2024-12-03
Estimated Expiration
2044-08-09

AI Technical Summary

Technical Problem

The prior art is difficult to trace agricultural product information accurately and credibly, and it is impossible to effectively evaluate the impact of external factors on crop quality, resulting in inaccuracy of traceability methods.

Method used

By obtaining the growth data of agricultural products under various key growth conditions, we determine the potential impact trends of extreme value points, and classifying and data-on-chaining different batches of agricultural products based on these trends and commonality, ensuring the accuracy and ability of traceability information Reliability.

Benefits of technology

It improves the representativeness and accuracy of agricultural product characteristics, avoids the inaccuracy of data judgment in a single time period, enhances the representativeness and accuracy of classification, and ensures the accuracy and credibility of traceability methods and systems.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119067679A_ABST
    Figure CN119067679A_ABST
Patent Text Reader

Abstract

The invention relates to the technical field of data processing, in particular to an agricultural product-oriented data information tracing method and system, and the method comprises the steps: obtaining the quality score of each batch of to-be-traced agricultural products, and determining the quality score difference between different batches; obtaining growth data of the agricultural product to be traced under various key growth conditions, and determining a potential influence trend of each extreme point; according to the time sequence difference between the extreme points under different key growth conditions and the potential influence trend difference between the extreme points under the same key growth condition, determining the generality degree between the different key growth conditions; determining the similarity between the to-be-traced agricultural products according to the generality degree and the quality score difference; and classifying different batches of to-be-traced agricultural products according to the similarity, and endowing the same traceability information and data to the to-be-traced agricultural products of the same category with the same traceability information. According to the method, the feature representativeness and accuracy of different agricultural products are improved, and the representativeness and accuracy of final clustering are improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular to a data information tracing method and system for agricultural products. Background Art

[0002] The introduction of technologies such as blockchain and smart contracts can effectively solve problems such as information opacity, difficulty in data transmission and easy tampering in the agricultural product traceability system, and achieve efficient, accurate and reliable traceability of the entire agricultural product production process.

[0003] The growth process of crops takes a relatively long time, and the impact of various external factors on the quality of crops cannot be simply evaluated. At the same time, since crops themselves have different requirements for external conditions, and for any crop, the external conditions need to be kept in a balanced state within the prescribed range, the surge of a single type of external factor will cause problems with the quality of crops, and even cause the entire batch of crops to die directly. If you simply collect the various data generated by agricultural products under external conditions and upload the data to the chain, then the accuracy and credibility of the information traceability of agricultural products based on these data cannot be guaranteed. Therefore, it is also necessary to process the collected crop data accordingly based on the short-term changes of various external factors of agricultural products and the overall balance relationship before uploading them to the chain, so as to ensure the accuracy and credibility of the traceability method and system.

[0004] Therefore, how to analyze the impact of short-term changes in various external conditions and the overall balance between them on agricultural products, and then better trace the information of agricultural products is an issue that needs to be urgently addressed in the industry. Summary of the invention

[0005] In order to solve the technical problem that it is difficult to accurately trace the information of agricultural products only by various data generated by external conditions of agricultural products, the purpose of the present invention is to provide a data information traceability method and system for agricultural products. The technical solution adopted is as follows:

[0006] According to the first aspect, an embodiment of the present invention provides a data information tracing method for agricultural products, the method comprising:

[0007] Obtaining the quality score of each batch of agricultural products to be traced, and determining the difference in quality scores between different batches of agricultural products to be traced;

[0008] Acquire the growth data of the agricultural product to be traced under various key growth conditions, and determine the potential impact trend of each extreme point according to the local abnormal data in the growth data; the potential impact trend has a mapping relationship with the quality of the agricultural product;

[0009] Determine the degree of commonality between different key growth conditions based on the time sequence differences between extreme points of the agricultural product to be traced under different key growth conditions and the potential impact trend differences between extreme points under the same key growth conditions;

[0010] Determining the similarities between different batches of the agricultural products to be traced based on the commonality levels and the quality score differences;

[0011] Different batches of the agricultural products to be traced are classified according to the similarity, and the same traceability information is assigned to the agricultural products to be traced of the same category, and the data of the agricultural products to be traced with the traceability information is uploaded to the chain.

[0012] In combination with the first aspect, in the first implementation of the first aspect, the acquiring of growth data of the agricultural product to be traced under various key growth conditions, and determining the potential impact trend of each extreme point according to local abnormal data in the growth data, specifically includes:

[0013] Acquire growth data of the agricultural product to be traced under each of the key growth conditions, sort the growth data in time sequence, and determine the extreme value points in the growth data corresponding to each of the key growth conditions;

[0014] The time period between adjacent extreme value points having the same characteristics is taken as the extreme value time period, and a plurality of extreme value time periods and the extreme value time periods corresponding to the extreme value points are obtained; the same characteristics are the same maximum value or the same minimum value;

[0015] According to the correspondence between the extreme time period where the extreme point is located and the appropriate range of the corresponding key growth condition, the potential influence trend of the extreme point is obtained; each phase of the key growth condition has the corresponding appropriate range.

[0016] In combination with the first implementation manner of the first aspect, in the second implementation manner of the first aspect, obtaining the potential influence trend of the extreme value point according to the correspondence between the extreme value time period where the extreme value point is located and the appropriate range of the corresponding key growth condition specifically includes:

[0017] Determining the extreme value time period in which the extreme value point is located;

[0018] Obtaining a first number and a second number within the extreme value time period; the first number is the total number of data points within the extreme value time period that are lower than the minimum value of the suitable range corresponding to the key growth condition; the second number is the total number of data points within the extreme value time period that exceed the maximum value of the suitable range;

[0019] Determining a maximum value of the extreme value time period;

[0020] Determining a mean value of the excess of the data points exceeding the suitable range;

[0021] The potential influence trend of the extreme value point is obtained according to the first quantity, the second quantity, the total number of data points in the extreme value time period, the maximum value of the extreme value time period and the mean value of the excess amount.

[0022] In combination with the first aspect, in a third implementation of the first aspect, determining the degree of commonality between different key growth conditions according to the time series differences between extreme points of the agricultural product to be traced under different key growth conditions and the potential impact trend differences between extreme points under the same key growth conditions specifically includes:

[0023] According to the extreme value points in the growth data, the time series difference between two different key growth conditions is obtained, and a difference time period is constructed according to the minimum time series difference to obtain a plurality of the difference time periods;

[0024] Obtaining a time series difference between an extreme point of one of the key growth conditions and an extreme point of another key growth condition within each of the difference time periods;

[0025] The degree of commonality between the two different key growth conditions is determined based on the time series difference between the extreme points of the two key growth conditions and the potential impact trend difference between the extreme points under the same key growth condition.

[0026] In combination with the third implementation manner of the first aspect, in the fourth implementation manner of the first aspect, determining the degree of commonality between the two different key growth conditions according to the time sequence difference between the extreme points of the two key growth conditions and the potential impact trend difference between the extreme points under the same key growth conditions specifically includes:

[0027] Respectively obtain the time series differences from the maximum point and the minimum point of the key growth condition of the current item to the maximum point of another key growth condition in each difference time period, and obtain a first maximum time series difference and a first minimum time series difference;

[0028] Obtain the potential impact trend differences between the maximum value points of the key growth conditions described in the current item;

[0029] According to the first maximum time series difference, the first minimum maximum time series difference and the potential impact trend difference of each maximum point in the key growth condition of the current item, a first commonality degree between different key growth conditions is obtained; the first commonality degree is the commonality degree between the maximum points of different key growth conditions;

[0030] Respectively obtain the time series difference from the maximum point and the minimum point of the key growth condition of the current item to the minimum point of another key growth condition in each difference time period, and obtain a second maximum time series difference and a second minimum difference;

[0031] Obtain the potential impact trend differences between the various minimum value points of the key growth conditions described in the current item;

[0032] According to the second maximum time series difference, the second minimum time series difference and the potential impact trend difference of each minimum point in the key growth condition of the current item, a second commonality degree between different key growth conditions is obtained; the second commonality degree is the commonality degree between the minimum points of different key growth conditions;

[0033] Based on the first commonality degree and the second commonality degree, the commonality degree between the two different key growth conditions is obtained.

[0034] In combination with the first aspect, in a fifth implementation of the first aspect, determining the similarity between different batches of the agricultural products to be traced based on the commonality degree and the quality score difference specifically includes:

[0035] Determining a similarity weight of each of the key growth conditions according to the degree of commonality;

[0036] Determine the extreme value point where the potential impact trend exceeds 0 for each batch of the agricultural products to be traced under the same key growth conditions, and determine the minimum time series difference and potential impact trend difference between the extreme value points where the potential impact trend exceeds 0 for different batches;

[0037] The similarities between different batches of the agricultural products to be traced are determined based on the minimum time series difference of the similarity weight, the potential impact trend difference and the quality score difference.

[0038] In combination with the first aspect, in a sixth implementation of the first aspect, the quality score is obtained by a trained quality evaluation model, and the quality evaluation model is used to input key features of the agricultural products to be traced and scores of the key features and output the quality score of the agricultural products to be traced.

[0039] In combination with the first aspect, in a seventh implementation of the first aspect, the quality evaluation model is based on key features of the sample agricultural products and the scores of the key features, and is trained by a gradient descent method.

[0040] In combination with the sixth implementation of the first aspect, in the eighth implementation of the first aspect, the quality assessment model is trained by the following steps:

[0041] Obtain key characteristics of sampled agricultural product quality;

[0042] Determine the weight of each of the above key characteristics and determine the score of the sample agricultural products in each key characteristic;

[0043] The key characteristics of sample agricultural products, the weights of key characteristics, and the scores of key characteristics are used as input data, input into the neural network model and trained using the gradient descent method to obtain a quality evaluation model for outputting the quality scores of agricultural products.

[0044] According to the second aspect, an embodiment of the present invention provides a data information traceability system for agricultural products, the system comprising:

[0045] A quality scoring module, used to obtain the quality score of each batch of agricultural products to be traced, and to determine the difference in quality scores between different batches of agricultural products to be traced;

[0046] A potential impact module is used to obtain the growth data of the agricultural product to be traced under various key growth conditions, and determine the potential impact trend of each extreme point according to the local abnormal data in the growth data; the potential impact trend has a mapping relationship with the quality of the agricultural product;

[0047] A commonality determination module, for determining the degree of commonality between different key growth conditions according to the time sequence differences between the extreme points of the agricultural products to be traced under different key growth conditions and the potential impact trend differences between the extreme points under the same key growth conditions;

[0048] A similarity determination module, used to determine the similarity between different batches of the agricultural products to be traced based on the commonality and the quality score differences;

[0049] The classification and chain-uploading module is used to classify different batches of the agricultural products to be traced according to the similarity, assign the same traceability information to the agricultural products to be traced of the same category, and upload the data of the agricultural products to be traced with the traceability information to the chain.

[0050] The present invention has the following beneficial effects: by determining the potential influence trend of each extreme point in the growth data, and then analyzing the local excess or missing amount of the key growth conditions in the external conditions affecting the quality of agricultural products in the historical data, judging its local abnormal state to obtain the potential influence trend of any extreme point, avoiding the use of historical data of a single time period to judge whether the crops are in dynamic equilibrium and the inaccurate impact on the quality, improving the representativeness and accuracy of the characteristics of different agricultural products, and at the same time, according to the time series difference between the extreme points of the agricultural products to be traced under different key growth conditions and the potential influence trend difference between the extreme points under the same key growth conditions, determining the degree of commonality between different key growth conditions, and obtaining the data change commonality of the key growth conditions by comparing the change consistency of different key growth conditions in the historical data, avoiding the weakening of the impact of short-term abnormalities in the historical data caused by the long growth time span of the crops, so that the specific differences of the final agricultural products of different batches are small, improving the representativeness and accuracy of the final classification, and ensuring the accuracy and credibility of the traceability method and system. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0052] Figure 1 A schematic diagram of a flow chart of a data information tracing method for agricultural products provided by an embodiment of the present invention;

[0053] Figure 2 A flow chart of step S20 in a data information traceability method for agricultural products provided by an embodiment of the present invention;

[0054] Figure 3 A schematic flow chart of step S23 in a data information traceability method for agricultural products provided by an embodiment of the present invention;

[0055] Figure 4 A schematic flow chart of step S30 in a data information traceability method for agricultural products provided by an embodiment of the present invention;

[0056] Figure 5 A schematic flow chart of step S40 in a data information traceability method for agricultural products provided by an embodiment of the present invention;

[0057] Figure 6A schematic diagram of the structure of a data information traceability system for agricultural products provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0058] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following is a detailed description of a data information traceability method and system for agricultural products proposed by the present invention, its specific implementation method, structure, features and effects, in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0059] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0060] The introduction of blockchain technology and smart contracts can effectively solve the problems of information opacity, data transmission difficulties and easy tampering in the agricultural product traceability system, and realize efficient, accurate and reliable traceability of the entire process of agricultural product production. Specifically, various data on the growth environment and production process of agricultural products and agricultural product-related indicators can be obtained through technologies such as the Internet of Things, and then these data can be pre-processed and uploaded to the blockchain network through the communication protocol. Then, the data in the blockchain network can be verified using smart contract technology to ensure the accuracy of data transmission. Finally, the data in the production process can be uploaded to the chain to ensure the accuracy and credibility of the traceability method and system.

[0061] However, the growth process of crops takes a relatively long time, and the impact of various external factors on the quality of crops cannot be simply evaluated. At the same time, since crops themselves have different requirements for external conditions, and for any crop, the external conditions need to be kept in a balanced state within the prescribed range, the surge in a single type of external factor will cause problems with the quality of the crops, and even cause the entire batch of crops to die directly. If you simply collect various data generated by external conditions of agricultural products and upload the data to the chain, the accuracy and credibility of the information traceability of agricultural products based on these data cannot be guaranteed.

[0062] Correspondingly, the data information of agricultural products corresponding to crops covers various key data from production to sales, which are crucial to ensuring the quality, safety and traceability of agricultural products. At the same time, since agricultural products are planted in different regions during the planting process, the actual environmental requirements and human intervention requirements (such as fertilizer types and fertilizer weight) for different agricultural products in different regions are different. The quality of agricultural products produced under different external conditions has certain differences, and the subsequent processing process has a relatively standardized processing method. Therefore, for the agricultural products that finally appear on the market, their final quality is closely related to the growth conditions in their historical growth process, so data traceability can be carried out for various key growth conditions in the historical production process of different agricultural products.

[0063] Therefore, how to analyze the impact of short-term changes in various external conditions and the overall balance between them on agricultural products, and then better trace the information of agricultural products is an issue that needs to be urgently addressed in the industry.

[0064] In order to solve the above problems, a data information traceability method and system for agricultural products are provided in this specification. The specific scheme of the data information traceability method and system for agricultural products provided by the present invention is described in detail below with reference to the accompanying drawings.

[0065] See also Figure 1 , which shows a flow chart of a data information traceability method for agricultural products provided by an embodiment of the present invention. The method may include:

[0066] S10. Obtain the quality score of each batch of agricultural products to be traced, and determine the difference in quality scores between different batches of agricultural products to be traced.

[0067] In this embodiment, the above-mentioned quality score is obtained by a trained quality evaluation model, that is, a trained quality evaluation model is used to perform classification network training on the actual quality of the batch of agricultural products corresponding to each region, and the quality score value of the batch of agricultural products is output. The quality evaluation model is a currently known neural network model for classification. There is no restriction on the specific form of the neural network model, and it is only necessary to ensure that the quality score value of the agricultural products can be obtained.

[0068] In this embodiment, agricultural products corresponding to a planting area are harvested in batches, and key production conditions for each preset collection interval in the historical period between sowing of agricultural products in the planting area and batch receipt are obtained and stored to form offline historical data to facilitate subsequent operations.

[0069] Specifically, the key production conditions of agricultural products in the historical period are collected to provide data support for subsequent quality evaluation. The preset collection interval refers to the time period for data collection, which can be 1h, 6h, 24h, one week, etc., without limitation. In addition, agricultural product data under different circumstances such as special time periods, such as germination period and maturity period, can be collected to determine the key production conditions in various situations, so as to provide better and more accurate quality evaluation.

[0070] In this embodiment, the above-mentioned key production conditions include at least: total light intensity, temperature, precipitation, soil pH value, nutrient content and other multi-dimensional data.

[0071] In this method, before processing agricultural products, preliminary quality inspections are conducted on agricultural products, and the actual quality of the corresponding batches of agricultural products in each region is pre-evaluated. It should be noted that, considering the large number of agricultural products in a single batch, in order to avoid the impact of a single case on the entire batch of agricultural products, a sampling of all agricultural products in a batch is screened and evaluated for quality. For example, 20% of the total number of agricultural products in a single batch is selected as sample agricultural products, and then the sample agricultural products are quality inspected and the actual quality of the sample agricultural products is used as the actual quality corresponding to the batch of agricultural products.

[0072] Preferably, the training process of the quality assessment model is:

[0073] A10. Obtain key features that affect the quality of sample agricultural products in a single batch of agricultural products. In this embodiment, the key features mentioned above are features of interest that affect the agricultural products. The key features may at least include: color, size, completeness, taste and other multi-dimensional data corresponding to the agricultural products.

[0074] A20. Determine the weight of each of the above key characteristics and determine the score of the sample agricultural products in each key characteristic. It can be understood that the higher the score of the sample agricultural products in a key characteristic, the higher (more) the characteristic information of the sample agricultural products in this key characteristic.

[0075] In this embodiment, the score of the sample agricultural product on each key feature may be assessed by industry experts, or by consumers, or may be a combination of assessments by industry experts, consumers, and other groups.

[0076] A30. The key features of the sample agricultural products, the weights of the key features, and the scores of the key features are used as input data, input into the neural network model and trained using the gradient descent method to obtain a quality evaluation model for outputting the quality score of the agricultural products. The quality evaluation model can combine the scores of the sample agricultural products on all key features and the weights of each key feature to output the quality score of the sample agricultural products. The value of the agricultural product quality score is between 0 and 1. Similarly, the higher the quality score, the higher the quality of the sample agricultural products, that is, the higher the quality of the corresponding batch of agricultural products, that is, the better the quality of the agricultural products obtained after a long growth process.

[0077] Take the neural network model whose network structure is a five-layer fully connected neural network model and whose loss function is a cross entropy function as an example for explanation. The training ratio set in the historical data during training is 7:3. The neural network model is trained by selecting link data and trained by the gradient descent method until the loss function converges.

[0078] By inputting the key characteristics of each batch of agricultural products to be traced and the scores of the key characteristics into the quality evaluation model, the quality scores of the corresponding batches of agricultural products to be traced and the differences in quality scores between different batches of the agricultural products to be traced can be obtained.

[0079] S20. Obtain growth data of the agricultural products to be traced under various key growth conditions, and determine the potential impact trend of each extreme point in the growth data, wherein there is a mapping relationship between the potential impact trend and the quality of the agricultural products. More specifically, the greater the potential impact trend, the greater the impact on the quality of the agricultural products, and vice versa.

[0080] Since any key growth condition has an almost positive effect on the growth state and quality of crops, when all key growth conditions are within a reasonable range and are in relative dynamic balance, the key growth condition has an almost positive promoting effect on crops, and this promoting effect has a small impact on the quality of agricultural products obtained from agricultural products in a single time period, such as one day, that is, using a single time period, such as one day, to judge whether agricultural products are in dynamic balance and have a positive or negative impact on quality is relatively one-sided. According to the analysis, the impact of a single key data on crops is reflected in how much the actual monitoring data of the condition exceeds the range (upper or lower limit) corresponding to the growth requirements of crops, and the duration of this situation; for example, for precipitation, there was a heavy rainstorm in a short period of time. Although it exceeded the standard range of crop requirements at this time, it had a short duration, so it had little impact on the quality of agricultural products.

[0081] The potential impact trend in step S20 is obtained by analyzing local abnormal data (including local excess and missing amounts) to determine its local abnormal state. This avoids the inaccuracy of using relevant data from a single time period to determine whether agricultural products are in dynamic equilibrium and the impact on quality, thereby improving the representativeness and accuracy of the characteristics of different agricultural products.

[0082] S30. Determine the degree of commonality between different key growth conditions based on the time series differences between extreme points of the agricultural products to be traced under different key growth conditions and the potential impact trend differences between extreme points under the same key growth conditions.

[0083] In this embodiment, considering that the actual production quality of crops depends on the joint influence of multiple key growth conditions, the so-called dynamic balance, that is, the increase of a key growth condition within a certain period of time, theoretically, if it is accompanied by the joint increase of some other key growth conditions with strong correlation, that is, external factors, its impact on crops will be reduced. Since there is an inevitable relationship between some key growth conditions, such as the negative correlation between precipitation and temperature, the negative correlation between precipitation and light, the random fluctuation between soil pH and precipitation, etc., the commonality of any two key growth conditions is first obtained, so as to avoid the weakening of the impact of short-term anomalies of historical data caused by the long growth time span of agricultural products, so that the specific differences of different batches of final agricultural products are relatively small, and the representativeness and accuracy of the final clustering are improved.

[0084] S40. Determine the similarities between different batches of agricultural products to be traced based on the degree of commonality and differences in quality scores.

[0085] For any batch of agricultural products, the higher the commonality (value) between a key growth condition and other key generation conditions, the lower its own impact trend is. This means that the actual numerical fluctuations caused by the current key growth conditions have little impact on the quality of agricultural products. Therefore, different batches of agricultural products can be clustered based on potential impact trends and commonality values, so that the same traceability information, such as traceability codes, can be given to the same type of agricultural products.

[0086] S50. Classify different batches of agricultural products to be traced according to similarities, assign the same traceability information to agricultural products to be traced of the same category, and upload data of agricultural products to be traced with traceability information to the chain.

[0087] For example, based on the kmeans clustering algorithm, the elbow method is used to obtain the optimal k, and the agricultural product batches are clustered based on the above method, and unsupervised learning clustering operations are performed. The data is divided into multiple clusters through the similarity within the data, and the similarity within the cluster is made as large as possible, while the similarity between clusters is made as small as possible. For the agricultural products of the same cluster calculated above, the agricultural products of different clusters are assigned quality score-related information based on the mean of the quality score, and the traceability information of the agricultural products in the same cluster, such as the traceability code, is assigned, and the information is uploaded to the chain.

[0088] The data information traceability method for agricultural products of the present invention determines the potential influence trend of each extreme point in the growth data, and then analyzes the local excess or missing amount of the key growth conditions in the external conditions affecting the quality of agricultural products in the historical data, and judges its local abnormal state to obtain the potential influence trend of any extreme point, thereby avoiding the use of historical data of a single time period to judge whether the crops are in dynamic equilibrium and the inaccurate impact on the quality, and improving the representativeness and accuracy of the characteristics of different agricultural products. At the same time, according to the time series difference between the extreme points of the agricultural products to be traced under different key growth conditions and the potential influence trend difference between the extreme points under the same key growth conditions, the degree of commonality between different key growth conditions is determined, and the data change commonality of the key growth conditions is obtained by comparing the change consistency of different key growth conditions in the historical data, thereby avoiding the weakening of the impact of short-term abnormalities in the historical data caused by the long growth time span of the crops, so that the specific differences of the final agricultural products of different batches are small, and the representativeness and accuracy of the final classification are improved, and the accuracy and credibility of the traceability method and system are guaranteed.

[0089] See also Figure 2 , step S20 specifically includes:

[0090] S21. Obtain growth data of the agricultural product to be traced under each key growth condition, sort the growth data in chronological order, and determine the extreme points in the growth data corresponding to each key growth condition.

[0091] For any key data, the relevant data collected historically are first arranged in chronological order on the coordinate axis, and then based on the a priori knowledge about any current crop and any key data, the theoretical most suitable range of production environment conditions can be used as the ideal adaptation range of the current key data.

[0092] S22. Taking the time period between adjacent extreme value points having the same characteristics as the extreme value time period, a plurality of extreme value time periods and extreme value time periods corresponding to the extreme value points are obtained, wherein the same characteristics are the same maximum value or the same minimum value.

[0093] At this time, for the key growth condition data collected historically, the difference between the maximum and minimum values ​​of the adaptation range is obtained at any time, and the maximum and minimum values ​​are used as subtrahend and minuend respectively, so that the data within the range is always negative and the value outside the range is positive. At this time, two differences can be obtained for the data at any time.

[0094] Next, for any key growth condition, all extreme points (maximum and minimum) in the collected historical data are obtained, and any pair of adjacent maximum points or a pair of adjacent minimum values ​​is taken as an extreme time period, so that the above extreme time periods respectively represent the prominence of the corresponding minimum or maximum values, and the length N of each extreme time period is obtained. i , where i represents the local extreme value time period corresponding to the i-th extreme value point (maximum or minimum).

[0095] S23. According to the correspondence between the extreme time period of the extreme point and the suitable range of the corresponding key growth condition, the potential influence trend of the extreme point is obtained, wherein each key growth condition has a corresponding suitable range.

[0096] See also Figure 3 , taking a pair of adjacent minimum values ​​as an extreme value time period as an example, step S23 specifically includes:

[0097] S231. Determine the extreme value time period where the extreme value point is located.

[0098] S232. Obtain a first quantity and a second quantity within the extreme value time period, wherein the first quantity is the total number of data points within the extreme value time period that are lower than the minimum value of the suitable range corresponding to the key growth condition, and the second quantity is the total number of data points within the extreme value time period that are higher than the maximum value of the suitable range.

[0099] Since the extreme time period N between the two minimum values i There may be data points below the minimum value of the appropriate range, and the extreme value time period N is obtained. i The total number of data points M below the minimum value of the suitable range min That is, the first quantity, and at the same time obtain the extreme value time period N i The total number of data points that exceed the maximum value of the suitable range M max That is the second quantity.

[0100] S233, determine the extreme value time period N i The maximum value ΔT t max .

[0101] S234, determine the extreme value time period N iThe mean value of the excess of data points outside the appropriate range

[0102] S235. Obtain the potential influence trend of the extreme value point according to the first quantity, the second quantity, the total number of data points in the extreme value time period, the maximum value of the extreme value time period, and the mean value of the excess amount.

[0103] In this way, the potential impact trend of the key growing conditions on the current type of agricultural products can be obtained.

[0104]

[0105] Among them, P i Indicates potential impact trend; M min Indicates the extreme value time period N i The total number of data points below the minimum value of the suitable range; M max Indicates the extreme value time period N i The total number of data points that exceed the maximum value of the suitable range; N i Indicates the extreme value time period N i The total number of internal data points, that is, N i -M min Indicates the extreme value time period N i The total number of data points except those exceeding the minimum range; Indicates the extreme value time period N i The maximum value of Indicates the extreme value time period N i The mean of the excess of data points outside the appropriate range; That is, the extreme time period N i The proportion of data points exceeding the maximum value of the range. The larger the value, the longer the possible negative impact on the current key growth conditions will last, and the greater the potential impact trend. At the same time, for the data points exceeding the maximum value of the range, the larger the mean and the maximum value point itself, the greater the magnitude of the excess and the greater the impact trend. Therefore, multiply the above two together to get P i The larger the value, the greater the impact on the trend.

[0106] It should be noted that the calculation principle of the potential impact trend of the minimum point is the same as above. For data points within the normal range, due to M max is 0, so the potential impact trend is 0.

[0107] Traversing all the maximum and minimum points of a key growth condition, each extreme point or the extreme time period corresponding to the extreme value can correspond to a potential impact trend. After the above processing, each extreme point corresponding to any batch of agricultural products can be obtained, and at the same time, the potential impact trend of these extreme points can also be obtained.

[0108] See also Figure 4 , step S30 specifically includes:

[0109] S31. According to the extreme value points in the growth data, the time series difference between two different key growth conditions is obtained, and a difference time period is constructed according to the minimum time series difference to obtain a plurality of difference time periods.

[0110] For each key growth condition, several time periods between the maximum and minimum values ​​can be obtained. At this time, for any extreme point of any key growth condition, there is a timing difference (value) between other key growth conditions and the extreme point, and the extreme point with the smallest timing difference between other key growth conditions and the extreme point can also be obtained.

[0111] A new time period, namely the difference time period, is constructed by taking the extreme points of the above-mentioned current key growth conditions and other external conditions with the smallest time series difference as examples.

[0112] S32: Obtain the time series difference between the extreme value point of one of the key growth conditions and the extreme value point of another key growth condition in each of the difference time periods.

[0113] S33. Determine the degree of commonality between the two different key growth conditions based on the time series difference between the extreme points of the two key growth conditions and the potential impact trend difference between the extreme points under the same key growth condition.

[0114] More specifically, step S33 specifically includes:

[0115] S331, traverse all the maximum points of the current key growth condition, respectively obtain the maximum point and minimum point of another key growth condition in different difference time periods corresponding to the maximum point, and then respectively calculate the difference in the numerical change of the time series difference between the maximum point and the minimum point of the other key growth condition and the maximum point of the current key growth condition, and obtain Δt i,max and Δt i,min , where Δt i,max represents the first maximum timing difference, Δt i,min Indicates the first minimum timing difference.

[0116] S332. Obtain the potential impact trend differences between the maximum points of the current key growth conditions.

[0117] S333. According to the first maximum time series difference, the first minimum maximum time series difference and the potential impact trend difference of each maximum point in the current key growth condition, the first commonality degree between different key growth conditions is obtained. The first commonality degree is the commonality degree between the maximum points of different key growth conditions:

[0118]

[0119] Among them, E x m ,y ax Indicates the commonality between the current key growth condition x and another key growth condition y about the maximum value point, that is, the first commonality; n x,max Indicates the total number of maximum points of the current key growth condition x; Δt i,max The difference in the numerical change of the first maximum time series difference, that is, the time series difference between the maximum point of another key growth condition y and the maximum point of the current key growth condition x, Δt i,min The difference in the numerical change of the first minimum time series difference, that is, the time series difference between the minimum point of another key growth condition y and the maximum point of the current key growth condition x, Δt i,max -Δt j,max , Δt i,min -Δt j,min The smaller the value of is, the more similar the regularity of the numerical changes between the maximum points of the two key growth conditions is. By using the Euclidean norm, we can get P i represents the potential impact trend difference of the maximum point i of the previous key growth condition x, P j represents the potential impact trend difference of the maximum point j of the previous key growth condition x, |P i -P j | represents the potential impact trend difference between the maximum points i and j of any current key growth condition x. The smaller the value, the higher the similarity between the key growth conditions, the more similar the numerical change law of the above calculation, the greater the reference weight for the judgment of common characteristics, and the latter's maximum regularity is weighted.

[0120] Traverse and accumulate all the maximum points, thus we get The smaller the value, the higher the commonality between the current key growth condition and other key growth conditions.

[0121] S334, traverse all minimum points of the current key growth condition, obtain the timing differences from the maximum point and minimum point of the key growth condition of the current item to the minimum point of another key growth condition in each difference time period, and obtain the second maximum timing difference and the second minimum difference.

[0122] S335. Obtain the potential impact trend differences between the various minimum points of the key growth conditions described in the current item.

[0123] S336. According to the second maximum time series difference, the second minimum time series difference and the potential impact trend difference of each minimum point in the current key growth condition, the second commonality degree between different key growth conditions is obtained; the second commonality degree is the commonality degree between the minimum points between different key growth conditions.

[0124] Similarly, for the minimum point, we get It indicates the commonality between the current key growth condition x and another key growth condition y about the minimum value point. The calculation process is not described here.

[0125] S337: Based on the first commonality degree and the second commonality degree, the commonality degree between the two different key growth conditions is obtained. At this time, for any two current key growth conditions, the actual commonality value of the two can be obtained:

[0126]

[0127] Among them, E x,y Indicates the commonality value between the current key growth condition x and another key growth condition y.

[0128] By traversing all the key growth conditions, the corresponding commonality value can be obtained. The larger the commonality value, the more regular the changes in any two key growth conditions in the historical data are, which also means that the current batch of agricultural products is more in line with the actual crop growth needs.

[0129] See also Figure 5 , step S40 specifically includes:

[0130] S41. Determine the similarity weight of each key growth condition based on the degree of commonality.

[0131] For any batch of agricultural products, the higher the commonality value between a key growth condition and other key generation conditions, the lower its own impact trend is. This means that the actual numerical fluctuations caused by the current key growth conditions have little impact on the quality of agricultural products. Therefore, different batches of agricultural products can be clustered based on potential impact trends and commonality values, so that the same traceability information, such as traceability codes, can be given to the same type of agricultural products.

[0132] In this embodiment, the similarity weights of the commonality values ​​of the key growth conditions are also determined:

[0133]

[0134] Among them, μ x represents the similarity weight of the key growth condition x; E x,y Indicates the commonality value between the current key growth condition x and another key growth condition y; n x,max Indicates the total number of data points of the current key growth condition x.

[0135] S42. Determine the extreme points where the potential impact trend exceeds 0 for each batch of agricultural products to be traced under the same key growth conditions, and determine the minimum time series difference and potential impact trend difference between the extreme points where the potential impact trend exceeds 0 for different batches.

[0136] S43. Determine the similarity between different batches of agricultural products to be traced based on the minimum time series difference of similarity weight, the difference in potential impact trend and the difference in quality score.

[0137] Therefore, by expressing the changes in the historical external condition data of the current batch of agricultural products in the form of a curve, for all batches of agricultural products, by comparing the curves of the same external condition data, the extreme points with a potential impact trend greater than 0 are compared with the minimum time series difference between the extreme points of other batches of agricultural products, and comparing the above data points of different batches one by one, so that for any two batches of agricultural products with the same external growth conditions, G can be obtained. x The data corresponds to the combination and records the minimum timing difference Δd r The difference between the potential impact trend ΔP r At the same time, obtain the commonality (value) differences between the same key growth conditions as a whole, as well as the difference in quality scores ΔQ in the final quality inspection process a,b .

[0138] Thus, the similarity between any two different batches of agricultural product data is obtained:

[0139]

[0140] Where f represents the similarity between a certain agricultural product of batch a and batch b; ΔQ a,b represents the difference in quality score of a certain agricultural product between batch a and batch b; μ a,x represents the similarity weight of a certain agricultural product of batch a under the key growth condition x; μ b,x G represents the similarity weight of a certain agricultural product of batch b under the key growth condition x; x Indicates the corresponding combination of agricultural product data of batch a and batch b; Δd r Indicates the minimum time series difference between a certain agricultural product data of batch a and batch b; ΔP rRepresents the difference between the potential impact trends of an extreme point of a certain agricultural product data of batch a and batch b; for a certain agricultural product of batch a and batch b, which contains a total of m key growth conditions, for the same key growth condition x, when the potential impact trend of the two batches of agricultural products is greater than 0 (extreme point) compared with the minimum time series difference Δd between the other batches of data points r The smaller the difference, the smaller the difference ΔP between the potential impact trends of the two data points. r The smaller it is, the more data points with potential influence trend greater than 0 are traversed, and the The smaller the value, the more similar the variation characteristics of the two batches under the current key growth condition x.

[0141] At the same time, for any key growth condition, the smaller the difference in commonality value between two batches of agricultural products, the lower the specificity of the change caused by the condition itself, the higher the overall similarity between the two key growth conditions, and the greater the similarity judgment weight of the current key growth conditions. By traversing m key growth conditions and combining the quality score differences, the similarity f between a certain agricultural product of batch a and batch b is obtained. The greater the similarity, the more similar the changes in the external conditions of the two batches of agricultural products are.

[0142] The following is a description of a data information traceability system for agricultural products provided by an embodiment of the present invention. The data information traceability system for agricultural products described below and the data information traceability method for agricultural products described above can be referenced to each other.

[0143] See also Figure 6 , which shows a schematic diagram of the structure of a data information traceability system for agricultural products provided by an embodiment of the present invention. The system may include:

[0144] The quality scoring module 10 is used to obtain the quality score of each batch of agricultural products to be traced, and to determine the difference in quality scores between different batches of agricultural products to be traced.

[0145] In this embodiment, the above-mentioned quality score is obtained by a trained quality evaluation model, that is, a trained quality evaluation model is used to perform classification network training on the actual quality of the batch of agricultural products corresponding to each region, and the quality score value of the batch of agricultural products is output. The quality evaluation model is a currently known neural network model for classification. There is no restriction on the specific form of the neural network model, and it is only necessary to ensure that the quality score value of the agricultural products can be obtained.

[0146] In this embodiment, agricultural products corresponding to a planting area are harvested in batches, and key production conditions for each preset collection interval in the historical period between sowing of agricultural products in the planting area and batch receipt are obtained and stored to form offline historical data to facilitate subsequent operations.

[0147] Specifically, the key production conditions of agricultural products in the historical period are collected to provide data support for subsequent quality evaluation. The preset collection interval refers to the time period for data collection, which can be 1h, 6h, 24h, one week, etc., without limitation. In addition, agricultural product data under different circumstances such as special time periods, such as germination period and maturity period, can be collected to determine the key production conditions in various situations, so as to provide better and more accurate quality evaluation.

[0148] In this embodiment, the above-mentioned key production conditions include at least: total light intensity, temperature, precipitation, soil pH value, nutrient content and other multi-dimensional data.

[0149] In this system, agricultural products are preliminarily inspected for quality before being processed, and the actual quality of the corresponding batch of agricultural products in each region is preliminarily evaluated. It should be noted that, considering the large number of agricultural products in a single batch, in order to avoid the impact of a single case on the entire batch of agricultural products, all agricultural products in a batch are sampled for quality screening and evaluation, for example, 20% of the total number of agricultural products in a single batch are selected as sample agricultural products, and then the sample agricultural products are quality inspected and the actual quality of the sample agricultural products is used as the actual quality corresponding to the batch of agricultural products.

[0150] By inputting the key characteristics of each batch of agricultural products to be traced and the scores of the key characteristics into the quality evaluation model, the quality scores of the corresponding batches of agricultural products to be traced and the differences in quality scores between different batches of the agricultural products to be traced can be obtained.

[0151] The potential impact module 20 is used to obtain the growth data of the agricultural products to be traced under various key growth conditions, and determine the potential impact trend of each extreme point in the growth data, wherein there is a mapping relationship between the potential impact trend and the quality of the agricultural products. More specifically, the greater the potential impact trend, the greater the impact on the quality of the agricultural products, and vice versa.

[0152] Since any key growth condition has an almost positive effect on the growth state and quality of crops, when all key growth conditions are within a reasonable range and are in relative dynamic balance, the key growth condition has an almost positive promoting effect on crops, and this promoting effect has a small impact on the quality of agricultural products obtained from agricultural products in a single time period, such as one day, that is, using a single time period, such as one day, to judge whether agricultural products are in dynamic balance and have a positive or negative impact on quality is relatively one-sided. According to the analysis, the impact of a single key data on crops is reflected in how much the actual monitoring data of the condition exceeds the range (upper or lower limit) corresponding to the growth requirements of crops, and the duration of this situation; for example, for precipitation, there was a heavy rainstorm in a short period of time. Although it exceeded the standard range of crop requirements at this time, it had a short duration, so it had little impact on the quality of agricultural products.

[0153] Among them, the potential impact trend is obtained by analyzing local abnormal data (including local excess and missing amounts) to determine its local abnormal state. This can avoid using relevant data from a single time period to determine whether agricultural products are in dynamic equilibrium and the inaccurate impact on quality, and improve the representativeness and accuracy of the characteristics of different agricultural products.

[0154] The commonality determination module 30 is used to determine the degree of commonality between different key growth conditions based on the time series differences between the extreme points of the agricultural products to be traced under different key growth conditions and the potential impact trend differences between the extreme points under the same key growth conditions.

[0155] In this embodiment, considering that the actual production quality of crops depends on the joint influence of multiple key growth conditions, the so-called dynamic balance, that is, the increase of a key growth condition within a certain period of time, theoretically, if it is accompanied by the joint increase of some other key growth conditions with strong correlation, that is, external factors, its impact on crops will be reduced. Since there is an inevitable relationship between some key growth conditions, such as the negative correlation between precipitation and temperature, the negative correlation between precipitation and light, the random fluctuation between soil pH and precipitation, etc., the commonality of any two key growth conditions is first obtained, so as to avoid the weakening of the impact of short-term anomalies of historical data caused by the long growth time span of agricultural products, so that the specific differences of different batches of final agricultural products are relatively small, and the representativeness and accuracy of the final clustering are improved.

[0156] The similarity determination module 40 is used to determine the similarities between different batches of agricultural products to be traced based on the degree of commonality and the difference in quality scores.

[0157] For any batch of agricultural products, the higher the commonality value between a key growth condition and other key generation conditions, the lower its own impact trend is. This means that the actual numerical fluctuations caused by the current key growth conditions have little impact on the quality of agricultural products. Therefore, different batches of agricultural products can be clustered based on potential impact trends and commonality values, so that the same traceability information, such as traceability codes, can be given to the same type of agricultural products.

[0158] The classification and chain-uploading module 50 is used to classify different batches of agricultural products to be traced according to similarities, and to upload data of agricultural products to be traced of the same category to the chain.

[0159] For example, based on the kmeans clustering algorithm, the elbow method is used to obtain the optimal k, and the agricultural product batches are clustered based on the above method, and unsupervised learning clustering operations are performed. The data is divided into multiple clusters through the similarity within the data, and the similarity within the cluster is made as large as possible, while the similarity between clusters is made as small as possible. For the agricultural products of the same cluster calculated above, the agricultural products of different clusters are assigned quality score-related information based on the mean of the quality score, and the traceability information of the agricultural products in the same cluster, such as the traceability code, is assigned, and the information is uploaded to the chain.

[0160] The data information traceability system for agricultural products of the present invention determines the potential influence trend of each extreme point in the growth data, and then analyzes the local excess or missing amount of the key growth conditions in the external conditions affecting the quality of agricultural products in the historical data, and judges its local abnormal state to obtain the potential influence trend of any extreme point, thereby avoiding the use of historical data of a single time period to judge whether the crops are in dynamic equilibrium and the inaccurate impact on the quality, and improving the representativeness and accuracy of the characteristics of different agricultural products. At the same time, according to the time series difference between the extreme points of the agricultural products to be traced under different key growth conditions and the potential influence trend difference between the extreme points under the same key growth conditions, the degree of commonality between different key growth conditions is determined, and the data change commonality of the key growth conditions is obtained by comparing the change consistency of different key growth conditions in the historical data, thereby avoiding the weakening of the impact of the short-term abnormality of the historical data caused by the long growth time span of the crops, so that the specific differences of the final agricultural products of different batches are small, and the representativeness and accuracy of the final classification are improved, and the accuracy and credibility of the traceability method and system are guaranteed.

[0161] It should be noted that the sequence of the above embodiments of the present invention is only for description and does not represent the advantages and disadvantages of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0162] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referenced to each other, and each embodiment focuses on the differences from other embodiments.

Claims

1. A data information traceability method for agricultural products, characterized in that: The method comprises: Obtaining a quality score for each batch of agricultural products to be traced, and determining the difference in quality scores between different batches of agricultural products to be traced; Acquire growth data of the agricultural product to be traced under various key growth conditions, and determine the potential impact trend of each extreme point based on local abnormal data in the growth data; the potential impact trend is mapped to the quality of the agricultural product; Determine the degree of commonality between the different key growth conditions based on the temporal differences between the extreme points of the agricultural product to be traced under different key growth conditions and the potential impact trend differences between the extreme points under the same key growth conditions; Determining the similarity between different batches of the agricultural products to be traced based on the degree of commonality and the difference in the quality scores; The agricultural products to be traced in different batches are classified according to the similarity, the agricultural products to be traced in the same category are given the same traceability information, and the data of the agricultural products to be traced with traceability information are uploaded to the chain.

2. The agricultural product data information traceability method according to claim 1, characterized in that: The step of obtaining growth data of the agricultural product to be traced under various key growth conditions and determining the potential impact trend of each extreme point based on local abnormal data in the growth data specifically includes: Obtaining growth data of the agricultural product to be traced under each of the key growth conditions, sorting the growth data in chronological order, and determining extreme points in the growth data corresponding to each of the key growth conditions; The time period between adjacent extreme value points having the same characteristics is taken as the extreme value time period, and a plurality of extreme value time periods and the extreme value time periods corresponding to the extreme value points are obtained; the same characteristics are the same maximum value or the same minimum value; According to the correspondence between the extreme time period of the extreme point and the appropriate range of the corresponding key growth condition, the potential influence trend of the extreme point is obtained; each phase of the key growth condition has the corresponding appropriate range.

3. The agricultural product data information traceability method according to claim 2, characterized in that: The step of obtaining the potential influence trend of the extreme point based on the correspondence between the extreme time period of the extreme point and the appropriate range of the corresponding key growth condition specifically includes: Determining the extreme value time period in which the extreme value point is located; Obtaining a first quantity and a second quantity within the extreme value time period; the first quantity being the total number of data points within the extreme value time period that are lower than a minimum value of a suitable range corresponding to the key growth condition; and the second quantity being the total number of data points within the extreme value time period that are higher than a maximum value of the suitable range; determining a maximum value of the extreme value time period; determining a mean of the excesses of the data points that exceed the suitable range; The potential influence trend of the extreme value point is obtained according to the first quantity, the second quantity, the total number of data points in the extreme value time period, the maximum value of the extreme value time period and the mean value of the excess amount.

4. The agricultural product data information traceability method according to claim 1, characterized in that: Determining the degree of commonality between different key growth conditions based on the temporal differences between extreme points of the agricultural product to be traced under different key growth conditions and the potential impact trend differences between extreme points under the same key growth conditions specifically includes: Obtaining the time series difference between two different key growth conditions according to the extreme value points in the growth data, and constructing a difference time period according to the minimum time series difference to obtain a plurality of the difference time periods; Obtaining a time series difference between an extreme point of one of the key growth conditions and an extreme point of another key growth condition within each of the difference time periods; The degree of commonality between the two different key growth conditions is determined based on the time sequence difference between the extreme points of the two key growth conditions and the potential impact trend difference between the extreme points under the same key growth condition.

5. The agricultural product data information traceability method according to claim 4, characterized in that: Determining the degree of commonality between the two different key growth conditions based on the time sequence difference between the extreme points of the two key growth conditions and the potential impact trend difference between the extreme points under the same key growth condition specifically includes: Respectively obtaining the time series differences between the maximum point and the minimum point of the key growth condition of the current item and the maximum point of another key growth condition within each difference time period to obtain a first maximum time series difference and a first minimum time series difference; Obtain the potential impact trend differences between the maximum value points of the key growth conditions described in the current item; Obtaining a first degree of commonality between different key growth conditions based on the first maximum value time series difference, the first minimum maximum value time series difference, and the potential impact trend difference of each maximum value point in the key growth condition of the current item; the first degree of commonality being the degree of commonality between the maximum value points of different key growth conditions; Respectively obtaining the time series differences between the maximum point and the minimum point of the key growth condition of the current item and the minimum point of the key growth condition of the other item within each difference time period to obtain a second maximum time series difference and a second minimum difference; Obtain the potential impact trend differences between the minimum points of the key growth conditions described in the current item; Obtaining a second degree of commonality between different key growth conditions based on the second maximum time series difference, the second minimum time series difference, and the potential impact trend difference of each minimum point in the key growth condition of the current item; the second degree of commonality being the degree of commonality between the minimum points of different key growth conditions; Based on the first degree of commonality and the second degree of commonality, the degree of commonality between the two different key growth conditions is obtained.

6. The agricultural product data information traceability method according to claim 1, characterized in that: Determining the similarity between different batches of the agricultural products to be traced based on the degree of commonality and the difference in quality scores specifically includes: determining a similarity weight for each of the key growth conditions based on the degree of commonality; Determine the extreme point at which the potential impact trend exceeds 0 for each batch of the agricultural products to be traced under the same key growth conditions, and determine the minimum time series difference and potential impact trend difference between the extreme points at which the potential impact trends exceed 0 for different batches; The similarity between different batches of the agricultural products to be traced is determined based on the minimum time series difference of the similarity weight, the potential impact trend difference and the quality score difference.

7. The agricultural product data information traceability method according to claim 1, characterized in that: The quality score is obtained by a trained quality evaluation model, which is used to input key features of the agricultural products to be traced and scores of the key features and output the quality score of the agricultural products to be traced.

8. The agricultural product data information traceability method according to claim 7, characterized in that: The quality evaluation model is based on the key features of sample agricultural products and the scores of the key features, and is trained by a gradient descent method.

9. The agricultural product data information traceability method according to claim 7, characterized in that: The quality evaluation model is trained by the following steps: Obtain key characteristics of sample agricultural product quality; Determine the weight of each key characteristic and the score of the sample agricultural products on each key characteristic; The key characteristics of sample agricultural products, the weights of key characteristics, and the scores of key characteristics are used as input data, input into the neural network model and trained using the gradient descent method to obtain a quality evaluation model for outputting the quality scores of agricultural products.

10. A data information traceability system for agricultural products, characterized by: The system comprises: A quality scoring module is used to obtain the quality score of each batch of agricultural products to be traced and to determine the difference in quality scores between different batches of agricultural products to be traced; A potential impact module is used to obtain growth data of the agricultural product to be traced under various key growth conditions, and determine the potential impact trend of each extreme point based on local abnormal data in the growth data; the potential impact trend is mapped to the quality of the agricultural product; A commonality determination module is configured to determine the degree of commonality between different key growth conditions based on the time sequence differences between the extreme points of the agricultural product to be traced under different key growth conditions and the potential impact trend differences between the extreme points under the same key growth conditions; A similarity determination module, configured to determine the similarity between different batches of the agricultural products to be traced based on the degree of commonality and the difference in quality scores; The classification and chain-uploading module is used to classify different batches of agricultural products to be traced according to the similarity, assign the same traceability information to the agricultural products to be traced of the same category, and upload the data of the agricultural products to be traced with traceability information to the chain.

Citation Information

Patent Citations

  • Secure agricultural product quality safety tracing chain correlation analysis system

    CN110880151A

  • Product tracing method and system based on block chain technology

    CN117196659A

  • Agricultural product traceability data management method and system based on block chain

    CN118093972A

Cited By

  • Anti-counterfeiting traceability method for gravure water-based ink printing production information

    CN119624484A

  • System and method for identifying and tracing producing areas of agricultural products based on characteristic fingerprints

    CN119807816A