A data analysis method based on collected trajectory checking

Abstract

This invention discloses a data analysis method based on the verification of the acquisition trajectory, comprising: locating the sequential break position in the acquisition trajectory through joint verification of trajectory spacing, acquisition sequence interval, and trajectory point association data residuals; splitting the original acquisition data into multiple data analysis sub-chains to avoid direct continuation analysis of jump points, broken chains, or looping data; establishing candidate continuation relationships between broken sub-chains by pairing the starting and ending endpoints one by one; eliminating reverse continuation and gradient boundary crossing endpoint pairs based on the trajectory direction continuation status and data gradient continuation status to reduce cross-location, cross-channel, or cross-segment misconnections; and finally rearranging the data analysis sub-chains according to the target endpoint pairs to form a verification data analysis link, enabling warehouse inspection acquisition data to enter the analysis and evidence storage process according to the actual cargo location arrangement order, thereby improving the consistency of the acquisition data link, the accuracy of anomaly location, and the credibility of data before blockchain evidence storage.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and in particular to a data analysis method based on the verification of acquisition trajectory. Background Technology

[0002] Existing warehouse inspection data processing technologies primarily focus on the identification, filtering, deduplication, and evidence storage of the collected data itself. For example, they identify location images, match electronic tag signals, identify anomalies in environmental sensor data, or upload data summaries to blockchain nodes. These technologies typically assume the collection order matches the actual location order, or simply archive the collection results using timestamps, device numbers, and location coordinates. For gaps in the collection trajectory caused by factors such as shelf obstruction, aisle turns, weak wireless signal areas, location drift, manual intervention, and inspection terminal looping, existing processing methods often only delete abnormal points or supplement missing points, making it difficult to identify misconnections in the data collection chain.

[0003] In blockchain supply chain warehousing inspection applications, once data summaries are uploaded to the blockchain, they form strong traceability constraints. However, blockchain notarization mainly ensures the immutability of data after it is uploaded to the blockchain, but it cannot automatically guarantee the correct spatial order of data collected before uploading. If the inspection terminal experiences a trajectory jump in a certain shelf aisle, incorrectly continuing the data of the later storage location segment to the previous storage location segment before performing storage location status analysis or blockchain notarization, although it can be verified that the data summary has not been tampered with, it is difficult to confirm whether the collected data corresponding to the data summary is formed according to the actual storage location arrangement order. This can lead to problems such as misjudgment of out-of-stock locations, spread of abnormal environmental segments, mismatch between storage location images and electronic tag signals, and inconsistency between the notarization link and the actual inspection route.

[0004] Furthermore, existing trajectory correction technologies primarily focus on smoothing trajectory point positions, filling in trajectory points, and removing trajectory anomalies. Their processing targets are usually the trajectory itself, rather than the analysis order of the trajectory-related collected data. For warehouse inspection data, simply restoring the geometric continuity of the trajectory is insufficient to guarantee the correctness of the data analysis chain. It is also necessary to consider the trajectory spacing between adjacent trajectory points, the interval between collection sequences, the residual data associated with trajectory points, the continuity of sub-chain endpoint directions, and the continuity of data gradients to perform endpoint matching and reordering of the connection sequence for the broken data analysis sub-chain. Therefore, existing technologies lack a technical solution capable of identifying trajectory breaks, splitting data analysis sub-chains, and reordering the verification chain for warehouse inspection collected data before blockchain notarization.

[0005] In summary, existing warehouse inspection data processing technologies suffer from problems such as broken data collection trajectories leading to misconnected data analysis links and difficulty in verifying the data order before blockchain storage. Our invention addresses the problem of verifying and rearranging the data analysis links based on the location of the broken data collection trajectory and the matching relationship between sub-chain endpoints. Summary of the Invention

[0006] The purpose of this section is to outline some aspects of the embodiments of the present invention and to briefly introduce some preferred embodiments. Some simplifications or omissions may be made in this section, as well as in the abstract and title of the present application, to avoid obscuring the purpose of this section, the abstract and title of the invention. Such simplifications or omissions shall not be used to limit the scope of the present invention.

[0007] In view of the aforementioned existing problems, the present invention is proposed.

[0008] To solve the above technical problems, the present invention provides the following technical solution: the position where at least one of the trajectory spacing and acquisition sequence interval of adjacent trajectory points in the acquisition trajectory does not meet the continuous boundary, and the residual of the trajectory point associated acquisition data does not meet the residual boundary is taken as the trajectory sequence break position, and the trajectory sequence break position is used as the boundary to split into multiple data analysis sub-chains; Extract the start and end endpoints of each of the data analysis subchains, and pair the end endpoints of each of the data analysis subchains with the start endpoints of the remaining data analysis subchains one by one to obtain candidate endpoint pairs; Based on the trajectory direction continuity and data gradient continuity between the candidate endpoint pairs, candidate endpoint pairs with opposite trajectory directions or data gradients that exceed the limits are eliminated, and the target endpoint pairs are retained. Adjust the connection order of multiple data analysis sub-chains according to the target endpoints, generate a verification data analysis link, and rearrange the collected data in the data analysis sub-chains according to the verification data analysis link.

[0009] As a preferred embodiment of the data analysis method based on acquisition trajectory verification according to the present invention, the method for obtaining the trajectory spacing, the acquisition sequence interval, and the residual of the trajectory point associated acquisition data includes: The trajectory points in the collected trajectory are arranged according to the sampling order, and two adjacent trajectory points are formed into adjacent trajectory point pairs. Calculate the spatial distance between two trajectory points in the adjacent trajectory point pair to obtain the trajectory spacing, and use the difference in the sequence number of the two trajectory points in the sampling order as the sampling sequence interval; Extract the collected data associated with the two trajectory points in the adjacent trajectory point pair, and calculate the difference between the two collected data to obtain the trajectory point associated collected data residual.

[0010] As a preferred embodiment of the data analysis method based on trajectory verification described in this invention, determining the location of the trajectory sequence break includes: The trajectory spacing is compared with the continuous distance boundary, the acquisition sequence interval is compared with the continuous sequence boundary, and the trajectory point associated acquisition data residual is compared with the residual boundary. When the trajectory spacing is greater than the continuous distance boundary, or the acquisition sequence interval is greater than the continuous sequence boundary, and the residual of the trajectory point associated acquisition data is greater than the residual boundary, the position between the adjacent trajectory point pairs is taken as the trajectory sequence break position.

[0011] As a preferred embodiment of the data analysis method based on trajectory verification described in this invention, the data is divided into multiple data analysis sub-chains with the trajectory sequence break position as the boundary, including: The starting position of the acquisition trajectory, the break position of the trajectory sequence, and the ending position of the acquisition trajectory are used as segment boundaries, and the trajectory points and their associated acquisition data between two adjacent segment boundaries are grouped into a data analysis sub-chain.

[0012] As a preferred embodiment of the data analysis method based on acquisition trajectory verification described in this invention, extracting the start and end endpoints of each data analysis sub-chain includes: According to the arrangement order of trajectory points in the data analysis sub-chain, the trajectory point at the beginning is determined as the starting endpoint, and the trajectory point at the end is determined as the ending endpoint. Extract the endpoint position, endpoint acquisition sequence, and endpoint associated acquisition data of the starting endpoint to form a starting endpoint parameter group; Extract the endpoint position, endpoint acquisition sequence, and endpoint associated acquisition data of the termination endpoint to form a termination endpoint parameter group.

[0013] As a preferred embodiment of the data analysis method based on acquisition trajectory verification described in this invention, obtaining the candidate endpoint pairs includes: Each of the termination endpoint parameter groups is combined one by one with the starting endpoint parameter groups that do not belong to the same data analysis subchain to form endpoint pairing groups; Delete endpoint pairs that have overlapping endpoint acquisition order or the same endpoint location, and keep the remaining endpoint pairs as candidate endpoint pairs.

[0014] As a preferred embodiment of the data analysis method based on acquisition trajectory verification described in this invention, retaining the target endpoint pairs includes: For each candidate endpoint pair, take the trajectory direction of the last segment of the data analysis subchain to which the terminating endpoint belongs, and take the trajectory direction of the first segment of the data analysis subchain to which the starting endpoint belongs; The angle between the direction of the last segment trajectory and the direction of the first segment trajectory is compared to obtain the trajectory direction continuity state, and the candidate endpoint pairs whose angle exceeds the direction boundary are marked as direction reverse endpoint pairs. Based on the endpoint association data collected in the termination endpoint parameter group and the endpoint association data collected in the starting endpoint parameter group, calculate the cross-endpoint data gradient of the candidate endpoint pair; The data gradient across endpoints is compared with the data gradient of the last segment of the data analysis subchain to which the termination endpoint belongs and the data gradient of the first segment of the data analysis subchain to which the starting endpoint belongs, respectively, to obtain the data gradient continuation state. The candidate endpoint pairs whose difference exceeds the gradient boundary are marked as gradient out-of-bounds endpoint pairs, and the reverse direction endpoint pairs and the gradient out-of-bounds endpoint pairs are deleted; The candidate endpoint pairs that have not been deleted are used as target endpoint pairs.

[0015] As a preferred embodiment of the data analysis method based on acquisition trajectory verification described in this invention, the included angle comparison includes: Based on the position of the adjacent trajectory points at the end of the data analysis sub-chain to which the termination endpoint belongs, the direction of the final trajectory segment is determined by connecting the positions of the adjacent trajectory points at the end of the sub-chain. Based on the positional connection of adjacent trajectory points in the data analysis subchain to which the starting endpoint belongs, the direction of the first trajectory segment is determined. Calculate the angle between the direction of the final trajectory segment and the direction of the first trajectory segment; When the included angle of the directions is less than or equal to the first angle boundary, the candidate endpoint pair is marked as a direction continuation endpoint pair; When the included angle of direction is greater than the first angle boundary and less than the second angle boundary, the candidate endpoint pair is marked as a direction deflection endpoint pair; When the included angle of the directions is greater than or equal to the second angle boundary, the candidate endpoint pair is marked as the reversed direction endpoint pair.

[0016] As a preferred embodiment of the data analysis method based on acquisition trajectory verification described in this invention, the first angle boundary is 30° to 60°; the second angle boundary is 120° to 150°.

[0017] As a preferred embodiment of the data analysis method based on acquisition trajectory verification described in this invention, generating the verification data analysis link includes: Each of the data analysis sub-chains is used as a link node, and the pairing direction from the end endpoint to the start endpoint in the target endpoint pair is used as the connection direction between the link nodes to generate a sub-chain connection relationship. When there are multiple subchain connection relationships in the same data analysis subchain, the subchain connection relationship with the closest trajectory direction continuation state and data gradient continuation state is retained. Multiple data analysis sub-chains are arranged sequentially according to the sub-chain connection relationship to generate a verification data analysis link; The collected data in each of the data analysis sub-chains is retained in the original order of the trajectory points, and the collected data in different data analysis sub-chains are rearranged according to the verification data analysis link.

[0018] The beneficial effects of this invention are as follows: This invention locates the sequential breakpoints in the acquisition trajectory by jointly verifying the trajectory spacing, acquisition sequence interval, and trajectory point association data residuals, and splits the original acquisition data into multiple data analysis sub-chains to avoid direct continuation analysis of jump points, broken chains, or looping data; by pairing the starting and ending endpoints one by one, candidate continuation relationships between broken sub-chains are established; then, based on the trajectory direction continuation status and data gradient continuation status, reverse continuation and gradient boundary crossing endpoint pairs are eliminated to reduce cross-location, cross-channel, or cross-segment misconnections; finally, the data analysis sub-chains are rearranged according to the target endpoint pairs to form a verification data analysis link, so that the warehouse inspection acquisition data enters the analysis and evidence storage process according to the actual cargo location arrangement order, improving the consistency of the acquisition data link, the accuracy of anomaly location, and the credibility of data before blockchain evidence storage. Attached Figure Description

[0019] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. Wherein: Figure 1 This is a flowchart illustrating the data analysis method based on acquisition trajectory verification as shown in this invention. Detailed Implementation

[0020] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0021] Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without inventive effort should fall within the scope of protection of this invention.

[0022] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0023] In a preferred embodiment, the present invention is applied to a blockchain supply chain warehouse inspection data collection link verification scenario; the inspection terminal moves along the warehouse rack aisle and forms a collection trajectory at different cargo locations, and the trajectory points in the collection trajectory are associated with at least one of the following collection data: cargo location image recognition data, electronic tag signal data, and cargo location environmental sensing data; wherein, the cargo location image recognition data reflects the cargo's location status, the electronic tag signal data reflects the matching status between the electronic tag and the cargo location, and the cargo location environmental sensing data reflects the temperature and humidity status of the local area where the cargo location is located; due to rack obstruction, aisle turns, weak wireless signal areas, positioning drift, or inspection terminal looping, adjacent trajectory points may be spatially discontinuous, the sampling sequence may be interrupted, or the cargo location status data may change abruptly. Therefore, in this embodiment, the collection trajectory is first broken at the location, and the data analysis sub-chain required for subsequent endpoint pairing is formed with the break location as the boundary.

[0024] According to an embodiment of the present invention, in combination Figure 1 The flowchart shown illustrates a data analysis method based on trajectory verification, which specifically includes the following steps: S1. The location where at least one of the trajectory spacing or acquisition sequence interval between adjacent trajectory points in the acquisition trajectory does not meet the continuity boundary, and where the residual of the associated acquisition data of the trajectory points does not meet the residual boundary, is designated as the trajectory sequence break point. This break point is then used as the boundary to split the data into multiple data analysis sub-chains. Note that the following should be noted in this step: S1.1 Arrange the trajectory points in the sampling trajectory according to the sampling order, and form adjacent trajectory point pairs with two adjacent trajectory points.

[0025] Specifically, the trajectory points formed by the inspection terminal as it moves within the warehouse aisle are arranged in the order of sampling to obtain the collected trajectory. Each trajectory point includes at least the trajectory point position and the sampling sequence number. The trajectory point position is the position of the inspection terminal in the aisle coordinate system, and the sampling sequence number is the arrangement number of the trajectory point in the sampling order.

[0026] Preferably, two trajectory points with adjacent sampling sequence numbers are grouped into an adjacent trajectory point pair so that the continuity of each adjacent trajectory point pair can be determined.

[0027] In warehouse inspection scenarios, the rack aisle coordinate system can be defined with the rack aisle entrance as the starting point, the rack extension direction as the first coordinate direction, and the aisle width direction as the second coordinate direction. Thus, the position of the trajectory point can reflect the positional change of the inspection terminal relative to the arrangement direction of the storage location, making it easier to determine whether adjacent trajectory points are still on the same continuous inspection path. For warehouse areas with multiple rack aisles arranged in parallel, a rack aisle coordinate system can be established for each rack aisle, or a rack aisle identifier can be added to each trajectory point under a unified warehouse coordinate system, so that the position of the trajectory point can correspond to the specific rack aisle and storage location arrangement direction.

[0028] S1.2 Calculate the spatial distance between two trajectory points in an adjacent trajectory point pair to obtain the trajectory spacing, and use the difference in the sequence number of the two trajectory points in the sampling order as the sampling sequence interval.

[0029] Specifically, for each pair of adjacent trajectory points, the spatial distance between them is calculated based on the positions of the two trajectory points, and this spatial distance is used as the trajectory spacing. The trajectory spacing reflects the spatial movement range of the inspection terminal between two adjacent samplings. For inspection terminals that move continuously along the shelf aisle, the trajectory spacing between adjacent trajectory points is usually adapted to the inspection terminal's moving speed, sampling frequency, and storage location spacing. If the trajectory spacing is significantly greater than the distance boundary under normal continuous movement conditions, it indicates that there may be positioning jumps, weak signal loss points, or the inspection terminal moving across aisles between adjacent trajectory points.

[0030] Meanwhile, the difference in sequence number is calculated based on the sampling sequence number of the two trajectory points, and the difference in sequence number is used as the sampling sequence interval. If the inspection terminal samples continuously, the sampling sequence interval is usually maintained at the interval between adjacent numbers. If the sampling sequence interval increases, it indicates that there is a sampling gap, upload delay or missing trajectory record between the two trajectory points. By obtaining the trajectory spacing and the sampling sequence interval at the same time, the suspected break position in the collected trajectory can be located from both the spatial continuity and the sampling sequence continuity.

[0031] In a preferred embodiment, the location of the trajectory point is represented by two-dimensional or three-dimensional coordinates. When the warehouse rack is a multi-layer structure, the location of the trajectory point also includes the rack level location. In this case, the trajectory spacing is calculated based on the spatial coordinate difference under the coordinate system of the same rack aisle, and the sampling sequence interval is calculated based on the difference between the sampling sequence numbers.

[0032] S1.3 Extract the collected data associated with the two trajectory points in the adjacent trajectory point pair, and calculate the difference between the two collected data to obtain the trajectory point associated collected data residual.

[0033] Specifically, for adjacent trajectory point pairs, the collected data associated with the two trajectory points are extracted respectively. The collected data can be one or more of the following: cargo location status value obtained from cargo location image recognition, electronic tag signal strength value, electronic tag matching status value, temperature value, and humidity value. For single-type collected data, the difference between the collected data associated with the two trajectory points is directly calculated, and the absolute value of the difference is used as the trajectory point associated collected data residual. For multiple types of collected data, the difference between the same type of collected data and the two trajectory points is calculated respectively, and the trajectory point associated collected data residual is synthesized according to the difference after data type normalization.

[0034] In a preferred embodiment, the cargo location image recognition data is represented by cargo location confidence, the electronic tag signal data by electronic tag matching strength, and the environmental sensor data by temperature and humidity values. After calculating the difference of the same type for each of the above different types of collected data, they are synthesized according to the contribution ratio of each type of collected data in the cargo location status analysis to obtain the trajectory point associated collected data residual. When normalizing different types of collected data, the normal value range of the collected data in the same shelf aisle can be used for normalization. The contribution ratio can be determined according to the cargo location status analysis task. For example, in the cargo location status analysis task, the contribution ratio of increasing cargo location confidence and electronic tag matching strength can be increased, and in the environmental anomaly analysis task, the contribution ratio of increasing temperature and humidity values ​​can be increased.

[0035] Preferably, this process avoids judging breakage solely based on trajectory location, while incorporating sudden changes in cargo location status into the trajectory breakage determination, making the trajectory sequence breakage location more closely match the actual misconnection location in the warehouse inspection data analysis link.

[0036] S1.4. Compare the trajectory spacing with the continuous distance boundary, compare the acquisition sequence interval with the continuous sequence boundary, and compare the trajectory point associated acquisition data residual with the residual boundary.

[0037] Specifically, the continuous distance boundary is determined based on the center distance between adjacent storage locations within the warehouse rack aisle, the normal moving speed of the inspection terminal, and the sampling frequency. For example, when the inspection terminal is inspecting at a constant speed along the rack aisle, the normal distance between adjacent trajectory points should not exceed the sum of the center distance between adjacent storage locations and the upper limit of the positioning error. When the trajectory spacing is greater than the continuous distance boundary, it indicates that the adjacent trajectory points do not satisfy the continuous movement relationship in space.

[0038] The continuous sequence boundary is determined according to the continuous numbering rule of the sampling sequence number of the inspection terminal; for example, if each valid sampling forms a sampling sequence number, the sampling sequence interval between adjacent valid trajectory points should be within the continuous numbering range; when the sampling sequence interval is greater than the continuous sequence boundary, it indicates that there is a sampling gap or missing data upload between the two trajectory points.

[0039] The residual boundary is determined based on the normal fluctuation range of the status of adjacent storage locations within the same shelf aisle. For storage location image recognition data, the residual boundary can be determined based on the allowable difference in location reliability between adjacent storage locations. For electronic tag signal data, the residual boundary can be determined based on the allowable attenuation range of the electronic tag signal strength between adjacent storage locations. For environmental sensing data, the residual boundary can be determined based on the normal spatial gradient of temperature and humidity within the same shelf aisle. When the residual of the trajectory point associated acquisition data is greater than the residual boundary, it indicates that there is a data abrupt change between the acquisition data corresponding to two adjacent trajectory points that does not conform to the continuous storage location arrangement relationship. The continuous distance boundary, continuous sequence boundary, and residual boundary can be jointly determined by the shelf structure parameters recorded in the warehousing system, the sampling rules of the inspection terminal, and historical normal inspection data, and should remain consistent under the same warehousing area, the same shelf aisle type, or the same inspection task type.

[0040] S1.5 When the trajectory spacing is greater than the continuous distance boundary, or the acquisition sequence interval is greater than the continuous sequence boundary, and the residual of the trajectory point association acquisition data is greater than the residual boundary, the position between adjacent trajectory point pairs shall be taken as the trajectory sequence break position.

[0041] Specifically, this embodiment does not directly determine the trajectory sequence break position by a single trajectory distance anomaly. Instead, it requires that the trajectory continuity anomaly and the residual of the collected data anomaly occur simultaneously. That is, when the trajectory spacing is greater than the continuous distance boundary and the residual of the trajectory point associated collected data is greater than the residual boundary, the position between adjacent trajectory point pairs is taken as the trajectory sequence break position. Alternatively, when the collection sequence interval is greater than the continuous sequence boundary and the residual of the trajectory point associated collected data is greater than the residual boundary, the position between adjacent trajectory point pairs is taken as the trajectory sequence break position.

[0042] In a blockchain supply chain warehouse inspection scenario, if the inspection terminal jumps from the end of shelf A aisle to the middle of shelf B aisle, the trajectory distance between two adjacent trajectory points will exceed the continuous distance boundary, and at the same time, the corresponding location image recognition result or electronic tag signal status will show obvious residuals. At this time, the position between the adjacent trajectory point pair is determined as the trajectory sequence break position. If there is only short-term positioning jitter but the location status data is continuous, this position is not directly used as the trajectory sequence break position, thereby reducing the mis-splitting caused by positioning noise. If the sampling sequence interval increases but the residual of the trajectory point associated sampling data does not exceed the residual boundary, it means that this position may only have a longer sampling interval without causing a misconnection of the location status link. In this embodiment, this position is not directly used as the trajectory sequence break position.

[0043] S1.6. The starting position of the acquisition trajectory, the position where the trajectory sequence breaks, and the ending position of the acquisition trajectory are used as segment boundaries, and the trajectory points between two adjacent segment boundaries and their associated acquisition data are merged into a data analysis sub-chain.

[0044] Specifically, after determining the breakage of all adjacent trajectory point pairs, the starting position of the collected trajectory, the breakage position of each trajectory sequence, and the ending position of the collected trajectory are arranged in the order of sampling to form multiple segment boundaries. For any two adjacent segment boundaries, the trajectory points located between them and the collected data associated with the trajectory points are merged into a data analysis sub-chain. The original trajectory point arrangement order is retained within each data analysis sub-chain, so that the location data collected in the same continuous shelf section is still arranged in the original sampling order. The original sampling continuation relationship is no longer directly used between different data analysis sub-chains, but the connection order is re-determined by the subsequent endpoint pairing and link rearrangement steps.

[0045] In a preferred embodiment, if there are multiple trajectory sequence breakpoints in the acquisition trajectory, multiple data analysis sub-chains are generated sequentially according to the sampling order; each data analysis sub-chain contains the trajectory points, trajectory point positions, sampling sequence numbers, and trajectory point associated acquisition data within that sub-chain.

[0046] It should be noted that, through the above processing, step S1 transforms the locations in the original collection trajectory that may contain misconnections, jumps, loops, or sampling gaps into clear trajectory sequence break locations, and splits the original collection data into multiple data analysis sub-chains with internal continuity. This processing provides structured objects for subsequent extraction of start endpoints, end endpoints, and screening of target endpoint pairs, and can prevent warehouse inspection data from directly entering the data analysis process before blockchain notarization according to the abnormal sampling order.

[0047] S2. Extract the start and end endpoints of each data analysis sub-chain, and pair the end endpoint of each data analysis sub-chain with the start endpoints of the remaining data analysis sub-chains one by one to obtain candidate endpoint pairs. Note that the following should be noted in this step: S2.1 According to the arrangement order of trajectory points in the data analysis sub-chain, the trajectory point at the first position is determined as the starting endpoint, and the trajectory point at the last position is determined as the ending endpoint.

[0048] Specifically, for each data analysis subchain, the endpoints are determined according to the arrangement order of the trajectory points within the data analysis subchain. Since the data analysis subchain is formed by merging the trajectory points between two adjacent segment boundaries and their associated collected data, the arrangement order of the trajectory points within the data analysis subchain is still the sampling order. The trajectory point at the beginning of this arrangement is determined as the starting endpoint, and the trajectory point at the end of this arrangement is determined as the ending endpoint.

[0049] In a blockchain-based supply chain warehouse inspection scenario, a data analysis sub-chain can correspond to a continuous storage location segment within a shelf aisle. The starting endpoint corresponds to the initial data collection position of this continuous storage location segment, and the ending endpoint corresponds to the final data collection position. In this way, each continuous storage location segment can be transformed into a data chain segment with a clear start and end position, facilitating subsequent determination of whether the data chain segment can form a reasonable continuation with other data chain segments. When a data analysis sub-chain contains only one trajectory point, this trajectory point serves as both the starting and ending endpoints, and participates in conflict resolution in subsequent endpoint pairing based on endpoint position, endpoint collection order, and associated data collection.

[0050] S2.2 Extract the endpoint position, endpoint acquisition sequence, and endpoint associated acquisition data of the starting endpoint to form the starting endpoint parameter group.

[0051] Specifically, for each starting endpoint, its endpoint position, endpoint acquisition sequence, and endpoint associated acquisition data are extracted and arranged into a starting endpoint parameter group in a unified manner; the endpoint position is the position of the starting endpoint in the warehouse rack aisle coordinate system; the endpoint acquisition sequence is the sampling sequence number corresponding to the starting endpoint; the endpoint associated acquisition data is at least one of the following: the cargo location image recognition data, electronic tag signal data, and cargo location environment sensing data corresponding to the starting endpoint.

[0052] In a preferred embodiment, the endpoint association data includes cargo location confidence, electronic tag matching strength, temperature value, and humidity value; the starting endpoint parameter group can be arranged in the order of endpoint location - endpoint acquisition sequence - endpoint association data.

[0053] S2.3 Extract the endpoint position, endpoint acquisition sequence, and endpoint associated acquisition data of the termination endpoint to form the termination endpoint parameter group.

[0054] Specifically, for each termination endpoint, its endpoint position, endpoint acquisition order, and endpoint-related acquisition data are extracted and arranged in the same way as the starting endpoint parameter group to form a termination endpoint parameter group. The endpoint position in the termination endpoint parameter group is used to determine the spatial continuity between the end position of the data analysis sub-chain and the start position of other data analysis sub-chains. The endpoint acquisition order is used to determine whether there is a acquisition order conflict between different data analysis sub-chains. The endpoint-related acquisition data is used for subsequent calculation of cross-endpoint data gradients.

[0055] By forming separate start endpoint parameter groups and end endpoint parameter groups, the continuity judgment between different data analysis subchains no longer depends on a single location coordinate, but simultaneously possesses endpoint information in three dimensions: location, order, and collected data.

[0056] In warehouse inspection, this process can avoid incorrectly connecting data segments of different shelf aisles simply because the two endpoints are close together, and it can also avoid continuing the original incorrect connection relationship simply by connecting according to the sampling time. The starting endpoint parameter group and the ending endpoint parameter group adopt the same data arrangement structure, so that subsequent endpoint pairing, direction comparison and data gradient comparison can be carried out under a unified parameter structure.

[0057] S2.4. Combine each terminal endpoint parameter group with the starting endpoint parameter group that does not belong to the same data analysis subchain one by one to form an endpoint pairing group.

[0058] Specifically, the terminal endpoint parameter group of each data analysis subchain is used as the endpoint to be connected. The terminal endpoint parameter group is combined with the starting endpoint parameter group of other data analysis subchains to form an endpoint pairing group. The terminal endpoint parameter group and the starting endpoint parameter group within the same data analysis subchain are not combined to avoid the same data analysis subchain being repeatedly connected to itself and forming a circular connection.

[0059] For example, when the acquisition trajectory is divided into a first data analysis sub-chain, a second data analysis sub-chain, and a third data analysis sub-chain, the termination endpoint parameter group of the first data analysis sub-chain is combined with the starting endpoint parameter groups of the second and third data analysis sub-chains, respectively; the termination endpoint parameter group of the second data analysis sub-chain is combined with the starting endpoint parameter groups of the first and third data analysis sub-chains, respectively; and the termination endpoint parameter group of the third data analysis sub-chain is combined with the starting endpoint parameter groups of the first and second data analysis sub-chains, respectively. This forms endpoint pairing groups that cover all possible connections between different data analysis sub-chains.

[0060] In the blockchain supply chain warehouse inspection scenario, endpoint pairing groups indicate that there may be a continuity relationship between the end position of one continuous storage location segment and the start position of another continuous storage location segment.

[0061] Preferably, this step does not directly determine the final continuation order, but first retains all possible continuation objects so that they can be filtered later based on the trajectory direction continuation status and the data gradient continuation status.

[0062] S2.5 Delete endpoint pairs with overlapping endpoint acquisition order or the same endpoint location, and use the remaining endpoint pairs as candidate endpoint pairs.

[0063] Specifically, basic conflict elimination is performed on endpoint pairing groups. If the endpoint sampling order of the terminating endpoint parameter group and the starting endpoint parameter group in the endpoint pairing group overlaps, it means that the sampling ranges corresponding to the two in the original sampling link overlap, and it is not suitable as a continuation relationship between two different data analysis sub-chains. Therefore, the endpoint pairing group is deleted. If the endpoint parameter group and the starting endpoint parameter group in the endpoint pairing group have the same endpoint position, it means that the two endpoints may correspond to the same trajectory point, repeated sampling point, or repeated records of the same cargo location. In order to avoid forming duplicate continuations, the endpoint pairing group is also deleted.

[0064] Overlapping endpoint sampling sequences can be determined based on the sampling sequence range of the data analysis sub-chains to which the two endpoints belong. Specifically, if the sampling sequence range of one data analysis sub-chain intersects with the sampling sequence range of another data analysis sub-chain, it is determined that the corresponding endpoint pairing group has overlapping endpoint sampling sequences. The same endpoint positions can be determined based on whether the endpoint positions fall within the same positioning error range. In the warehouse racking aisle, if the positions of two endpoints correspond to the same warehouse location center position or the same positioning grid, it is determined that the endpoint positions are the same.

[0065] After deleting the endpoint pairings with basic conflicts, the remaining endpoint pairings are used as candidate endpoint pairs. Candidate endpoint pairs represent endpoint combinations that may still have reasonable continuity relationships between different data analysis sub-chains. Through this process, subsequent steps do not need to judge the trajectory direction and data gradient for obviously invalid or duplicate endpoint combinations, thereby reducing the number of invalid matches in link rearrangement and improving the stability of subsequent target endpoint pair screening. Among them, the basic conflict elimination only deletes endpoint pairings that obviously do not meet the continuity conditions, and does not replace the subsequent screening of trajectory direction continuity status and data gradient continuity status.

[0066] It should be noted that, through the above S2 step, the multiple data analysis sub-chains obtained in S1 can be converted into an endpoint combination structure that can be used for continuity judgment; the starting endpoint parameter group and the ending endpoint parameter group retain the endpoint position, endpoint acquisition order and endpoint associated acquisition data, so that the candidate endpoint pairs have a basis for spatial continuity judgment, as well as a basis for acquisition order conflict elimination and data gradient continuation judgment; this step provides a clear technical object for the subsequent elimination of candidate endpoint pairs with opposite trajectory directions or data gradients that exceed the limits, avoiding misconnections caused by directly restoring the warehouse inspection data link according to the original sampling order.

[0067] S3. Based on the trajectory direction continuity and data gradient continuity between candidate endpoint pairs, eliminate candidate endpoint pairs with opposite trajectory directions or out-of-bounds data gradients, and retain the target endpoint pairs. Note that the following should be noted in this step: S3.1 For each candidate endpoint pair, take the trajectory direction of the last segment of the data analysis sub-chain to which the terminating endpoint belongs, and take the trajectory direction of the first segment of the data analysis sub-chain to which the starting endpoint belongs.

[0068] Specifically, for each candidate endpoint pair, first determine the data analysis sub-chain to which the termination endpoint parameter group in the candidate endpoint pair belongs, and select the previous trajectory point adjacent to the termination endpoint at the end of the data analysis sub-chain. The direction of the line connecting the position of the previous trajectory point and the position of the termination endpoint is taken as the direction of the final trajectory segment. Then determine the data analysis sub-chain to which the starting endpoint parameter group in the candidate endpoint pair belongs, and select the next trajectory point adjacent to the starting endpoint at the beginning of the data analysis sub-chain. The direction of the line connecting the position of the starting endpoint and the position of the next trajectory point is taken as the direction of the first trajectory segment.

[0069] In blockchain supply chain warehouse inspection scenarios, the direction of the final trajectory segment indicates the direction of travel of the inspection terminal when it leaves the previous continuous storage location segment, and the direction of travel of the first trajectory segment indicates the direction of travel of the inspection terminal when it enters the next continuous storage location segment. If the two directions are basically consistent, it usually indicates that the two data analysis sub-chains may belong to the continuous collection results of the same shelf aisle or adjacent storage location segments. If the two directions are nearly opposite, it usually indicates that the inspection terminal has looped back or the candidate endpoints have crossed the data chain segments with opposite inspection directions. When the number of trajectory points of the data analysis sub-chain is insufficient to determine the direction of the final trajectory segment or the direction of the first trajectory segment, the corresponding trajectory direction can be determined by combining the positions of the adjacent endpoints of the data analysis sub-chain and the direction of the shelf aisle.

[0070] S3.2. Compare the angle between the direction of the last segment of the trajectory and the direction of the first segment of the trajectory to obtain the trajectory direction continuity state, and mark the candidate endpoint pairs whose angle exceeds the direction boundary as direction reversal endpoint pairs.

[0071] Specifically, the trajectory direction continuity state of the candidate endpoint pair is determined based on the directional angle between the trajectory direction of the last segment and the trajectory direction of the first segment. The directional angle is the smaller angle between the trajectory direction of the last segment and the trajectory direction of the first segment, and its value ranges from 0° to 180°. The smaller the directional angle, the closer the trajectory directions of the two data analysis sub-chains are; the larger the directional angle, the greater the deviation in the trajectory directions of the two data analysis sub-chains.

[0072] Furthermore, the angle comparison specifically includes: determining the direction of the final trajectory segment by connecting the positions of adjacent trajectory points at the end of the data analysis subchain to which the termination endpoint belongs; determining the direction of the first trajectory segment by connecting the positions of adjacent trajectory points at the beginning of the data analysis subchain to which the starting endpoint belongs; calculating the angle between the direction of the final trajectory segment and the direction of the first trajectory segment; marking the candidate endpoint pair as a direction continuation endpoint pair when the angle is less than or equal to the first angle boundary; marking the candidate endpoint pair as a direction deflection endpoint pair when the angle is greater than the first angle boundary and less than the second angle boundary; and marking the candidate endpoint pair as a direction reversal endpoint pair when the angle is greater than or equal to the second angle boundary.

[0073] The first angle boundary is 30° to 60°, preferably 45°; the second angle boundary is 120° to 150°, preferably 135°. In the warehouse racking aisle scenario, when the included angle is less than or equal to 45°, it indicates that the two data analysis sub-chains have a good continuity relationship in the racking aisle direction and can be used as the priority object for subsequent link continuation. When the included angle is greater than 45° and less than 135°, it indicates that the two data analysis sub-chains may correspond to aisle turns, rack end detours, or diagonal cross-area movements, and further judgment is needed in combination with the data gradient continuity status. When the included angle is greater than or equal to 135°, it indicates that the two data analysis sub-chains have a reverse continuation relationship and are marked as reverse direction connection point pairs. The first angle boundary and the second angle boundary can be determined according to the turning angle of the warehouse racking aisle, the aisle intersection angle, and the allowable turning range of the inspection terminal.

[0074] S3.3 Calculate the cross-endpoint data gradient of the candidate endpoint pair based on the endpoint association data collected in the termination endpoint parameter group and the endpoint association data collected in the starting endpoint parameter group.

[0075] Specifically, for each candidate endpoint pair, the endpoint-related data collected from the termination endpoint parameter group and the endpoint-related data collected from the start endpoint parameter group are extracted. The difference between the endpoint-related data collected from the start endpoint parameter group and the endpoint-related data collected from the termination endpoint parameter group is calculated, and combined with the endpoint distance between the termination endpoint and the start endpoint, the cross-endpoint data gradient of the candidate endpoint pair is obtained. The cross-endpoint data gradient represents the transition amplitude of the collected data at the candidate continuation position of the two data analysis subchains.

[0076] In a preferred embodiment, when the endpoint-associated data is of a single data type, the cross-endpoint data gradient is the ratio of the difference between the two endpoint-associated data to the endpoint distance; when the endpoint-associated data includes multiple data types such as cargo location confidence, electronic tag matching strength, temperature value, and humidity value, the individual cross-endpoint data gradients of each type of data are calculated separately, and the cross-endpoint data gradient is synthesized according to the normalized individual cross-endpoint data gradients.

[0077] Preferably, this method allows the cross-endpoint data gradient to simultaneously reflect the continuity of the cargo location status, electronic tag signal, and environmental sensor status at the endpoint connection position; when the endpoint distance between the terminating endpoint and the starting endpoint is less than the positioning error range, the endpoint distance can be corrected to the minimum distance corresponding to the positioning error range, so as to avoid abnormal amplification of the cross-endpoint data gradient due to the endpoint distance being too small.

[0078] For example, within the same shelving aisle, the temperature and humidity data of adjacent storage locations usually change slowly, and the signal strength of electronic tags also exhibits a continuous attenuation characteristic as distance changes. If the endpoints of a candidate endpoint pair are close in distance but the goods show significant jumps in location reliability, electronic tag matching strength, or temperature and humidity data, it indicates that the candidate endpoint pair may not be a reasonable continuation relationship of the same continuous storage location segment.

[0079] S3.4. Compare the difference between the cross-endpoint data gradient and the last segment data gradient of the data analysis sub-chain to which the termination endpoint belongs, and the first segment data gradient of the data analysis sub-chain to which the starting endpoint belongs, to obtain the data gradient continuation state.

[0080] Specifically, the gradient of the last segment of data is calculated by first collecting data associated with the endpoints of adjacent trajectory points at the end of the data analysis sub-chain to which the termination endpoint belongs; then, the gradient of the first segment of data is calculated by collecting data associated with the endpoints of adjacent trajectory points at the beginning of the data analysis sub-chain to which the starting endpoint belongs. The gradient of the last segment of data represents the trend of data change of the previous data analysis sub-chain near the termination endpoint, and the gradient of the first segment of data represents the trend of data change of the next data analysis sub-chain near the beginning endpoint.

[0081] Subsequently, the gradient of the cross-endpoint data is compared with the gradient of the last segment and the gradient of the first segment respectively. When the difference between the gradient of the cross-endpoint data and the gradient of the last segment is small, and the difference between the gradient of the cross-endpoint data and the gradient of the first segment is also small, it indicates that the two data analysis sub-chains have a continuous data gradient relationship at the candidate continuation position. When either difference is significantly large, it indicates that the data gradient between the two data analysis sub-chains does not have continuity, and there may be incorrect continuation across shelves, aisles or state segments.

[0082] In the blockchain supply chain warehouse inspection scenario, the data gradient continuity status is used to reflect the transition relationship of the location status data along the shelf arrangement direction; for example, the temperature data in the same cold chain shelf usually changes continuously along the aisle direction; if the cross-endpoint data gradient of a candidate endpoint pair deviates significantly from the local data gradient of the preceding and following data analysis sub-chains, then the candidate endpoint pair should not be used as the continuation relationship of the location data link; the final data gradient and the first data gradient are calculated using the same data type and normalization method as the cross-endpoint data gradient to maintain the consistency of the scale of the difference comparison.

[0083] S3.5 Mark candidate endpoint pairs whose difference exceeds the gradient boundary as gradient out-of-bounds endpoint pairs, and delete reverse direction endpoint pairs and gradient out-of-bounds endpoint pairs.

[0084] Specifically, the gradient boundary is determined based on the allowable gradient fluctuation range of data collected from adjacent storage locations within the same shelving aisle; for storage location image recognition data, the gradient boundary can be determined based on the normal differences in location reliability between adjacent storage locations; for electronic tag signal data, the gradient boundary can be determined based on the normal attenuation amplitude of electronic tag signal strength between adjacent storage locations; for environmental sensing data, the gradient boundary can be determined based on the normal spatial variation amplitude of temperature and humidity within the same storage area or the same shelving aisle.

[0085] When the difference between the gradient of the cross-endpoint data and the gradient of the last segment data exceeds the gradient boundary, or when the difference between the gradient of the cross-endpoint data and the gradient of the first segment data exceeds the gradient boundary, the candidate endpoint pair is marked as a gradient out-of-bounds endpoint pair; then, the marked reverse-direction endpoint pairs and gradient out-of-bounds endpoint pairs are deleted.

[0086] Preferably, this process can simultaneously eliminate two types of unreasonable continuity relationships: one type is the combination of endpoints that are obviously opposite in the trajectory direction, i.e., the reverse direction endpoint pair; the other type is the combination of endpoints that may be close in trajectory direction but the data collection transition is discontinuous, i.e., the gradient out-of-bounds endpoint pair; thus, the retained endpoint pairs satisfy both the continuity relationship of the shelf aisle inspection direction and the continuity relationship of the location data collection along the spatial arrangement direction.

[0087] S3.6. Select the candidate endpoint pairs that have not been deleted as the target endpoint pairs.

[0088] Specifically, after deleting the reverse direction endpoint pairs and gradient out-of-bounds endpoint pairs, the remaining candidate endpoint pairs are taken as target endpoint pairs. The target endpoint pair represents the combination of endpoints that have a continuation relationship between two data analysis subchains, which is verified by both the trajectory direction continuation state and the data gradient continuation state.

[0089] In a preferred embodiment, if the same data analysis subchain corresponds to multiple candidate endpoint pairs that have not been deleted, then the candidate endpoint pair with the trajectory direction continuation state being the direction continuation endpoint pair and the data gradient continuation state being the closest is retained as the priority continuation object; if there is no direction continuation endpoint pair, then the candidate endpoint pair with the closest data gradient continuation state is selected from the direction deflection endpoint pairs as the target endpoint pair.

[0090] Preferably, this processing can adapt to scenarios such as turning at the end of the warehouse rack and switching of aisles, while avoiding the inclusion of endpoint combinations with reversed direction or abnormal gradients in subsequent link rearrangement; the closest data gradient continuity state can be determined by the difference between the cross-endpoint data gradient and the final segment data gradient, and the difference between the cross-endpoint data gradient and the first segment data gradient.

[0091] It should be noted that, through the above S3 step, this embodiment filters candidate endpoint pairs from two dimensions: trajectory geometry and data acquisition gradient, eliminating erroneous continuity relationships that are difficult to eliminate based solely on endpoint position or acquisition order. For blockchain supply chain warehouse inspection data, this step can reduce data link misconnections caused by rack aisle loops, cross-aisle jumps, repeated acquisition, or sudden changes in storage location status, making the subsequently generated verification data analysis link more consistent with the actual storage location arrangement order and the continuity of acquired data.

[0092] S4. Adjust the connection order of multiple data analysis sub-chains according to the target endpoints to generate a verification data analysis link, and rearrange the collected data within the data analysis sub-chains according to the verification data analysis link. Note that the following should be noted in this step: S4.1. Using each data analysis sub-chain as a link node, and using the pairing direction from the end endpoint to the start endpoint in the target endpoint pair as the connection direction between link nodes, generate the sub-chain connection relationship.

[0093] Specifically, each data analysis subchain is treated as a link node. The link node retains the order of trajectory points, start endpoint, end endpoint, and collected data within the data analysis subchain. For each target endpoint pair, the data analysis subchain to which the end endpoint belongs is taken as the preceding link node, and the data analysis subchain to which the start endpoint belongs is taken as the following link node. The direction from the end endpoint to the start endpoint is determined as the connection direction between the two link nodes, thereby generating the subchain connection relationship.

[0094] In the blockchain supply chain warehouse inspection scenario, a data analysis subchain can correspond to a continuous storage location segment, and a subchain connection relationship indicates the connection relationship between the previous continuous storage location segment and the next continuous storage location segment. For example, if the termination endpoint of the first data analysis subchain and the starting endpoint of the third data analysis subchain form a target endpoint pair, then the connection relationship between the first data analysis subchain and the third data analysis subchain is determined as a subchain connection relationship.

[0095] Preferably, this method can reorganize multiple data chain segments that have been broken at the trajectory sequence break points into a link structure with directional constraints; the sub-chain connection relationship includes the previous link node, the next link node, and the connection direction, and the subsequent arrangement order of each data analysis sub-chain is determined according to the sub-chain connection relationship.

[0096] S4.2 When there are multiple sub-chain connections in the same data analysis sub-chain, retain the sub-chain connection relationship where the trajectory direction continuation state is the direction continuation endpoint pair and the data gradient continuation state is closest.

[0097] When multiple target endpoints simultaneously involve the same data analysis subchain, it is possible that the same link node corresponds to multiple subsequent link nodes, or the same link node corresponds to multiple preceding link nodes. To avoid a data analysis subchain being repeatedly connected to different link positions, this embodiment resolves conflicts in the connection relationships of multiple subchains of the same data analysis subchain.

[0098] When multiple sub-chain connections exist within the same data analysis sub-chain that can serve as subsequent continuations, the sub-chain connections where the trajectory direction continuation state is a pair of direction continuation endpoints are prioritized for retention. If multiple sub-chain connections where the trajectory direction continuation state is a pair of direction continuation endpoints, the data gradient continuation states of each sub-chain connection are compared, and the sub-chain connection with the closest data gradient continuation state is retained. The closest data gradient continuation state refers to a continuation state where the difference between the cross-endpoint data gradient and the final segment data gradient, and the first segment data gradient, is small and does not exceed the gradient boundary. When comparing the data gradient continuation states of multiple sub-chain connections, the difference between the cross-endpoint data gradient and the final segment data gradient, and the difference between the cross-endpoint data gradient and the first segment data gradient, can be added together. The sub-chain connection with the smaller sum of differences is considered the sub-chain connection with the closest data gradient continuation state.

[0099] When multiple preceding subchain connections exist within the same data analysis subchain, the above rules are applied to filter them, ensuring that each data analysis subchain retains only one preceding subchain connection and one subsequent subchain connection. For directional deflection endpoint pairs, they are retained only when no directional continuation endpoint pairs exist, and the closest data gradient continuation state is still used as the filtering criterion. Endpoint combinations that have already been marked as directional reversal endpoint pairs or gradient out-of-bounds endpoint pairs have been deleted in step S3 and will no longer participate in the connection filtering in this step.

[0100] In warehouse inspection scenarios, this processing can solve the problem of multiple suspicious connection relationships generated after the inspection terminal goes around the end of the shelf, passes through the intersection of the aisle repeatedly, or partially goes around. By prioritizing the selection of sub-chain connection relationships with continuous direction and continuous data gradient, the risk of the same storage location segment being incorrectly connected to multiple analysis links can be reduced.

[0101] S4.3 Arrange multiple data analysis sub-chains sequentially according to their sub-chain connection relationships to generate a verification data analysis link.

[0102] Specifically, after resolving conflicts in the sub-chain connections, the order of each link node is determined based on the sub-chain connections. If a data analysis sub-chain does not have a preceding sub-chain connection, then that data analysis sub-chain is used as the starting link node for verifying the data analysis link. Subsequently, the next link node is searched sequentially along the sub-chain connections until there is no subsequent sub-chain connection. The resulting sequence of data analysis sub-chains is then used as the verification data analysis link.

[0103] If there are multiple data analysis subchains that do not have a preceding subchain connection relationship, multiple verification data analysis links can be generated respectively, and each verification data analysis link corresponds to an independent continuous storage location inspection section; if there is a data analysis subchain that is not connected by any subchain connection relationship, the data analysis subchain is retained as an independent verification data analysis link and is treated as a link to be verified in subsequent warehouse inspection data analysis.

[0104] Preferably, this method can avoid forcibly continuing data analysis subchains without reliable endpoint matching relationships, thereby reducing the impact of faulty chains on the data order before blockchain notarization.

[0105] In a preferred embodiment, the verification data analysis links can be represented according to the link node number, the preceding link node number, the subsequent link node number, and the link direction sequence. For warehouse racking aisles, the verification data analysis links reflect the arrangement order of the storage location data after trajectory sequence break verification. This differs from the original sampling order and is an analysis order formed by the combined constraints of the trajectory direction continuity state and the data gradient continuity state. When multiple verification data analysis links are generated, they can be distinguished according to racking aisle identifiers, storage location area identifiers, or the starting and ending points of the links.

[0106] S4.4. Retain the collected data in each data analysis subchain according to the original trajectory point arrangement order, and rearrange the collected data in different data analysis subchains according to the verification data analysis link.

[0107] Specifically, for each data analysis subchain, since its internal trajectory points are not divided by the trajectory sequence break position, it indicates that the data analysis subchain has good spatial continuity and data collection continuity. Therefore, the original trajectory point arrangement order of the collected data within the data analysis subchain is retained. For the arrangement relationship between different data analysis subchains, the original sampling order is no longer used, but is rearranged according to the link node order determined by the verification data analysis link.

[0108] During the rearrangement process, the collected data in the first data analysis sub-chain of the verification data analysis link is arranged first, and the collected data in the next data analysis sub-chain connected to it is arranged last. The collected data in the remaining data analysis sub-chains are arranged in sequence according to the verification data analysis link. For multiple independent verification data analysis links, corresponding data rearrangement results are formed respectively. For links to be verified, an independent identifier can be retained, and manual verification or re-collection can be prompted in the subsequent blockchain notarization pre-verification process.

[0109] In blockchain supply chain warehouse inspection scenarios, the rearranged collected data can serve as the basis for location status analysis, abnormal location location, environmental anomaly segment determination, and data summary calculation before blockchain notarization. For example, during aisle inspection, if the inspection terminal experiences a trajectory break due to a weak wireless signal area, the original sampling sequence might continue the data from the third aisle segment after the first aisle segment. After generating the verification data analysis link through this step, the data from the second aisle segment can be reconnected between the first and third aisle segments, ensuring that the rearranged collected data order matches the actual location arrangement. Before blockchain notarization, a data summary can be calculated on the collected data rearranged according to the verification data analysis link, and the data summary, the link identifier of the verification data analysis link, and the identifier of the trajectory sequence break point can be submitted to the blockchain node.

[0110] It should be noted that through the above S4 step, multiple data analysis sub-chains are transformed from a scattered state into a verification data analysis link with a connection direction and arrangement order. This step not only preserves the original continuity of the data collected within each data analysis sub-chain, but also corrects the order deviation caused by trajectory breaks, loops, or misconnections between different data analysis sub-chains, thereby providing a consistent and traceable data collection foundation for subsequent warehouse inspection big data analysis and blockchain evidence storage preprocessing.

[0111] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A data analysis method based on the verification of the acquisition trajectory, characterized in that, include: The location where at least one of the trajectory spacing and acquisition sequence interval of adjacent trajectory points in the acquisition trajectory does not meet the continuous boundary, and the residual of the trajectory point associated acquisition data does not meet the residual boundary, is taken as the trajectory sequence break position, and the trajectory sequence break position is used as the boundary to split into multiple data analysis sub-chains; Extract the start and end endpoints of each of the data analysis subchains, and pair the end endpoints of each of the data analysis subchains with the start endpoints of the remaining data analysis subchains one by one to obtain candidate endpoint pairs; Based on the trajectory direction continuity and data gradient continuity between the candidate endpoint pairs, candidate endpoint pairs with opposite trajectory directions or data gradients that exceed the limits are eliminated, and the target endpoint pairs are retained. Adjust the connection order of multiple data analysis sub-chains according to the target endpoints, generate a verification data analysis link, and rearrange the collected data in the data analysis sub-chains according to the verification data analysis link.

2. The data analysis method based on acquisition trajectory verification according to claim 1, characterized in that, The methods for obtaining the trajectory spacing, the acquisition sequence interval, and the residual of the trajectory point associated acquisition data include: The trajectory points in the collected trajectory are arranged according to the sampling order, and two adjacent trajectory points are formed into adjacent trajectory point pairs. Calculate the spatial distance between two trajectory points in the adjacent trajectory point pair to obtain the trajectory spacing, and use the difference in the sequence number of the two trajectory points in the sampling order as the sampling sequence interval; Extract the collected data associated with the two trajectory points in the adjacent trajectory point pair, and calculate the difference between the two collected data to obtain the trajectory point associated collected data residual.

3. The data analysis method based on acquisition trajectory verification according to claim 2, characterized in that, Determining the location of the trajectory break includes: The trajectory spacing is compared with the continuous distance boundary, the acquisition sequence interval is compared with the continuous sequence boundary, and the trajectory point associated acquisition data residual is compared with the residual boundary. When the trajectory spacing is greater than the continuous distance boundary, or the acquisition sequence interval is greater than the continuous sequence boundary, and the residual of the trajectory point associated acquisition data is greater than the residual boundary, the position between the adjacent trajectory point pairs is taken as the trajectory sequence break position.

4. The data analysis method based on acquisition trajectory verification according to claim 3, characterized in that, The data is divided into multiple sub-chains for analysis, using the break points in the trajectory sequence as boundaries. The starting position of the acquisition trajectory, the break position of the trajectory sequence, and the ending position of the acquisition trajectory are used as segment boundaries, and the trajectory points and their associated acquisition data between two adjacent segment boundaries are grouped into a data analysis sub-chain.

5. The data analysis method based on acquisition trajectory verification according to claim 1, characterized in that, Extract the start and end endpoints of each of the data analysis subchains, including: According to the arrangement order of trajectory points in the data analysis sub-chain, the trajectory point at the beginning is determined as the starting endpoint, and the trajectory point at the end is determined as the ending endpoint. Extract the endpoint position, endpoint acquisition sequence, and endpoint associated acquisition data of the starting endpoint to form a starting endpoint parameter group; Extract the endpoint position, endpoint acquisition sequence, and endpoint associated acquisition data of the termination endpoint to form a termination endpoint parameter group.

6. The data analysis method based on acquisition trajectory verification according to claim 5, characterized in that, Obtaining the candidate endpoint pair includes: Each of the termination endpoint parameter groups is combined one by one with the starting endpoint parameter groups that do not belong to the same data analysis subchain to form endpoint pairing groups; Delete endpoint pairs that have overlapping endpoint acquisition order or the same endpoint location, and keep the remaining endpoint pairs as candidate endpoint pairs.

7. The data analysis method based on acquisition trajectory verification according to claim 6, characterized in that, Retaining the target endpoint pair includes: For each candidate endpoint pair, take the trajectory direction of the last segment of the data analysis subchain to which the terminating endpoint belongs, and take the trajectory direction of the first segment of the data analysis subchain to which the starting endpoint belongs; The angle between the direction of the last segment trajectory and the direction of the first segment trajectory is compared to obtain the trajectory direction continuity state, and the candidate endpoint pairs whose angle exceeds the direction boundary are marked as direction reverse endpoint pairs. Based on the endpoint association data collected in the termination endpoint parameter group and the endpoint association data collected in the starting endpoint parameter group, calculate the cross-endpoint data gradient of the candidate endpoint pair; The data gradient across endpoints is compared with the data gradient of the last segment of the data analysis subchain to which the termination endpoint belongs and the data gradient of the first segment of the data analysis subchain to which the starting endpoint belongs, respectively, to obtain the data gradient continuation state. The candidate endpoint pairs whose difference exceeds the gradient boundary are marked as gradient out-of-bounds endpoint pairs, and the reverse direction endpoint pairs and the gradient out-of-bounds endpoint pairs are deleted; The candidate endpoint pairs that have not been deleted are used as target endpoint pairs.

8. The data analysis method based on acquisition trajectory verification according to claim 7, characterized in that, The angle comparison includes: Based on the position of the adjacent trajectory points at the end of the data analysis sub-chain to which the termination endpoint belongs, the direction of the final trajectory segment is determined by connecting the positions of the adjacent trajectory points at the end of the sub-chain. Based on the positional connection of adjacent trajectory points in the data analysis subchain to which the starting endpoint belongs, the direction of the first trajectory segment is determined. Calculate the angle between the direction of the final trajectory segment and the direction of the first trajectory segment; When the included angle of the directions is less than or equal to the first angle boundary, the candidate endpoint pair is marked as a direction continuation endpoint pair; When the included angle of direction is greater than the first angle boundary and less than the second angle boundary, the candidate endpoint pair is marked as a direction deflection endpoint pair; When the included angle of the directions is greater than or equal to the second angle boundary, the candidate endpoint pair is marked as the reversed direction endpoint pair.

9. The data analysis method based on acquisition trajectory verification according to claim 8, characterized in that, The first angular boundary is 30° to 60°; the second angular boundary is 120° to 150°.

10. The data analysis method based on acquisition trajectory verification according to claim 1, characterized in that, Generating the verification data analysis link includes: Each of the data analysis sub-chains is used as a link node, and the pairing direction from the end endpoint to the start endpoint in the target endpoint pair is used as the connection direction between the link nodes to generate a sub-chain connection relationship. When there are multiple subchain connection relationships in the same data analysis subchain, the subchain connection relationship with the closest trajectory direction continuation state and data gradient continuation state is retained. Multiple data analysis sub-chains are arranged sequentially according to the sub-chain connection relationship to generate a verification data analysis link; The collected data in each of the data analysis sub-chains is retained in the original order of the trajectory points, and the collected data in different data analysis sub-chains are rearranged according to the verification data analysis link.

Images