Blockchain-based red onion product traceability system

Abstract

The present application relates to the technical field of block chain, especially to a red onion product traceability system based on block chain, which receives corresponding information sets of each participant through a data input module; a traceability data construction module generates a traceability data chain based on the information sets; a time sequence feature extraction module constitutes a time sequence vector based on the traceability data chain; an abnormal point identification module determines an abnormal propagation point set or a first abnormal propagation point based on the traceability data chain and the time sequence vector; an abnormal point determination module determines an abnormal starting point based on the first abnormal propagation point or a second abnormal propagation point, and determines an abnormal ending point based on the second abnormal propagation point and the abnormal propagation point set; and a feedback adjustment module determines a data problem type based on data blocks of corresponding participants of an abnormal propagation chain, so as to determine an adjustment mode of corresponding participants of the next batch. The present application can significantly improve the authenticity and integrity of the on-chain information by identifying the abnormal propagation point and verifying the data blocks corresponding to the abnormal propagation point.

Description

Technical Field

[0001] This invention relates to the field of blockchain technology, and in particular to a blockchain-based traceability system for onions. Background Technology

[0002] Currently, the agricultural product traceability industry utilizes the immutability, decentralization, and secure encryption characteristics of blockchain to establish digital archives covering the entire chain of agricultural products. However, existing traceability systems lack a closed-loop incentive and constraint mechanism for the quality of data uploaded to the blockchain, which fails to incentivize all participants to report truthfully, easily leading to false reporting and omissions, thereby reducing the authenticity and completeness of the data reported on the blockchain.

[0003] Chinese Patent Application Publication No. CN120296088A discloses a blockchain-based method and device for managing agricultural product traceability data. The related technical solution constructs a consortium blockchain architecture consisting of a management chain and a data chain. When a new node is added to a data chain, a request is sent to the management chain corresponding to this data chain to enable consensus among the nodes of the data chain. Based on the consensus result, the access result of the new node is determined. In response to external data access requests, permission matching is performed on the data access requests, permissions are assigned to the data access requests based on the matching results, and the data of the data chains that are allowed to be accessed by the permissions are displayed.

[0004] However, the above technical solutions can only guarantee that the entered data will not be tampered with, but cannot verify the authenticity of the source data. Furthermore, they do not involve supporting incentive and punishment mechanisms, which cannot prompt all participants to complete the data on the blockchain in a standardized manner, nor can they guarantee the authenticity and integrity of the information on the blockchain. Summary of the Invention

[0005] To address this, the present invention provides a blockchain-based traceability system for onion products, which overcomes the problems in existing technologies that cannot verify the authenticity of data from abnormal sources and cannot prompt all participating parties to truthfully and completely report information sets.

[0006] To achieve the above objectives, this invention provides a blockchain-based traceability system for shallot products, comprising: The data input module is used to receive the corresponding information sets from the participants, including growers, processors, transporters, and sellers. The traceability data construction module is used to write each of the aforementioned information sets into the blockchain in the form of timestamps to generate a traceability data chain; The time-series feature extraction module is used to extract the timestamp differences of each link based on the traceability data chain to form a time-series vector; An anomaly identification module is used to determine an anomaly propagation point set or a first anomaly propagation point based on the source data chain and the time-series vector. If it is an anomaly propagation point set, a time-series offset vector is calculated based on the time-series vector to determine the main anomaly propagation point, and the time-series deviation between the remaining candidate anomaly propagation points that have not been determined as the main anomaly propagation point and the current main anomaly propagation point is calculated to determine whether the remaining candidate anomaly propagation points are independent anomaly propagation points. An anomaly point determination module is used to compare the main anomaly propagation point and the independent anomaly propagation point to determine the second anomaly propagation point, determine the corresponding anomaly starting point based on the first anomaly propagation point or the second anomaly propagation point, and determine the corresponding anomaly ending point based on the second anomaly propagation point and the set of anomaly propagation points. The feedback adjustment module is used to determine the anomaly propagation chain based on the anomaly start point and the anomaly end point, extract the data blocks of the participants corresponding to the anomaly propagation chain from the source tracing data chain, determine the corresponding data problem type based on the corresponding information set in the data block, and determine to adjust the data reporting frequency or data collection accuracy of the next batch of the corresponding participants based on the data problem type. The data problem type includes data omission or data misreporting.

[0007] Furthermore, the data input module is used to receive information sets from growers, processors, transporters, and sellers; The planting information set includes fertilizer application amount and planting density; the processing information set includes allicin content, moisture content and processing time; the transportation information set includes changes in storage and transportation humidity; and the sales information set includes shelf time.

[0008] Furthermore, the traceability data construction module is also used to encapsulate the information sets of each participant into corresponding data blocks, and link each data block sequentially based on the timestamp to form the traceability data chain.

[0009] Furthermore, the temporal feature extraction module includes: The timestamp acquisition submodule is used to extract the timestamps corresponding to the grower, the processor, the transporter, and the seller based on the traceability data chain. The timestamp difference acquisition submodule is used to calculate a first timestamp difference based on the difference between the timestamps corresponding to the processor and the planter. The timestamp difference acquisition submodule is further used to calculate a second timestamp difference based on the difference between the timestamps corresponding to the transporter and the processor; The timestamp difference acquisition submodule is further used to calculate a third timestamp difference based on the difference between the timestamps corresponding to the seller and the transporter; The timestamp total value acquisition submodule is used to calculate the timestamp total value based on the first timestamp difference, the second timestamp difference, and the third timestamp difference; The time-series vector acquisition submodule is used to construct the time-series vector based on the first timestamp difference, the second timestamp difference, the third timestamp difference, and the total timestamp value.

[0010] Furthermore, the anomaly identification module is also used to determine the comparison of the data tracing chain based on the fact that each component of the time-series vector is greater than the corresponding standard link time-consuming vector.

[0011] Furthermore, the anomaly identification module includes: The propagation point identification submodule is used to determine the planting site as the first abnormal propagation point based on the fact that the allicin content is less than the standard allicin content. The propagation point identification submodule is also used to determine the processing party as the first abnormal propagation point based on the processing time being greater than the standard processing time; The propagation point identification submodule is also used to determine the transporter as the first abnormal propagation point based on the fact that the change in storage and transportation humidity is greater than the standard change in storage and transportation humidity. The propagation point identification submodule is also used to determine the seller as the first abnormal propagation point based on the fact that the listing duration is greater than the standard listing duration; Specifically, the standard allicin content is determined based on the amount of fertilizer applied and the planting density; the standard processing time and the standard storage and transportation humidity change are determined based on the moisture content; and the standard shelf-reading time is determined based on the storage and transportation humidity change.

[0012] Furthermore, the anomaly identification module also includes: The propagation point set identification submodule is used to identify several abnormal participants as the abnormal propagation point set based on the occurrence of two or more of the first abnormal propagation points. The propagation point set identification submodule is further used to determine the propagation point corresponding to the largest component in the time offset vector of the abnormal propagation point set as the main abnormal propagation point; The propagation point set identification submodule is further used to identify the candidate abnormal propagation points corresponding to the time deviation degree being greater than the standard time deviation degree as the independent abnormal propagation points; The timing offset vector is obtained based on the absolute difference between each component of the timing vector and the corresponding standard stage time consumption vector, and the timing deviation is obtained based on the absolute difference between the corresponding components in the timing offset vectors of the candidate anomaly propagation point and the main anomaly propagation point.

[0013] Furthermore, the anomaly point determination module is also used to compare the timestamps corresponding to the main anomaly propagation point and the independent anomaly propagation point, and determine the propagation point with the earliest timestamp as the second anomaly propagation point, and determine the second anomaly propagation point as the anomaly starting point.

[0014] Furthermore, the anomaly point determination module is also used to take the first anomaly propagation point as a reference, traverse along the supply chain direction in ascending order of timestamps, and determine that the first anomaly propagation point is both the anomaly start point and the anomaly end point. The anomaly point determination module is also used to take the second anomaly propagation point as a reference, traverse along the supply chain direction in ascending order of timestamps, determine the candidate anomaly propagation point located downstream of the second anomaly propagation point in the set of anomaly propagation points, and determine the candidate anomaly propagation point with the last timestamp as the anomaly endpoint. If there are multiple independent abnormal propagation points in the abnormal propagation point cluster, then each independent abnormal propagation point is used as a reference to traverse downstream along its corresponding supply chain path to determine the abnormal endpoint of each independent abnormal propagation chain.

[0015] Furthermore, the feedback adjustment module includes: The problem type identification submodule is used to obtain the upstream information set of the corresponding participant or re-obtain the information set of the participant based on the fact that the corresponding information set in the data block has not been uploaded or the corresponding information set does not contain complete information. The missing data identification submodule is used to determine the data problem type of the corresponding participant or upstream party as the data missing data based on the upstream information set containing complete information and the re-acquired participant information being incomplete, or the upstream information set not containing the complete information. The error message identification submodule is used to determine the data problem type of the corresponding participant as the data error message based on the corresponding information set in the data block and the fact that the upstream information set contains the complete information. The adjustment method determination submodule is used to determine, based on the data problem type being data underreporting, to increase the data reporting frequency of the corresponding participants in the next batch, or based on the data problem type being data misreporting, to determine to increase the data collection accuracy of the corresponding participants in the next batch.

[0016] Compared with the prior art, the beneficial effect of the present invention is that it receives the corresponding information sets of each participating party; Each information set is used to generate a traceability data chain in the form of timestamps. The timestamp differences between each stage are extracted from the traceability data chain to form a time-series vector. Based on the traceability data chain and the time-series vector, an anomaly propagation point set or the first anomaly propagation point is determined. If it is an anomaly propagation point set, the main anomaly propagation point is determined based on the corresponding time-series vector, and the time-series deviation of the remaining candidate anomaly propagation points that were not determined as the main anomaly propagation point is calculated to determine whether the remaining candidate anomaly propagation points are independent anomaly propagation points. Based on the main anomaly propagation point and independent anomaly propagation points, the second anomaly propagation point is determined. Based on the first or second anomaly propagation point, the corresponding anomaly starting point is determined, and based on the second anomaly propagation point and the anomaly propagation point set, the anomaly ending point is determined. Based on the anomaly starting point and the anomaly ending point, an anomaly propagation chain is determined, and data blocks of the participants corresponding to the anomaly propagation chain are extracted from the traceability data chain to determine the data problem type. Thus, this invention automatically identifies the participants in a time-series anomaly, accurately determines whether the anomaly is caused by the corresponding link itself or by the transmission of an anomaly from upstream, thereby reconstructing the propagation path of the anomaly in the supply chain, and further assesses the information completeness and data accuracy of each participant in the anomaly propagation chain, and then adopts corresponding feedback adjustment strategies based on different problem types, thereby incentivizing each participant to truthfully and completely report information sets from the source.

[0017] Furthermore, this invention also determines the comparison of information within the data traceability chain by based on the fact that each component of the time-series component is greater than the corresponding standard step time vector, avoiding the need to screen the entire information set. While ensuring the effective identification of abnormal propagation points, it significantly reduces the frequency of blockchain data reading. In addition, it combines the corresponding information sets of each participant to determine the standard allicin content, standard processing time, standard storage and transportation humidity change, and standard shelf time, avoiding the misjudgment or omission of abnormal propagation points due to isolated threshold settings.

[0018] Furthermore, this invention compares the timestamps corresponding to the main anomaly propagation point and independent anomaly propagation points, and determines the propagation point with the earliest timestamp as the second anomaly propagation point. This second anomaly propagation point is then identified as the anomaly starting point. The process proceeds downstream along the supply chain, traversing in ascending order of timestamps, to identify the anomaly propagation point with the corresponding timestamp as the anomaly ending point. Since the anomaly propagates from upstream to downstream along the supply chain, the anomaly propagation point with the earliest timestamp is the earliest anomaly event in the entire anomaly propagation chain, and is the source of the entire anomaly propagation chain. By taking this as the anomaly starting point and using it as a reference to traverse to the anomaly propagation point with the latest timestamp, the last affected link in the anomaly propagation chain can be identified as the anomaly ending point. This completely reconstructs the propagation path from the anomaly source to the anomaly ending point, overcoming the deficiency of existing technologies that can only identify a single anomaly point but cannot determine the causal relationship between anomalies.

[0019] Furthermore, this invention determines the data problem type by combining the upstream information set with the information sets of the corresponding participants, thereby determining the corresponding feedback adjustment method for the data problem type. This avoids directly attributing the source of the anomaly to the current participant. It comprehensively considers the situation where the downstream cannot report complete information due to the lack of upstream information, thereby avoiding unnecessary adjustment strategies for non-anomaly sources due to the difficulty in determining the source of the anomaly. This overcomes the problem in the prior art that both data omissions and data errors are handled in a uniform manner, making it impossible to make targeted improvements to the reporting behavior of the participants. Attached Figure Description

[0020] Figure 1 This is a block diagram of a blockchain-based traceability system for shallot products according to an embodiment of the present invention. Figure 2 This is a flowchart illustrating the blockchain-based traceability system for shallot products used in an embodiment of the present invention. Figure 3 This is a logic decision diagram for determining independent abnormal propagation points based on the temporal vector components of the corresponding propagation points in the abnormal propagation point set and the standard process time vector in an embodiment of the present invention. Figure 4 This is a logic decision diagram for determining the corresponding data problem type based on the corresponding information set in the data block in an embodiment of the present invention. Detailed Implementation

[0021] To make the objectives and advantages of the present invention clearer, the present invention will be further described below with reference to embodiments; it should be understood that the specific embodiments described herein are merely for explaining the present invention and are not intended to limit the present invention.

[0022] Preferred embodiments of the present invention will now be described with reference to the accompanying drawings. Those skilled in the art should understand that these embodiments are merely illustrative of the technical principles of the present invention and are not intended to limit the scope of protection of the present invention.

[0023] Please see Figure 1 The diagram shown is a block diagram of a blockchain-based traceability system for shallot products according to an embodiment of the present invention. In this embodiment, the system includes: The data input module is used to receive the corresponding information sets from the participants, including growers, processors, transporters, and sellers. The traceability data construction module, which is connected to the data input module, is used to write each information set into the blockchain in the form of timestamps to generate a traceability data chain; The time-series feature extraction module is connected to the source data construction module to extract the timestamp differences of each link based on the source data chain to form a time-series vector. An anomaly identification module is connected to the source data construction module and the time-series feature extraction module, respectively. It is used to determine the set of anomaly propagation points or the first anomaly propagation point based on the source data chain and the time-series vector. If it is an anomaly propagation point set, the time-series offset vector is calculated based on the time-series vector to determine the main anomaly propagation point. The time-series deviation between the remaining candidate anomaly propagation points that have not been determined as the main anomaly propagation point and the current main anomaly propagation point is calculated to determine whether the remaining candidate anomaly propagation points are independent anomaly propagation points. An anomaly determination module, which is connected to the anomaly identification module, is used to compare the main anomaly propagation point and independent anomaly propagation points to determine the second anomaly propagation point, determine the corresponding anomaly starting point based on the first anomaly propagation point or the second anomaly propagation point, and determine the corresponding anomaly ending point based on the second anomaly propagation point and the set of anomaly propagation points. The feedback adjustment module, connected to the anomaly point determination module, is used to determine the anomaly propagation chain based on the anomaly start point and anomaly end point. It extracts data blocks of the participants corresponding to the anomaly propagation chain from the source data chain, determines the corresponding data problem type based on the corresponding information set in the data block, and determines to adjust the data reporting frequency or data collection accuracy of the corresponding participants in the next batch based on the data problem type. The data problem type includes data omission or data misreporting.

[0024] In this embodiment, the blockchain adopts a consortium blockchain architecture, and the participating nodes include planting nodes, processing nodes, transportation nodes, sales nodes, and regulatory nodes. Each node reaches a consensus on the data blocks written to the blockchain through the Practical Byzantine Fault-Tolerant Consensus Algorithm (PBFT).

[0025] It should be noted that the PBFT consensus algorithm is suitable for consortium blockchain scenarios with a relatively fixed number of nodes and certain requirements for transaction throughput. When a node initiates a data write request, the other nodes vote on the request. When no less than two-thirds of the total number of nodes confirm the request, the data block is confirmed to be written into the blockchain.

[0026] The data traceability module calls the data on-chain smart contract deployed on the blockchain to serialize the information sets of each participant in JSON format and then encapsulates them into data blocks through the smart contract's write interface.

[0027] Before writing, the smart contract pre-validates the format integrity and timestamp validity of the information set; if the information set field is missing, the timestamp is empty, or the timestamp is earlier than the timestamp of the previous block, it refuses to write and returns an error code.

[0028] Each participating node only has the permission to write data blocks for its current stage and read data across the entire chain; it does not have the permission to modify or delete historical block data. The supervisory node has full-chain audit permissions and can read the complete content of all data blocks, but similarly does not have modification permissions. Node access control is implemented through an access control list in the smart contract, which is configured uniformly by the consortium management node when a node joins the consortium blockchain.

[0029] Please see Figure 2 The diagram shown is a flowchart illustrating the blockchain-based traceability system for shallot products according to an embodiment of the present invention. S1 receives the corresponding information set of the participants, including growers, processors, transporters and sellers; S2, write each information set into the blockchain in the form of a timestamp to generate a traceability data chain; S3, extract the timestamp difference of each link based on the traceability data chain to form a time-series vector; S4. Determine the set of abnormal propagation points or the first abnormal propagation point based on the source data chain and the time series vector. If it is an abnormal propagation point set, calculate the time series offset vector based on the time series vector to determine the main abnormal propagation point, and calculate the time series deviation between the remaining candidate abnormal propagation points that have not been determined as the main abnormal propagation point and the current main abnormal propagation point to determine whether the remaining candidate abnormal propagation points are independent abnormal propagation points. S5, compare the main anomaly propagation point with the independent anomaly propagation points to determine the second anomaly propagation point, determine the corresponding anomaly starting point based on the first anomaly propagation point or the second anomaly propagation point, and determine the corresponding anomaly ending point based on the second anomaly propagation point and the set of anomaly propagation points. S6. Determine the anomaly propagation chain based on the anomaly start point and anomaly end point. Extract the data blocks of the participants corresponding to the anomaly propagation chain from the source tracing data chain. Determine the corresponding data problem type based on the corresponding information set in the data block. Adjust the data reporting frequency or data collection accuracy of the next batch of corresponding participants based on the data problem type. The data problem type includes data omission or data misreporting.

[0030] In this embodiment, the data input module receives corresponding information sets uploaded by growers, processors, transporters, and sellers in the red shallot supply chain through user terminals. The grower information set includes fertilizer application amount and planting density, the processor information set includes allicin content, moisture content, and processing time, the transporter information set includes changes in storage and transportation humidity, and the seller information set includes shelf time.

[0031] The traceability data construction module encapsulates the information sets of each participant into corresponding data blocks, uses the current receiving time as the timestamp, and writes the hash value of the previous data block into the header of the current block. This allows each data block to be linked sequentially through the hash value of the previous block, thereby generating a traceability data chain in the order of grower, processor, transporter, and seller.

[0032] The data block consists of a block header and a block body. The block body stores the information set of the corresponding participants, while the block header stores the timestamp of this data block, the hash value of the previous block, and the hash value of this block.

[0033] It should be noted that if the information set of a certain participant is incomplete or the timestamp is missing, the traceability data construction module will not write data into the data block corresponding to the participant, but will retain the corresponding data block so that the traceability data chain maintains the complete correspondence between the positions of each participant in the structure. At the same time, the propagation point identification submodule in the anomaly identification module directly classifies the participant in the corresponding data block as the first anomaly propagation point.

[0034] The time-series feature extraction module reads the timestamps of the planting, processing, transportation, and sales data blocks respectively from the traceability data chain.

[0035] The timestamp acquisition submodule is used to calculate the first timestamp difference based on the difference between the timestamp of the processing data block and the timestamp of the planting data block, which is used to determine the time required for the red shallots to go from the end of the planting stage to the start of the processing stage.

[0036] The timestamp acquisition submodule is also used to calculate a second timestamp difference based on the difference between the timestamp of the transporter's data block and the timestamp of the processor's data block, in order to determine the time required for the red shallots to travel from the end of the processing stage to the start of the transportation stage.

[0037] The timestamp acquisition submodule is also used to calculate a third timestamp difference based on the difference between the timestamp of the seller's data block and the timestamp of the transporter's data block, which is used to determine the time required for the red shallots to go from the end of the transportation process to being put on the shelves in the sales process.

[0038] The timestamp total value acquisition submodule is used to calculate the total timestamp value by summing the first timestamp difference, the second timestamp difference, and the third timestamp difference, in order to determine the total time required for red shallots to go from planting to sales.

[0039] Furthermore, the time-series vector acquisition submodule is used to sort the first timestamp difference, the second timestamp difference, the third timestamp difference, and the total timestamp value according to the supply chain flow order of planting, processing, transportation, and sales to generate a time-series vector.

[0040] The anomaly identification module determines whether to further compare the upstream and downstream information sets within the data traceability chain based on the comparison process of each component of the time-series vector with the time consumption of the standard process, in order to identify the first anomaly propagation point or a set of anomaly propagation points.

[0041] In this embodiment, the anomaly identification module determines that the actual time consumption of the corresponding link exceeds the normal range based on the fact that each component of the time sequence vector is greater than the corresponding standard link time consumption vector. This means that there are problems such as operation delay, equipment failure, and product backlog within the link. It is difficult to determine the real cause of the product quality problem based solely on the time sequence anomaly. It is necessary to further compare the upstream and downstream information sets within the data block. Therefore, it is determined to compare the upstream and downstream information sets within the data traceability chain.

[0042] Specifically, comparing upstream and downstream information sets within the data traceability chain is used to determine the point of anomaly propagation, so as to trace the source of the anomaly and achieve accurate traceability.

[0043] The anomaly identification module also determines that the actual time consumption of the corresponding step is within the normal range based on the fact that each component of the time sequence vector is less than or equal to the time consumption vector of the corresponding standard step. The indicators such as the allicin content and moisture content of the red shallots will not deteriorate significantly, the time sequence is not abnormal, and there is no need to further analyze the information set inside the data block, thereby saving computing resources. Therefore, it is determined that no comparison of the upstream and downstream information sets inside the data traceability chain will be performed.

[0044] Specifically, a standard time vector for each stage of the red onion supply chain is pre-determined based on the industry-standard circulation time of each link. The standard time vector includes a first component, a second component, a third component, and a fourth component. Specifically, the first component is calculated based on the sum of the updated post-harvest processing time, the updated in-transit transportation time from the production site to the processor, and the updated receiving preparation time of the processor, and is used to determine the standard circulation time from the grower to the processor; the second component is calculated based on the sum of the updated product packaging time after processing, the updated labeling time, and the updated loading and dispatching time, and is used to determine the standard circulation time from the processor to the transporter; the third component is calculated based on the sum of the updated in-transit transportation time of red onions and the updated receiving and acceptance time of the seller, and is used to determine the standard circulation time from the transporter to the seller; the fourth component is calculated based on the sum of the first, second, and third components, and is used to determine the standard total time.

[0045] Specifically, the following data were collected from the most recent 20 consecutive batches of products in the same supply chain under normal production conditions, with a timestamp span of no more than 12 months: post-harvest processing time, in-transit transportation time from the place of origin to the processor, processing and receiving preparation time, product packaging time, labeling time, loading and scheduling time, in-transit transportation time, and receiving and acceptance time. The corresponding historical data median was then calculated.

[0046] The difference between the historical median and the measured data is calculated, and then multiplied by the smoothing factor to obtain the update coefficient. The updated data is then multiplied by the updated coefficient to obtain the corresponding updated data.

[0047] Specifically, based on the components of the measured time series vectors from the most recent 20 consecutive batches under normal production conditions within the same supply chain, with a timestamp span not exceeding 12 months, a grid search is performed to obtain the smoothing factor, with the objective of minimizing the mean square error between the measured value of each component and its corresponding standard component. The grid search employs leave-one-out cross-validation to avoid overfitting.

[0048] Furthermore, the propagation point identification submodule extracts information sets based on the time sequence vector where each component is greater than the corresponding standard link time vector, in order to identify the first abnormal propagation point or abnormal propagation point set.

[0049] In this embodiment, the propagation point identification submodule further determines that the allicin content of the current shallots is abnormal based on the allicin content being lower than the standard allicin content. The allicin content of the shallots is determined during the planting process and will only decrease during subsequent processing due to thermal degradation and oxidation reactions; it cannot be increased through any processing method. The abnormal allicin content originates from improper fertilization management or planting density during planting, rather than from the processing stage. Therefore, the planting party is identified as the first abnormal propagation point.

[0050] The propagation point identification submodule also determines that, based on the allicin content being greater than or equal to the standard allicin content, the sulfur assimilation pathway of red shallots is functioning normally under the current planting density and fertilization conditions, and can meet the needs of allicin synthesis. Therefore, it is determined that the grower is not the first abnormal propagation point.

[0051] Specifically, the standard allicin content is determined based on the fertilization amount and planting density of the grower's information set. Specifically, multiple batches of red onion samples under known planting density and fertilization conditions are obtained. The samples are divided into several groups based on planting density and fertilization, with samples within the same group having the same planting density and fertilization conditions. The allicin content of each group is measured, forming a sample dataset for each group. The mean and standard deviation of the allicin content in each sample dataset are calculated. The standard allicin content is determined by the difference between the mean and three times the standard deviation. The sample dataset used to calculate the standard allicin content is taken from batches of red onions that have already been tested and whose quality indicators are within the normal range, to ensure that the baseline data used for calculating the standard allicin content is not contaminated by abnormal batches.

[0052] In this embodiment, the propagation point identification submodule also determines that the actual processing time of red shallots is abnormal based on the processing time being longer than the standard processing time. This means that there are problems such as equipment performance degradation, non-standard operation, or production backlog in the processing process, which directly leads to a reduction in processing efficiency. Therefore, the processing party is identified as the first abnormal propagation point.

[0053] The propagation point identification submodule also determines that the equipment in the processing stage of the red shallot is operating well, the operation is in accordance with regulations, and the processing efficiency meets the expected requirements, based on the processing time being less than or equal to the standard processing time, and no abnormalities have occurred. Therefore, it is determined that the processor is not the first abnormal propagation point.

[0054] Specifically, the standard processing time is determined based on the moisture content information collected from the processors. Specifically, based on the dehydrated shallot processing procedure, under drying conditions of 55–60℃, shallot samples with different initial moisture contents are selected and dried. Samples are taken and weighed every 5 minutes until the moisture content of the shallot shreds drops to 4.5%. The initial moisture content and corresponding total drying time of each group of samples are recorded to establish a mapping table between the corresponding moisture content and the standard processing time. In application, the standard processing time corresponding to the moisture content value uploaded by the processor is used as the comparison benchmark by looking up the corresponding standard processing time in the mapping table. For moisture contents not appearing in the mapping table, linear interpolation is used to calculate the corresponding standard processing time. The shallot samples used for the drying experiment are taken from batches of shallots with allicin content within the normal range and no appearance defects, to avoid interference from quality abnormalities in the samples themselves.

[0055] In this embodiment, the propagation point identification submodule also determines that the change in storage and transportation humidity of shallots exceeds the allowable range based on the fact that the change in storage and transportation humidity is greater than the standard change in storage and transportation humidity. If the humidity of the storage and transportation environment is too high or too low or the humidity fluctuates frequently, it will disrupt the moisture balance of the shallot products, causing the products to lose water and shrink or absorb moisture and become damp, thereby causing quality deterioration. Therefore, the transporter is identified as the first abnormal propagation point.

[0056] The propagation point identification submodule also determines that, based on the fact that the change in storage and transportation humidity is less than or equal to the standard change in storage and transportation humidity, the actual environmental humidity fluctuation during the storage and transportation process of the shallots does not exceed the preset allowable range, the moisture balance of the shallots is not disrupted by the environmental humidity fluctuations, and the quality of the shallots is effectively guaranteed during the storage and transportation process. Therefore, it is determined that the transporter is not the first abnormal propagation point.

[0057] Specifically, the standard humidity variation during storage and transportation is determined based on the moisture content information set from the processor. Specifically, a sample of shallots with a certain moisture content is placed in an environment with periodic fluctuations in relative humidity at 25°C for testing. During the test, the sample is weighed every two hours, and the weight loss or gain rate is calculated. The test is stopped when the weight loss or gain rate reaches 5%. The difference between the highest and lowest humidity values ​​during the test is recorded to calculate the cumulative fluctuation range. This cumulative fluctuation range is used as the standard humidity variation during storage and transportation under the given moisture content to determine the maximum humidity fluctuation range that shallots can tolerate during storage and transportation at that moisture content. The shallot samples used for testing are from batches of shallots whose moisture content and allicin content are within the normal range.

[0058] Among them, 5% is the critical value for quality deterioration in the post-harvest storage of vegetables. When the water loss or weight gain of fresh vegetables exceeds 5% of their fresh weight, the cell turgor pressure decreases significantly, and the cell tissues wilt and shrink. Therefore, a weight loss or weight gain rate of 5% is used as the trigger condition for stopping the test on the quality deterioration of red shallots during storage and transportation.

[0059] In this embodiment, the propagation point identification submodule also determines that the actual shelf time of the red shallots exceeds the standard shelf time based on the shelf time being longer than the standard shelf time. The seller failed to complete the sale within the standard shelf time corresponding to the change in storage and transportation humidity. Therefore, the seller is identified as the first abnormal propagation point.

[0060] The propagation point identification submodule also determines that the seller completed the sale within the standard shelf time corresponding to the change in storage and transportation humidity, based on the shelf time being less than or equal to the standard shelf time. The inventory turnover efficiency and shelf management meet the requirements. Therefore, it is determined that the seller is not the first abnormal propagation point.

[0061] Specifically, the standard shelf life is determined based on the humidity variation during storage and transportation data collected from the transporter's information. Specifically, several groups of shallot samples are obtained. Each group consists of shallot bulbs from the same batch, uniform in size, and without mechanical damage. The initial weight of each group is the same. Each group is placed under different humidity variation conditions for storage experiments. The weight of each group is recorded every 24 hours. When the weight loss rate of any group reaches 5%, the experiment for that group is stopped. The number of days elapsed from the start to the end of the experiment for that group is used as the standard shelf life for that group under the current humidity variation conditions. Each group of samples is taken from the same batch of shallots whose humidity variation during storage and transportation is within the standard range.

[0062] It should also be noted that during the initial operation phase of the system, before accumulating 20 consecutive batches of normal historical data, the standard thresholds are determined using the following cold start rules: the standard process time vector is directly assigned the initial value of each component using the industry-standard circulation time; after accumulating 20 batches of data, dynamic calibration is performed according to the aforementioned update formula; the standard allicin content uses the industry standard data of the corresponding red onion variety under conventional planting density and recommended fertilization conditions as the initial threshold; the standard processing time uses the standard drying time specified in the corresponding dehydrated red onion processing procedure as the initial threshold; the standard storage and transportation humidity variation uses the cumulative fluctuation amplitude measured under standard environmental conditions of 25℃ and 60%±5% relative humidity as the initial threshold; and the standard shelf time uses the number of storage days measured under standard storage and transportation humidity variation conditions as the initial threshold. After the system starts operating, each batch of data that has been determined to be normal through the anomaly identification process is updated according to the aforementioned threshold determination method until the corresponding sample size requirement is met, at which point the system switches to normal dynamic update mode.

[0063] Furthermore, if the corresponding component of the time-series vector of a certain link is greater than the corresponding standard link's time-consuming vector, an internal comparison of the data traceability chain is determined. However, during the comparison process, the information set corresponding to this link does not exceed the corresponding standard threshold, and the time-series anomaly does not meet the identification conditions of any of the aforementioned first anomaly propagation points. It is determined that there is an operational delay due to non-quality factors in this link, such as backlog in logistics transfers, waiting in warehouse scheduling, or lag in information system entry. Therefore, it is determined that the participants in the corresponding link are marked as operational delay-type anomaly nodes and will not be used as first anomaly propagation points for the time being. At the same time, in the next batch of data processing, the anomaly point identification module continuously monitors the time-series component of this link. If the time-series component of three consecutive batches is greater than the corresponding standard link's time-consuming vector, it is determined that there is a systemic operational efficiency problem in this link. Therefore, it is determined that the corresponding participant is upgraded to the first anomaly propagation point.

[0064] Among them, the propagation point set identification submodule, based on the occurrence of two or more first abnormal propagation points, determines the corresponding abnormal participants as an abnormal propagation point set, which is used to collect and filter multiple abnormal propagation points, thereby determining the main abnormal propagation point and independent abnormal propagation points.

[0065] Please see Figure 3 As shown, this is a logic decision diagram for determining independent anomaly propagation points based on the temporal vector components of the corresponding propagation points in the anomaly propagation point set and the standard process time vector, according to an embodiment of the present invention. In this embodiment, the process of determining independent anomaly propagation points includes: The propagation point set identification submodule is also used to calculate the absolute value of the difference between the time sequence vector component of the corresponding propagation point in the abnormal propagation point set and the corresponding component in the standard link time consumption vector to obtain the time sequence offset vector, which is used to determine the main abnormal propagation point in the abnormal propagation point set.

[0066] The propagation point set identification submodule also determines, based on the propagation point corresponding to the largest component in the temporal offset vector within the abnormal propagation point set, that the corresponding propagation point in the abnormal propagation point set has the most severe temporal deviation and is the node with the most significant deviation accumulation and the greatest damage to the supply chain in the entire abnormal propagation chain. It should be prioritized for tracing and correction in order to block the abnormal propagation point from continuing to propagate and accumulate along the supply chain direction to the greatest extent. Therefore, the corresponding propagation point is identified as the main abnormal propagation point.

[0067] Specifically, identifying the corresponding propagation point as the main anomaly propagation point allows for precise location of the node with the most severe time sequence deviation from the anomaly propagation point, serving as the priority entry point for tracing. This enables a single-point tracing path to cover the entire anomaly propagation chain caused by the main anomaly propagation point spreading along the supply chain, avoiding the inefficiency of tracing due to checking multiple anomaly nodes one by one.

[0068] The propagation point set identification submodule also determines that, based on the set of abnormal propagation points, the propagation point whose temporal offset vector is not the largest component is less affected by the temporal offset of the main abnormal propagation point. Therefore, it is not the node with the most significant deviation in the entire abnormal propagation chain and should not be used as a priority entry point for tracing the source. Thus, the corresponding propagation point is determined not to be the main abnormal propagation point.

[0069] Specifically, it is determined that the corresponding propagation point is not the main anomaly propagation point, so as to exclude candidate nodes that are not the maximum time series offset, so that the tracing focuses on a single most serious node and avoids path crossing caused by parallel tracing from multiple starting points.

[0070] Furthermore, the propagation point set identification submodule is also used to calculate the time deviation degree by performing the absolute value difference between the remaining candidate anomaly propagation points that have not been determined as the main anomaly propagation point and the corresponding components in the time offset vector corresponding to the current main anomaly propagation point, in order to determine whether the remaining candidate anomaly propagation points are independent anomaly propagation points.

[0071] In this embodiment, the propagation point set identification submodule further determines that the timing deviation of the other candidate abnormal propagation points is greater than the standard timing deviation. It then determines that the timing offset difference between the corresponding candidate abnormal propagation point and the main abnormal propagation point is large. The timing deviation of the candidate abnormal propagation point is not caused by the deviation of the main abnormal propagation point propagating along the supply chain direction, but by an anomaly caused by an independent disturbance factor, constituting an independent abnormal propagation path. Therefore, the corresponding candidate abnormal propagation point is determined as an independent abnormal propagation point.

[0072] Specifically, the corresponding candidate anomaly propagation points are identified as independent anomaly propagation points in order to identify independent anomaly sources that have no temporal transmission relationship with the main anomaly propagation point, and to trace them as independent tracing starting points to avoid missed detections caused by a single tracing path.

[0073] The propagation point set identification submodule also determines that the time offset difference between the corresponding candidate abnormal propagation point and the main abnormal propagation point is small, based on the fact that the time offset of the other candidate abnormal propagation points is less than or equal to the standard time offset. The offset changes synchronously with the time offset of the main abnormal propagation point. That is, the time offset of the main abnormal propagation point is transmitted to this link along the supply chain direction, causing the time components of this link to deviate from the normal range synchronously. This is a chain reaction caused by the transmission of the main abnormal propagation point along the supply chain direction. Therefore, it is determined that the corresponding candidate abnormal propagation point is not an independent abnormal propagation point.

[0074] Specifically, the standard deviation of the temporal offsets of all candidate anomaly propagation points is calculated, and the standard deviation is compared with the minimum standard temporal deviation value to determine the maximum value as the standard temporal deviation, which is used to determine the overall distribution range of the temporal offsets. The minimum standard temporal deviation value is calculated based on the mean of the components of the temporal offsets of each propagation point in the anomaly propagation point set that are smaller than the temporal offset of the main anomaly propagation point.

[0075] Furthermore, the anomaly detection module compares the timestamps of the main anomaly propagation point and the independent anomaly propagation points to determine the second anomaly propagation point.

[0076] In this embodiment, the anomaly point determination module also determines that the timestamp of the main anomaly propagation point is earlier than that of the independent anomaly propagation points, based on the fact that the anomaly event corresponding to the main anomaly propagation point occurs first in the time dimension and is the earliest anomaly event to appear in the entire supply chain. Therefore, the main anomaly propagation point is determined as the second anomaly propagation point and is identified as the anomaly starting point.

[0077] The anomaly point determination module also determines that the timestamp of the independent anomaly propagation point is earlier than that of the main anomaly propagation point. The anomaly event corresponding to the independent anomaly propagation point occurs first in the time dimension and is the earliest anomaly event in the entire supply chain. Therefore, the independent anomaly propagation point is determined as the second anomaly propagation point and is identified as the anomaly starting point.

[0078] Specifically, the earliest point of the timestamp is taken as the starting point of the anomaly to ensure that the direction of tracing is consistent with the direction of anomaly propagation. That is, starting from the source, we trace downstream along the supply chain to fully cover all affected nodes in the anomaly propagation chain.

[0079] Furthermore, the anomaly point determination module also uses the first anomaly propagation point as a benchmark and traverses it in the form of timestamps to determine that there is only one anomaly propagation point in the supply chain, and there is no propagation relationship between multiple anomalies. There are no anomalies that precede this anomaly propagation point, nor are there any anomalies that follow this anomaly propagation point. Therefore, the first anomaly propagation point is determined to be both the anomaly start point and the anomaly end point.

[0080] The anomaly identification module also uses the second anomaly propagation point as a benchmark, traversing downstream along the supply chain in ascending order of timestamps. Within the set of anomaly propagation points, it identifies the last anomaly propagation point downstream of the second anomaly propagation point, determining it to be the last link affected in the anomaly propagation originating from the second anomaly propagation point. Therefore, the anomaly propagation point with the last timestamp is identified as the anomaly endpoint.

[0081] Specifically, the point where the anomaly propagation last appears in the timestamp is identified as the anomaly endpoint. This is used to accurately determine the final boundary of the anomaly propagation chain downstream, so as to avoid including links that have not been affected by the anomaly in the tracing chain and to avoid omitting abnormal links that have been affected by the anomaly.

[0082] If there are multiple independent abnormal propagation points in the cluster, then each independent abnormal propagation point is used as a reference to traverse downstream along its corresponding supply chain path. The abnormal propagation point with the last timestamp on the transmission path corresponding to each independent abnormal propagation point is determined as the corresponding abnormal endpoint, so as to form several independent abnormal propagation chains.

[0083] Please see Figure 4As shown, this is a logic decision diagram for determining the corresponding data problem type based on the corresponding information set in the data block according to an embodiment of the present invention. In this embodiment, the process of determining the corresponding data problem type based on the corresponding information set in the data block includes: The feedback adjustment module determines the starting link of the anomaly propagation chain based on the anomaly starting point and the ending link of the anomaly propagation chain based on the anomaly ending point. It traverses the supply chain from the anomaly starting point to the anomaly ending point, and all the links passed in between constitute the anomaly propagation chain.

[0084] The feedback adjustment module also extracts data blocks of the participants corresponding to the anomaly propagation chain from the source data chain, and determines the corresponding data problem type based on the corresponding information set in the data block.

[0085] Furthermore, the problem type identification submodule identifies situations where the corresponding information set in the data block has not been uploaded, or the corresponding information set does not contain complete information. That is, although the participant has uploaded an information set, the corresponding information set is missing complete information. For example, the information set of the grower is missing one or more of the fertilizer amount and planting density; the information set of the processor is missing one or more of the allicin content, moisture content, and processing time; the information set of the transporter is missing the storage and transportation humidity change; and the information set of the seller is missing the shelf time. The module then obtains the upstream information set of the corresponding participant or re-obtains the information set of the participant.

[0086] In this embodiment, the problem type identification submodule further determines that, based on the fact that the upstream information set contains complete information but the re-acquired participant information set is incomplete, the corresponding participant, even with complete upstream information, still failed to provide a complete information set after re-acquisition. This is not because the upstream information set lacks complete information, but because the current participant can obtain complete data but failed to report the complete information set. Therefore, the data problem type of the corresponding participant is determined to be data omission.

[0087] Specifically, the data problem type of the corresponding participant is identified as data omission, and the data problem type and the responsible party are accurately determined. This avoids misjudging the data incompleteness caused by the current participant's own lack of reporting behavior as the cause of data missing from the upstream party, and also avoids adopting the wrong adjustment method due to confusion of data problem types.

[0088] The problem type identification submodule also determines that the corresponding participant does not have complete information because the upstream participant itself failed to provide complete data. The transmission of supply chain data has a chain structure. When the upstream information set does not contain complete information, even if the downstream participant reports completely truthfully, it is impossible to generate a data block containing complete information. Therefore, the data problem type of the corresponding upstream party is determined to be data omission.

[0089] Specifically, identifying the data problem type of the corresponding upstream party as data underreporting is used to determine that the root cause of the data underreporting lies with the upstream participants, thus avoiding the invalidation of subsequent adjustment methods due to the division of responsibility to subsequent participants.

[0090] The problem type identification submodule also determines that the corresponding participant's information set contains complete information, and the upstream information set also contains complete information, but the current participant's information set has information distortion, such as allicin content being less than the standard allicin content, processing time being longer than the standard processing time, change in storage and transportation humidity being greater than the standard change in storage and transportation humidity, and shelf time being longer than the standard shelf time. Therefore, the data problem type of the corresponding participant is determined to be data misreporting.

[0091] Specifically, the data problem type of the corresponding participant is identified as a data misreport, which is used to determine that the data error occurred in the current participant's process and avoid directly making an error tracing method because the data error was not identified.

[0092] Furthermore, the missing data identification submodule determines to increase the data reporting frequency of the corresponding participant in the next batch based on the data problem type of missing data reporting, in order to reduce the risk of data loss due to a single missed report.

[0093] The adjustment method determination submodule is used to determine the improved data reporting frequency of the next batch of corresponding participants based on the product of the adjustment coefficient and the data reporting frequency of the current corresponding participant. The adjustment coefficient is determined based on the ratio of the corresponding component of the time sequence vector of the current corresponding participant to the time consumption vector of the corresponding standard link.

[0094] Specifically, when the ratio falls within the first interval, the corresponding adjustment coefficient is determined. When the ratio falls within the second interval, the adjustment coefficient is calculated as the sum of the lower limit of the second interval and the incremental adjustment value. When the ratio is greater than or equal to the upper limit of the second interval, the adjustment coefficient is set to 2 to prevent over-adjustment in cases of severe anomalies. The abnormal offset is calculated based on the difference between the ratio and the lower limit of the second interval; the incremental adjustment value is calculated based on the ratio of half of the abnormal offset to the lower limit of the second interval.

[0095] Based on the postharvest physiological critical values ​​of shallots, the adjacent first and second intervals are determined. When the actual harvest time exceeds the standard harvest time by 1.5 times, the water loss rate of shallots reaches 3% to 5% and the cell turgor pressure decreases significantly. 1.5 can be used as the upper limit of the first interval. When the actual harvest time reaches 3 times the standard harvest time, the rot rate of shallots rises to 10% to 15% and allicin degradation exceeds 40%, and the shallots have lost their commercial value. 3 can be used as the upper limit of the second interval.

[0096] The adjustment method determination submodule is also used to determine the accuracy of data collection for the next batch of data for the corresponding participants based on the data problem type of data misreporting, thereby improving the accuracy and reliability of the data.

[0097] The adjustment method determination submodule is also used to determine the improved data acquisition accuracy of the next batch of corresponding participants based on the product of the adjustment coefficient and the data acquisition accuracy of the current corresponding participant. The data acquisition accuracy is used to determine the number of sampling points set by the corresponding participant in a unit batch of data acquisition.

[0098] The technical solution of the present invention has been described above with reference to the preferred embodiments shown in the accompanying drawings. However, it will be readily understood by those skilled in the art that the scope of protection of the present invention is obviously not limited to these specific embodiments. Without departing from the principles of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will all fall within the scope of protection of the present invention.

Claims

1. A blockchain-based traceability system for shallot products, characterized in that, include: The data input module is used to receive the corresponding information sets from the participants, including growers, processors, transporters, and sellers. The traceability data construction module is used to write each of the aforementioned information sets into the blockchain in the form of timestamps to generate a traceability data chain; The time-series feature extraction module is used to extract the timestamp differences of each link based on the traceability data chain to form a time-series vector; An anomaly identification module is used to determine an anomaly propagation point set or a first anomaly propagation point based on the source data chain and the time-series vector. If it is an anomaly propagation point set, a time-series offset vector is calculated based on the time-series vector to determine the main anomaly propagation point, and the time-series deviation between the remaining candidate anomaly propagation points that have not been determined as the main anomaly propagation point and the current main anomaly propagation point is calculated to determine whether the remaining candidate anomaly propagation points are independent anomaly propagation points. An anomaly point determination module is used to compare the main anomaly propagation point and the independent anomaly propagation point to determine the second anomaly propagation point, determine the corresponding anomaly starting point based on the first anomaly propagation point or the second anomaly propagation point, and determine the corresponding anomaly ending point based on the second anomaly propagation point and the set of anomaly propagation points. The feedback adjustment module is used to determine the anomaly propagation chain based on the anomaly start point and the anomaly end point, extract the data blocks of the participants corresponding to the anomaly propagation chain from the source tracing data chain, determine the corresponding data problem type based on the corresponding information set in the data block, and determine to adjust the data reporting frequency or data collection accuracy of the next batch of the corresponding participants based on the data problem type. The data problem type includes data omission or data misreporting.

2. The blockchain-based traceability system for shallot products according to claim 1, characterized in that, The data input module is used to receive information sets from growers, processors, transporters, and sellers. The planting information set includes fertilizer application amount and planting density; the processing information set includes allicin content, moisture content and processing time; the transportation information set includes changes in storage and transportation humidity; and the sales information set includes shelf time.

3. The blockchain-based traceability system for shallot products according to claim 2, characterized in that, The traceability data construction module is also used to encapsulate the information sets of each participant into corresponding data blocks, and link the data blocks sequentially based on the timestamp to form the traceability data chain.

4. The blockchain-based traceability system for shallot products according to claim 3, characterized in that, The temporal feature extraction module includes: The timestamp acquisition submodule is used to extract the timestamps corresponding to the grower, the processor, the transporter, and the seller based on the traceability data chain. The timestamp difference acquisition submodule is used to calculate a first timestamp difference based on the difference between the timestamps corresponding to the processor and the planter. The timestamp difference acquisition submodule is further used to calculate a second timestamp difference based on the difference between the timestamps corresponding to the transporter and the processor; The timestamp difference acquisition submodule is further used to calculate a third timestamp difference based on the difference between the timestamps corresponding to the seller and the transporter; The timestamp total value acquisition submodule is used to calculate the timestamp total value based on the first timestamp difference, the second timestamp difference, and the third timestamp difference; The time-series vector acquisition submodule is used to construct the time-series vector based on the first timestamp difference, the second timestamp difference, the third timestamp difference, and the total timestamp value.

5. The blockchain-based traceability system for shallot products according to claim 4, characterized in that, The anomaly identification module is also used to determine the comparison of the data tracing chain based on the fact that each component of the time-series vector is greater than the corresponding standard link time-consuming vector.

6. The blockchain-based traceability system for shallot products according to claim 5, characterized in that, The anomaly identification module includes: The propagation point identification submodule is used to determine the planting site as the first abnormal propagation point based on the fact that the allicin content is less than the standard allicin content. The propagation point identification submodule is also used to determine the processing party as the first abnormal propagation point based on the processing time being greater than the standard processing time; The propagation point identification submodule is also used to determine the transporter as the first abnormal propagation point based on the fact that the change in storage and transportation humidity is greater than the standard change in storage and transportation humidity. The propagation point identification submodule is also used to determine the seller as the first abnormal propagation point based on the fact that the listing duration is greater than the standard listing duration; Specifically, the standard allicin content is determined based on the amount of fertilizer applied and the planting density; the standard processing time and the standard storage and transportation humidity change are determined based on the moisture content; and the standard shelf-reading time is determined based on the storage and transportation humidity change.

7. The blockchain-based traceability system for shallot products according to claim 6, characterized in that, The anomaly identification module further includes: The propagation point set identification submodule is used to identify several abnormal participants as the abnormal propagation point set based on the occurrence of two or more of the first abnormal propagation points. The propagation point set identification submodule is further used to determine the propagation point corresponding to the largest component in the time offset vector of the abnormal propagation point set as the main abnormal propagation point; The propagation point set identification submodule is further used to identify the candidate abnormal propagation points corresponding to the time deviation degree being greater than the standard time deviation degree as the independent abnormal propagation points; The timing offset vector is obtained based on the absolute difference between each component of the timing vector and the corresponding standard stage time consumption vector, and the timing deviation is obtained based on the absolute difference between the corresponding components in the timing offset vectors of the candidate anomaly propagation point and the main anomaly propagation point.

8. The blockchain-based traceability system for shallot products according to claim 7, characterized in that, The anomaly point determination module is further used to compare the timestamps corresponding to the main anomaly propagation point and the independent anomaly propagation point, and determine the propagation point with the earliest timestamp as the second anomaly propagation point, and determine the second anomaly propagation point as the anomaly starting point.

9. The blockchain-based traceability system for shallot products according to claim 8, characterized in that, The anomaly point determination module is also used to take the first anomaly propagation point as a reference, traverse along the supply chain direction and in ascending order of timestamps, and determine that the first anomaly propagation point is both the anomaly start point and the anomaly end point. The anomaly point determination module is also used to take the second anomaly propagation point as a reference, traverse along the supply chain direction in ascending order of timestamps, determine the candidate anomaly propagation point located downstream of the second anomaly propagation point in the set of anomaly propagation points, and determine the anomaly propagation point with the last timestamp as the anomaly endpoint. If there are multiple independent abnormal propagation points in the abnormal propagation point cluster, then each independent abnormal propagation point is used as a reference to traverse downstream along its corresponding supply chain path to determine the abnormal endpoint of each independent abnormal propagation chain.

10. The blockchain-based traceability system for shallot products according to claim 9, characterized in that, The feedback adjustment module includes: The problem type identification submodule is used to obtain the upstream information set of the corresponding participant or re-obtain the information set of the participant based on the fact that the corresponding information set in the data block has not been uploaded or the corresponding information set does not contain complete information. The missing data identification submodule is used to determine the data problem type of the corresponding participant or upstream party as the data missing data based on the upstream information set containing complete information and the re-acquired participant information being incomplete, or the upstream information set not containing the complete information. The error message identification submodule is used to determine the data problem type of the corresponding participant as the data error message based on the corresponding information set in the data block and the fact that the upstream information set contains the complete information. The adjustment method determination submodule is used to determine, based on the data problem type being data underreporting, to increase the data reporting frequency of the corresponding participants in the next batch, or based on the data problem type being data misreporting, to determine to increase the data collection accuracy of the corresponding participants in the next batch.

Images