Bulk material entering and leaving yard and inventory collaborative management method and system based on end-cloud collaboration

Abstract

This invention discloses a method and system for collaborative management of bulk material entry and exit and inventory based on edge-cloud collaboration, belonging to the field of bulk material logistics and inventory management technology. It obtains the first stable weighing segment and vehicle axle sequence segment cached at the weighbridge edge terminal, generates a pending closure entry token, and establishes a shadow inventory unit isolated from available inventory in the cloud. It generates an unloading witness segment by combining material change segments formed by the unloading point edge terminal. It constructs candidate closed chains for back-weighing related conflicts, and the edge terminal replays locally cached data, selecting a unique closed chain based on the direction of the net weighing difference and the correspondence between the vehicle axle sequence. The cloud then converts the actual inventory accordingly and synchronously updates the vehicle release status and available outbound quantity. This invention not only directly avoids inflated inventory, premature release of outbound quotas, and erroneous vehicle releases, but also synchronously updates the vehicle release status and available outbound quantity.

Description

Technical Field

[0001] This invention relates to the field of bulk material logistics and inventory management technology, specifically to a method and system for collaborative management of bulk material entry and exit and inventory based on end-to-end cloud collaboration. Background Technology

[0002] When bulk materials arrive at the site, vehicles typically complete access control verification, initial weighing, unloading at the material shed, and final weighing in sequence. In weak network environments, the weighbridge terminal may cache records and upload them centrally once the network is restored. If, during the unloading of purchased vehicles, on-site transport vehicles temporarily borrow the weighbridge, and vehicle identification is affected by trailer changes, obscured license plates, or disordered upload sequences, the cloud can easily combine weighing records from different vehicles into an incorrect closed loop. This incorrect closed loop superficially meets the weighing process requirements but distorts inventory records and further affects vehicle release and outbound scheduling. Existing methods typically associate data based on license plates, order numbers, or upload times, making it difficult to identify cross-vehicle pseudo-closed loops that have formed a complete business chain. Summary of the Invention

[0003] The purpose of this invention is to provide a method and system for collaborative management of bulk material entry and exit and inventory based on end-to-end cloud collaboration, so as to solve the shortcomings of the prior art.

[0004] To achieve the above objectives, the present invention provides the following technical solution: a method for collaborative management of bulk material entry and exit and inventory based on edge-cloud collaboration, comprising:

[0005] Obtain the first weighing stability segment and vehicle axle sequence segment cached by the weighbridge edge terminal in local incremental order, and generate a token to be closed for storage.

[0006] Upload the pending closure entry token to the cloud, and establish a shadow inventory unit isolated from the available inventory for the pending closure entry token;

[0007] Obtain material change segments formed at the edge terminal of the unloading point during the vehicle's stay, and attach the material change segments to the warehousing token to be closed to form an unloading witness segment.

[0008] Obtain stable back-weighing segments cached at the edge terminal of the weighbridge; when a stable back-weighing segment can be associated with multiple pending closure entry tokens, or when the upload order is inconsistent with the local incremental order, the cloud generates multiple candidate closure chains based on the spatial order of access control, first weighing, unloading and back-weighing, and according to the rule that a unloading witness segment can only be occupied by one pending closure entry token.

[0009] The candidate closed chain is sent to the corresponding edge terminal, and the corresponding edge terminal replays the locally cached weighing stability segment, vehicle axle sequence segment and material change segment. Candidate closed chains that are inconsistent with the direction of material change or whose vehicle axle sequence cannot be matched end to end are eliminated to obtain a unique closed chain.

[0010] Based on the unique closed chain, the cloud platform converts the corresponding shadow inventory unit into actual inventory changes and synchronously updates the vehicle release status and available quantity for outbound shipment.

[0011] Preferably, the token to be closed for warehousing includes: a local incrementing sequence number, a stable weight value, a vehicle axle sequence summary, and the time when the first weighing occurred.

[0012] Preferably, establishing a shadow inventory unit isolated from available inventory includes:

[0013] The cloud uses the edge terminal number and the local incrementing sequence number to form a joint verification condition to perform duplicate transmission identification on the received token to be closed in the database; when a corresponding temporary storage entry already exists, only the most recent reception time and reception count are updated.

[0014] When there is no corresponding temporary storage entry, the cloud creates a unique shadow inventory unit and writes the shadow inventory unit into an isolated storage area that is independent of the formal inventory ledger, so that the weight snapshot in it does not participate in the real-time available inventory calculation.

[0015] The cloud establishes the first index layer based on the edge terminal number, the second index layer based on the local incremental sequence number, and the third index layer based on the axle sequence retrieval tags formed by arranging the vehicle's direction of entry, the number of axles, and the trigger time interval between adjacent axles.

[0016] The cloud enables the same shadow inventory unit to be simultaneously written into a sequential chain arranged according to the local incremental sequence number and the corresponding axis sequence candidate retrieval group, which is used to restore the initial weighing formation order when the offline retransmission order is disordered.

[0017] Preferably, the unloading witness segment includes:

[0018] Based on the change in geomagnetic occupancy edge and axle triggering sequence formed when the vehicle enters the unloading position, the edge terminal of the unloading point determines the start time when the vehicle fully occupies the unloading position and the end time when the vehicle leaves the unloading position, and the time interval between the two is determined as the vehicle dwell window.

[0019] Within the vehicle parking window, the edge terminal of the unloading point continuously acquires the changes in the bearing surface of the unloading position, the continuous period of hopper vibration, and the continuous period of dust obstruction, and deletes the change records that occur before the vehicle fully occupies the unloading position or after the vehicle leaves the unloading position.

[0020] At the edge of the unloading point, the records of changes in the bearing surface that are consistent in direction, overlap with the duration of the hopper vibration, and are covered by the duration of dust obstruction are merged into a material change segment.

[0021] Preferably, multiple candidate closed chains are generated, including:

[0022] The cloud receives stable back-to-tape weighing segments cached by the edge terminal of the weighbridge and establishes a back-to-tape weighing record pool according to the edge terminal number, upload batch and local incremental sequence number;

[0023] Each stable segment of weighing return is associated with multiple pending closure tokens in the axle sequence candidate retrieval group. When the same stable segment of weighing return can be associated with multiple pending closure tokens, or when its upload order is inconsistent with the local increment order of the corresponding weighbridge edge terminal, the stable segment of weighing return is marked as a conflict segment.

[0024] Based on the spatial sequence of access control records, initial weighing stabilization segment, vehicle parking window, attached unloading witness segment, and conflict segment, a preliminary closed chain is established, consisting of the initial weighing stabilization segment, the unloading witness segment, and the return weighing stabilization segment in sequence.

[0025] According to the rule that a single unloading witness segment can only be occupied by one preliminary closed chain, preliminary closed chains that satisfy spatial order but have different matching combinations are retained as multiple candidate closed chains.

[0026] Preferably, the candidate closed chain is sent to the corresponding edge terminal, including:

[0027] The cloud reads the local occurrence time and the most recent clock calibration offset recorded by the edge terminal of the weighbridge and the edge terminal of the unloading point, and converts the local occurrence time of the corresponding segments of each candidate closed chain into a unified playback timing sequence.

[0028] When no new clock calibration offset is obtained during a network outage, the cloud uses the most recent valid clock calibration offset before the network outage and adds a mark indicating that it needs to be re-verified to the corresponding candidate closed chain.

[0029] The weighbridge edge terminal uses the stable start edge of the first weighing stable segment as the front playback anchor point and the stable end edge of the return weighing stable segment as the rear playback anchor point. It reads the weighing stable segment and vehicle axle sequence segment located between the two playback anchor points according to the local writing order.

[0030] The edge terminal at the unloading point reads the corresponding material change segment and vehicle parking window based on the unloading witness segment identifier, and marks the material change segment whose start time or end time is outside the vehicle parking window as a boundary abnormal segment.

[0031] Preferably, candidate closed chains are eliminated, including:

[0032] The weighbridge edge terminal assigns the trigger time intervals of adjacent axles in the first weighing phase to a pre-set continuous time interval according to the order in which the axles pass the detection position, and arranges them sequentially to form the first axle sequence fingerprint; the trigger time intervals of adjacent axles in the return weighing phase are formed into the tail axle sequence fingerprint in the same way.

[0033] When the edge terminal of the weighbridge determines that the direction of travel for the first weighing stage is opposite to that for the return weighing stage based on the direction of travel marking, the rear axle sequence fingerprint is reversed and compared with the front axle sequence fingerprint item by item; when the directions of travel are the same, the fingerprints are compared item by item in the original order.

[0034] For inbound unloading operations, when the stable weight value of the first stable weighing segment is greater than the stable weight value of the return weighing segment, the weighing net difference direction is determined as the unloading direction; when the bearing surface change direction is the increasing direction, the material change direction is determined as the unloading direction.

[0035] Candidate closed chains are removed from the cloud if the direction of net weighing difference does not correspond to the direction of material change, the first and last axis sequence fingerprints cannot be matched, or the material change segment is marked as a boundary abnormal segment.

[0036] Preferably, a unique closed chain is obtained, including:

[0037] The cloud performs sequential continuous verification on the candidate closed chains that have not been eliminated. The sequential continuous verification includes: in the local write records of the same weighbridge edge terminal, the first weighing stable segment is formed before the return weighing stable segment; the vehicle dwell window corresponding to the unloading witness segment is located between the first weighing stable segment and the return weighing stable segment; there are no other pending closure entry tokens that have occupied the same unloading witness segment.

[0038] When only one candidate closed chain that satisfies the sequential continuous verification is retained, the cloud determines it as the unique closed chain and writes the unique closed chain identifier into the conversion record field of the corresponding shadow inventory unit;

[0039] When more than two candidate closed chains are retained, the cloud maintains the isolation status of the corresponding shadow inventory unit and marks it as pending manual review;

[0040] When all candidate closed chains are eliminated, a closure failure flag is generated in the cloud, and the corresponding shadow inventory unit is prohibited from being converted into a formal inbound record.

[0041] Preferably, the corresponding shadow inventory unit is converted into actual inventory changes and the vehicle release status and available quantity for outbound shipment are updated synchronously, including:

[0042] The cloud generates an inventory conversion token that can only be used once based on a unique closed chain, and writes the unique closed chain identifier, shadow inventory unit identifier, vehicle identifier, storage location identifier, and conversion version into the inventory conversion token in a fixed order;

[0043] After confirming that the inventory conversion token has not been used, the cloud will convert the corresponding shadow inventory unit into a formal inbound record and create an available quota fragment in the managed quota pool that corresponds one-to-one with the formal inbound record. The available quota fragment carries the quota weight, remaining quota weight, occupancy status and conversion version.

[0044] The cloud establishes a causal barrier based on the sequential dependency of the formal entry record writing event, the available quota fragment creation event, and the vehicle release event. Only after both the formal entry record writing event and the available quota fragment creation event are completed will the corresponding vehicle release status be updated to allow release.

[0045] When a duplicate inventory conversion token is detected, an available quota segment is already occupied, or any preceding event is not completed, the cloud will not repeatedly increase the actual inventory, will not repeatedly release the available outbound quantity, will keep the vehicle release status unchanged, and will write the corresponding update request to the compensation queue.

[0046] This invention also provides a collaborative management system for bulk material entry and exit and inventory based on edge-cloud collaboration, including:

[0047] The warehouse entry token generation module obtains the first weighing stability segment and vehicle axle sequence segment cached by the weighbridge edge terminal in local incremental order, and generates a warehouse entry token to be closed.

[0048] The shadow inventory isolation establishment module uploads the pending closure entry token to the cloud and establishes a shadow inventory unit isolated from the available inventory for the pending closure entry token;

[0049] The unloading witness segment formation module acquires the material change segment formed at the edge terminal of the unloading point during the vehicle's stay, and attaches the material change segment to the warehousing token to be closed to form the unloading witness segment.

[0050] The candidate closed chain generation module obtains the stable back-taper weighing segments cached by the weighbridge edge terminal. When a stable back-taper weighing segment can be associated with multiple pending closure entry tokens, or when the upload order is inconsistent with the local incremental order, the cloud generates multiple candidate closed chains based on the spatial order of access control, first weighing, unloading, and back-taper weighing, and according to the rule that a unloading witness segment can only be occupied by one pending closure entry token.

[0051] The unique closed chain determination module sends the candidate closed chains to the corresponding edge terminals. The corresponding edge terminals replay the locally cached weighing stability segments, vehicle axle sequence segments, and material change segments, and eliminate candidate closed chains that are inconsistent with the direction of material change or whose vehicle axle sequence cannot be matched end to end, thus obtaining the unique closed chain.

[0052] The collaborative status update module converts the corresponding shadow inventory unit into actual inventory changes based on the unique closed chain in the cloud, and synchronously updates the vehicle release status and available quantity for outbound shipment.

[0053] The technical effects and advantages provided by the present invention in the above technical solution are as follows:

[0054] This invention differs from existing methods that rely solely on license plate numbers, order numbers, or cloud-received order for inventory verification. Instead, it caches the initial weighing stability segment and vehicle axle sequence segment at the weighbridge edge terminal in a locally incremental order, and establishes a shadow inventory unit in the cloud, isolated from available inventory, for the inventory entry token awaiting closure. Therefore, even in cases of weak network retransmission, duplicate uploads, or cross-weighing by multiple vehicles, unverified weight information will not prematurely enter the official inventory ledger, directly preventing inflated inventory, premature release of outbound quotas, and accidental vehicle release.

[0055] This invention further utilizes material change segments collected by the edge terminal at the unloading point to form non-reusable unloading witness segments, and sends candidate closed chains to the edge terminal to replay locally cached data. Cross-elimination is performed based on the direction of net weighing difference, the direction of material change, and the correspondence between the first and last vehicle axle sequences. This process can identify pseudo-closed chains across vehicles that appear complete but are formed by incorrect splicing of records from different vehicles. Therefore, only the shadow inventory unit corresponding to the unique closed chain is converted into actual inventory change, and the vehicle release status and available outbound quantity are updated simultaneously. Attached Figure Description

[0056] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0057] Figure 1 This is a flowchart of a method for collaborative management of bulk material entry and exit and inventory based on end-to-end cloud collaboration according to the present invention.

[0058] Figure 2 This is a flowchart of a module of a bulk material entry and exit and inventory collaborative management system based on end-to-end cloud collaboration according to the present invention.

[0059] Figure 3 This is a flowchart of the unique closed-chain acquisition method of the present invention. Detailed Implementation

[0060] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0061] Example 1, please refer to Figure 1 As shown in this embodiment, a method for collaborative management of bulk material entry and exit and inventory based on edge-cloud collaboration includes:

[0062] Obtain the first weighing stability segment and vehicle axle sequence segment cached by the weighbridge edge terminal in local incremental order, and generate a token to be closed for storage.

[0063] In one implementation, the weighbridge edge terminal is communicatively connected to the weighbridge instrument, the gate controller, and the wheel axle detector. After a vehicle enters the weighbridge, the weighbridge edge terminal continuously acquires weight values ​​according to a preset sampling period. When the weight fluctuation does not exceed a preset range within multiple consecutive sampling periods, the corresponding time period is defined as the first stable weighing segment. The first stable weighing segment includes at least a local record number, weighing start time, weighing end time, stable weight value, and weighbridge equipment identifier.

[0064] During the process of a vehicle entering and leaving the weighbridge, the wheel axle detectors sequentially acquire the trigger signals of each axle as it passes the detection position. The weighbridge edge terminal generates a vehicle axle sequence segment according to the trigger order. The vehicle axle sequence segment includes at least the number of axles, the trigger order of each axle, the trigger time interval between adjacent axles, and the local record number corresponding to the first weighing stability segment.

[0065] The weighbridge edge terminal sets up an incrementing counter in its local storage area. Each time a first stable weighing segment is formed, a unique, incrementing local sequence number is assigned to that segment, and vehicle axle sequence segments with the same weighing process are stored associated with this incrementing local sequence number. During network interruption, the first stable weighing segment and vehicle axle sequence segments are still written to the buffer sequentially according to their incrementing local sequence numbers; after network recovery, they are uploaded sequentially according to their incrementing local sequence numbers.

[0066] The weighbridge edge terminal generates a pending closure entry token based on the initial weighing stability segment and the vehicle axle sequence segment. The pending closure entry token includes at least a locally incrementing sequence number, weighbridge equipment identifier, stable weight value, number of axles, vehicle axle sequence summary, the time of the initial weighing, and a token status field. Initially, the token status field is set to a pending closure state, awaiting the association of subsequent unloading witness segments and tare weighing stability segments.

[0067] For example, after a vehicle enters the weighbridge, the weighbridge edge terminal continuously collects a stable weight of 48.6 tons. The wheel axle detectors sequentially identify the six axles, and the trigger sequence of each axle is complete. The weighbridge edge terminal assigns a local incrementing sequence number 108 to this weighing and generates a pending entry token containing the stable weight, a summary of the six axle sequence, and the local incrementing sequence number 108.

[0068] The pending closure entry token is uploaded to the cloud, and a shadow inventory unit isolated from the available inventory is established for the pending closure entry token.

[0069] In this embodiment, a token receiving service, a shadow inventory management service, and a formal inventory ledger service are set up in the cloud. The token receiving service is used to receive the pending inbound tokens uploaded by the weighbridge edge terminal; the shadow inventory management service is used to establish, retrieve, and maintain shadow inventory units; and the formal inventory ledger service is used to calculate available inventory, verify outbound quantities, and generate outbound release results. The shadow inventory management service and the formal inventory ledger service each correspond to independent data tables and access interfaces, ensuring that weight information in the pending state does not directly participate in the calculation of available inventory.

[0070] After generating a pending entry token, the weighbridge edge terminal uploads the tokens to the cloud sequentially according to their local incrementing sequence numbers. Each pending entry token includes at least a token identifier, a local incrementing sequence number, an edge terminal number, a weighbridge device identifier, the time of the first weighing, the first stable weighing segment, the stable weight value, a vehicle axle sequence summary, and a token status field. The vehicle axle sequence summary includes at least the number of axles, the vehicle's direction of entry, and the trigger time interval between adjacent axles arranged in the order of axle triggering. The token status field is set to a pending closure state during upload.

[0071] After receiving the token to be closed in the cloud, the edge terminal number and the local incrementing sequence number are used as joint verification conditions to query whether the corresponding temporary storage entry already exists in the temporary storage area.

[0072] If the entry already exists, this upload will be identified as a duplicate upload. Only the most recent reception time and reception count of the temporary storage entry will be updated, and the shadow inventory unit will not be created again.

[0073] If it does not exist, a new temporary storage entry is created based on the pending closure token.

[0074] The temporary storage entry includes at least a token identifier, edge terminal number, local incrementing sequence number, first weighing stable segment, stable weight value, vehicle axle sequence summary, first reception time, most recent reception time, reception batch identifier, and temporary storage status.

[0075] The receiving batch identifier is used to mark the tokens to be closed and uploaded in a centralized manner after the network is restored. For multiple tokens to be closed and uploaded by the same edge terminal in the same receiving batch, the cloud restores the order of the first weighing according to the local incremental sequence number, rather than arranging them according to the order in which the tokens to be closed and uploaded to the cloud.

[0076] The shadow inventory management service establishes a unique shadow inventory unit for each newly created temporary storage entry. The shadow inventory unit includes at least a shadow inventory unit identifier, a token identifier, an edge terminal number, a locally incrementing sequence number, a weight snapshot, a vehicle axle sequence summary, an isolation status marker, a field for unloading witness to be bound, a field for returning to tare weighing to be bound, and a conversion record field.

[0077] Among them, the weight snapshot is generated based on the stable weight value; the isolation status flag is set to the isolation status when the shadow inventory unit is created; the unbound unloading witness field and the unbound return weighing field are empty in the initial state; the conversion record field is used to record whether the shadow inventory unit has been converted into a formal inbound record.

[0078] To isolate shadow inventory units from the formal inventory ledger, the shadow inventory management service writes shadow inventory units to an isolated storage area. When calculating real-time available inventory, determining whether the outbound quantity exceeds available inventory, and generating vehicle departure release results, the formal inventory ledger service only reads confirmed inventory records from the formal inventory ledger and does not read weight snapshots from the isolated storage area. Therefore, although weight information can be retrieved and traced in the cloud after the closing entry token is uploaded, the available outbound quantity will not increase until subsequent closing verification is completed.

[0079] To facilitate subsequent identification of cross-vehicle pseudo-closures, a hierarchical cross-index is established in the cloud for shadow inventory units. Specifically, firstly, an index layer is established based on the edge terminal number, classifying shadow inventory units originating from the same weighbridge edge terminal into the same terminal index partition; secondly, a second index layer is established based on the local incremental sequence number, arranging shadow inventory units within the same terminal index partition into a sequential chain according to the order of their first weighing; thirdly, a third index layer is established based on the vehicle axle sequence summary, enabling the same shadow inventory unit to be simultaneously written into the corresponding axle sequence candidate retrieval group.

[0080] The axle sequence candidate retrieval group is established as follows: The cloud platform arranges the number of axles and the preset intervals containing the trigger time intervals of adjacent axles sequentially according to the vehicle's direction of entry, generating axle sequence retrieval tags. Adjacent axle trigger time intervals can be categorized according to multiple preset continuous time intervals. For axle sequence retrieval tags with the same number of axles, the same vehicle direction of entry, and most adjacent axle trigger time intervals falling within the same preset interval, the cloud platform writes the corresponding shadow inventory unit into the same axle sequence candidate retrieval group. Through this method, the same shadow inventory unit is both retained in the sequential chain arranged according to its local ascending sequence number and can be cross-retrieved based on vehicle axle sequence characteristics.

[0081] When the cloud subsequently receives unloading witness segments or tare weighing stabilization segments, the shadow inventory management service first reads the isolation status flag of the corresponding shadow inventory unit, and then filters the shadow inventory units to be matched from the axis sequence candidate search group based on the edge terminal number, occurrence time, and axis sequence search tag. If only one shadow inventory unit is filtered out, the unloading witness segment or tare weighing stabilization segment is written into the corresponding pending binding field. If two or more shadow inventory units are filtered out, or if the same unloading witness segment has already requested binding to other shadow inventory units, formal binding is not performed; the relevant segments are only temporarily stored in the corresponding pending binding field, and the isolation status flag remains unchanged. Shadow inventory units in an isolated state must not be written into the formal inventory ledger.

[0082] After the unique closed-chain verification passes, the cloud reads the weight snapshot of the corresponding shadow inventory unit and combines it with the initial weighing stability segment and the tare weighing stability segment to generate a formal inventory entry record. Within the same inventory update transaction, the cloud writes the formal inventory entry record to the formal inventory ledger, associates the formal inventory entry record with the shadow inventory unit identifier, and changes the isolation status marker to the converted state. Shadow inventory units in the converted state no longer participate in axis sequence candidate retrieval and cannot generate formal inventory entry records again.

[0083] For example, during periods of weak network connectivity, the weighbridge edge terminal sequentially caches three pending inventory closure tokens with local incrementing sequence numbers 201, 202, and 203. After network recovery, due to retransmission retry, the cloud receives these pending inventory closure tokens in the order of 202, 201, and 203. The cloud restores the original weighing order based on the edge terminal number and the local incrementing sequence number, and creates three shadow inventory units for each. Although each of the three shadow inventory units stores a corresponding weight snapshot, they remain isolated. Before the unique closed-chain verification is successful, the weight snapshots do not participate in the real-time available inventory calculation. This reduces the risk of premature inventory release due to out-of-order offline retransmissions, duplicate uploads, and incorrect cross-vehicle splicing.

[0084] Obtain material change segments formed at the edge terminal of the unloading point during vehicle dwell time, and attach the material change segments to the closure token to form an unloading witness segment.

[0085] In this embodiment, the edge terminal of the unloading point is communicatively connected to a geomagnetic detector, an axle detector, a bearing surface detector, a vibration detector, and an obstruction detector. The geomagnetic detector is located near the unloading position entrance and exit to identify changes in the geomagnetic occupancy edge generated when vehicles enter and leave the unloading position. The axle detector is located at the unloading position entrance to generate an axle trigger sequence according to the axle passage order. The bearing surface detector is located in the bearing area of ​​the unloading position or on the bearing structure connected to the unloading hopper to acquire changes in the force on the bearing surface during unloading. The vibration detector is located on the unloading hopper, unloading grid, or adjacent bearing structure to acquire the vibration generated when material falls into the unloading hopper. The obstruction detector is located above the unloading hopper or to the side of the unloading position to acquire changes in the light path obstruction caused by dust during material unloading.

[0086] After receiving the entrance geomagnetic detector trigger signal when the vehicle enters the unloading position, the edge terminal of the unloading point records the time when the change of geomagnetic occupancy edge occurs and starts the temporary data acquisition task corresponding to the current vehicle parking process.

[0087] When the inlet geomagnetic detector changes from an unoccupied state to an occupied state, an entry edge is formed; when the outlet geomagnetic detector changes from an occupied state to an unoccupied state, a departure edge is formed. The unloading point edge terminal records the number of axles entering the unloading position, the triggering order of each axle, and the triggering time interval between adjacent axles, according to the triggering sequence of the axle detectors.

[0088] The cloud-based or unloading point edge terminal extracts the vehicle axle sequence summary from the pending closure entry token generated before the vehicle enters the unloading position. The unloading point edge terminal compares the axle triggering sequence generated at the entrance with the vehicle axle sequence summary. When the number of axles is consistent and the triggering time interval of most adjacent axles falls within the corresponding preset range, a temporary association is established between the current vehicle dwell process and the pending closure entry token. Thus, even if the vehicle license plate is obscured by dust, the unloading process can still be associated with the corresponding pending closure entry token based on the axle triggering sequence.

[0089] The unloading point edge terminal determines the vehicle dwell window based on the entry edge formed by the inlet geomagnetic detector, the completion time of all axles entering the unloading position, and the departure edge formed by the outlet geomagnetic detector. Specifically, the start time of the vehicle fully occupying the unloading position is the moment after the last axle entering the unloading position completes its triggering, and the end time of the vehicle dwell window is the moment when the outlet geomagnetic detector detects the vehicle starting to leave or the inlet geomagnetic detector returns to an unoccupied state. The time interval between the start and end times is defined as the vehicle dwell window. Within the vehicle dwell window, the unloading point edge terminal continuously acquires data on changes in the unloading position bearing surface, the duration of hopper vibration, and the duration of dust obstruction according to a preset acquisition cycle.

[0090] The bearing surface detector continuously collects the stress state of the bearing structure. The edge terminal at the unloading point compares the stress state at adjacent collection times, and when the stress state changes in the same direction over multiple consecutive collection cycles, the corresponding record is marked as a bearing surface change record. For vehicle unloading scenarios, after the material is transferred from the vehicle to the unloading hopper or unloading grid, the stress state of the bearing structure typically changes in an increasing direction; for vehicle loading scenarios, the direction of change of the stress state can be set accordingly to a decreasing direction. The bearing surface change record includes at least the start time of the change, the end time of the change, and the direction of change.

[0091] The vibration detector continuously collects the vibration status of the unloading hopper or unloading grid. When the vibration status continuously exceeds the preset reference state and the duration reaches the preset duration condition, the edge terminal of the unloading point determines the corresponding time segment as the hopper vibration duration segment. For short-term vibrations caused by vehicle engines, personnel movement, or the operation of nearby equipment, if their duration does not reach the preset duration condition, they will not be included in the hopper vibration duration segment.

[0092] The dust obstruction detector acquires changes in dust obstruction by monitoring the optical path between the transmitter and receiver. When the optical path signal received by the receiver is continuously weaker than a preset reference state, and the obstruction state persists for a preset duration, the edge terminal at the unloading point defines the corresponding time segment as the dust obstruction duration. At least two sets of obstruction detectors can be configured, located above the unloading hopper and to the side of the unloading position, respectively. The corresponding dust obstruction duration is recorded only when at least one set of obstruction detectors establishes a continuous obstruction state.

[0093] The edge terminal at the unloading point performs boundary filtering on the collected records. Records of changes in the bearing surface, continuous periods of hopper vibration, and continuous periods of dust obstruction occurring before the vehicle fully occupies the unloading position are not included in the current vehicle dwell process; corresponding records occurring after the vehicle leaves the unloading position are also deleted. Through the above processing, the interference caused by residual unloading from the previous vehicle, the early entry of the next vehicle, or operations at a nearby unloading position can be reduced to the current unloading process.

[0094] After boundary screening, the unloading point edge terminal combines the bearing surface change records, hopper vibration duration segments, and dust suppression duration segments according to the occurrence time. For multiple bearing surface change records with the same direction of change, if their time interval does not exceed a preset interval condition, they are merged into the same candidate change segment. For each candidate change segment, the unloading point edge terminal determines whether its occurrence time overlaps with at least one hopper vibration duration segment, whether the candidate change segment is within the coverage area of ​​the dust suppression duration segment, or whether the start and end times of the candidate change segment fall near the front and rear boundaries of the dust suppression duration segment, respectively.

[0095] When the direction of change of the bearing surface is consistent, the candidate change segment overlaps with the continuous vibration segment of the hopper, and the candidate change segment is surrounded by the continuous dust blocking segment, the corresponding records of the unloading point edge terminal will be merged into a material change segment. The material change segment includes at least the change start time, change end time, change direction, associated hopper vibration continuous segment, associated dust blocking continuous segment, vehicle parking window identifier, and unloading point edge terminal number.

[0096] The unloading point edge terminal generates an unloading witness segment based on the material change segment. The unloading witness segment includes at least an unloading witness segment identifier, a vehicle parking window identifier, a corresponding pending closure entry token identifier, a change start time, a change end time, a change direction, an unloading point edge terminal number, a segment occupancy marker, and additional status fields.

[0097] The segment occupancy marker is set to an occupancy state when the unloading witness segment is formed, and the additional status field is set to an additional status state.

[0098] The edge terminal at the unloading point uploads the unloading witness segment to the cloud. After receiving the unloading witness segment, the cloud queries the corresponding shadow inventory unit based on the pending closure entry token identifier and reads the pending unloading witness field from that shadow inventory unit.

[0099] If the unloading witness field to be bound is empty, and the unloading witness fragment has not been occupied by other pending closure entry tokens, then the unloading witness fragment is written into the unloading witness field to be bound, and the fragment occupation mark is changed to non-reusable state, and the attachment state field is changed to attached state.

[0100] If the same unloading witness fragment has already been attached to other pending inbound tokens, or if multiple pending inbound tokens simultaneously request the attachment of the same unloading witness fragment, the cloud will not perform duplicate attachment. Instead, it will keep the relevant pending inbound tokens in an isolated state and generate a pending review flag. Unloading witness fragments in a non-reusable state must not be written again into the pending unloading witness field of other shadow inventory units.

[0101] For example, after a vehicle completes its initial weighing, it enters the unloading position. An entrance geomagnetic detector forms an entry edge, and an axle detector sequentially detects six axles, with the axle triggering order matching the vehicle axle sequence summary in the corresponding pending closure token. After the vehicle fully enters the unloading position, a bearing surface detector continuously acquires bearing surface change records along the increasing direction, a vibration detector acquires the hopper vibration duration overlapping with the bearing surface change records, and an obstruction detector acquires the dust obstruction duration covering the bearing surface change records. The unloading point edge terminal merges these records into a material change segment and generates an unloading witness segment. The cloud attaches the unloading witness segment to the corresponding pending closure token and modifies the segment occupancy mark to a non-reusable state.

[0102] It should be noted that the unloading witness segment is not generated solely based on license plates, reservation orders, or single sensor signals. Instead, it is generated based on the temporal relationship between the vehicle parking window, changes in the bearing surface, hopper vibration, and dust obstruction. This ensures that the same unloading process can only be occupied by one pending warehousing token, thereby reducing the possibility of incorrect inventory conversion caused by false closures across vehicles.

[0103] Obtain stable back-weighing segments cached at the weighbridge edge terminal; when a stable back-weighing segment can be associated with multiple pending closure entry tokens, or when the upload order is inconsistent with the local incremental order, the cloud generates multiple candidate closure chains based on the spatial sequence of access control, first weighing, unloading, and back-weighing, and according to the rule that a unloading witness segment can only be occupied by one pending closure entry token.

[0104] In this embodiment, the cloud system includes a tare weighing receiving module, a closed-loop generation module, a conflict analysis module, a candidate closed-loop sorting module, and a shadow inventory unit association module. The tare weighing receiving module is responsible for receiving stable tare weighing segments cached by the weighbridge edge terminal and establishing a tare weighing record pool. The closed-loop generation module generates a preliminary closed loop based on the initial weighing, unloading witness segments, and tare weighing segments. The conflict analysis module identifies abnormal upload order or cross-vehicle pseudo-closed loops. The candidate closed-loop sorting module prioritizes the preliminary closed loops, eliminates conflicts, and associates them with shadow inventory units.

[0105] After the local weighbridge edge terminal completes the vehicle's return weighing, it uploads the stable weighing segment to the cloud according to the local incremental sequence number. Upon receiving the data, the cloud establishes a return weighing record pool. Each record includes the return weighing time, stable weight value, vehicle axle sequence summary, edge terminal number, upload batch, and local incremental sequence number. The return weighing record pool not only records weight information but also the network status and number of retransmissions during upload, enabling data traceability in weak or offline network conditions.

[0106] The cloud platform performs multi-dimensional correlation analysis on each segment in the tare weighing record pool. Specifically, this includes:

[0107] Time window matching calculates the degree of overlap of segments based on vehicle access time, first weighing time, unloading witness segment time, and return weighing time, to determine whether they may correspond to the same vehicle.

[0108] Weight continuity analysis compares the difference between the tare weighing segment and the first weighing to see if it is within the allowable range. If it exceeds the threshold, it is marked as a potential abnormal closed chain.

[0109] The axle sequence alignment compares the axle sequence summary of the weighing return segment with the axle sequence summary in the closure token to ensure sequence matching; otherwise, the priority of the candidate closure chain is reduced.

[0110] Conflict flags are used to mark a single tare weighing fragment as a conflict fragment if it matches multiple pending closure tokens, or if the upload order is inconsistent with the local incrementing order.

[0111] After identifying conflicting segments, the cloud system constructs an initial closed chain based on the spatial-temporal sequence of access control records, the initial stable weighing segment, the attached unloading witness segment, and the conflicting return weighing segment. Each closed chain consists of the initial weighing segment, the corresponding unloading witness segment, and the return weighing segment in sequence, while adhering to the unique occupancy rule for unloading witness segments: the same unloading witness segment can only be used by one closed chain and cannot be repeatedly bound to other closed chains. During the closed chain generation process, the system combines the vehicle parking window and axle trigger sequence to determine physical feasibility, ensuring that the closed chain sequence is consistent with the actual material flow.

[0112] After the initial closed chain is generated, the candidate closed chain sorting module performs priority calculation and conflict elimination. The sorting criteria include the order of the first weighing time, the order of the unloading witness segments, the continuity of the weight between the return weighing and the first weighing, and the axle sequence matching degree.

[0113] The conflict elimination rule is as follows: when multiple closed-loop candidates correspond to the same weighing segment, only the one that best matches the time window of the initial weighing and unloading witness segments is retained; other candidate closed loops are temporarily stored in an isolation queue. Closed loops in the isolation queue can be processed after verification at the edge terminal or after subsequent closure verification. This method effectively eliminates conflicting closed loops caused by cross-vehicle pseudo-closed loops, disordered upload order, and multiple vehicles sharing the weighbridge, ensuring that the final closed loop is highly consistent with the actual material flow path.

[0114] To enhance the reliability of generated closed chains, triple verification is performed on each candidate closed chain:

[0115] Time continuity verification: The initial weighing, unloading witness segment, and return weighing segment should be completed within a reasonable time interval; if the threshold is exceeded, an anomaly will be marked.

[0116] Weight continuity verification: The weight of the return weighing should match the sum of the initial weighing and unloading changes; otherwise, the priority should be reduced or the material should be moved to an isolation queue.

[0117] Axle sequence verification: The axle sequence for return weighing must be consistent with the axle sequence for initial weighing and the axle sequence logic during unloading; otherwise, the priority of the candidate closed chain will be reduced.

[0118] Once the candidate closed chain is generated, its identifier is associated with the shadow inventory unit, providing a unique basis for subsequent edge terminal verification and inventory closure conversion. For example, if a vehicle completes its first weighing and unloading under weak network conditions, and the order of the tare weighing uploads is out of sync with the local incrementing order, the cloud identifies the conflicting tare weighing segments, constructs two preliminary closed chains, and verifies them for time, weight, and axle order. Finally, a unique closed chain is determined for the shadow inventory unit closure conversion.

[0119] Please see Figure 3 As shown, the candidate closed chain is sent to the corresponding edge terminal, which then replays the locally cached weighing stability segment, vehicle axle sequence segment, and material change segment. Candidate closed chains that are inconsistent with the direction of material change or whose vehicle axle sequence cannot be matched end to end are eliminated, resulting in a unique closed chain.

[0120] In this embodiment, after generating a set of candidate closed chains, the cloud distributes each candidate closed chain to the corresponding weighbridge edge terminal and unloading point edge terminal. Each candidate closed chain includes at least a candidate closed chain identifier, a token identifier for closure into the warehouse, an identifier for the first stable weighing segment, an identifier for the tare weighing stable segment, an identifier for the unloading witness segment, a local incrementing sequence number, a weighbridge edge terminal number, and an unloading point edge terminal number. The weighbridge edge terminal stores the first stable weighing segment, the tare weighing stable segment, and the vehicle axle sequence segments corresponding to the first stable weighing segment and the tare weighing stable segment in its local buffer. The unloading point edge terminal stores material change segments, vehicle dwell windows, and conveyor operation markers in its local buffer. When writing the above segments, each edge terminal simultaneously records the local occurrence time and the most recent clock calibration offset.

[0121] When candidate closed chains are distributed from the cloud, the local occurrence times recorded by the weighbridge edge terminal and the unloading point edge terminal are converted to a unified playback time sequence based on the clock calibration offset corresponding to each segment. If no new clock calibration offset is obtained during a network interruption, the most recent valid clock calibration offset before the network interruption is used, and a "pending secondary verification" mark is added to the corresponding candidate closed chain. The "pending secondary verification" mark does not prevent the edge terminal from performing playback, but it does not directly trigger the shadow inventory unit conversion when there are multiple candidate closed chains that have not been eliminated.

[0122] For each candidate closed chain, the weighbridge edge terminal reads the corresponding initial weighing stabilization segment and the return weighing stabilization segment. Both the initial weighing stabilization segment and the return weighing stabilization segment include at least a stabilization start edge, a stabilization end edge, a stabilization weight value, an entry direction marker, and a corresponding locally incrementing sequence number. The weighbridge edge terminal uses the stabilization start edge of the initial weighing stabilization segment as the front playback anchor point and the stabilization end edge of the return weighing stabilization segment as the rear playback anchor point, and reads the weighing stabilization segment and vehicle axle sequence segment located between the two playback anchor points according to the local writing order.

[0123] The unloading point edge terminal reads the corresponding material change segment and vehicle dwell window based on the unloading witness segment identifier in the candidate closed chain. If both the start and end times of the material change segment are within the vehicle dwell window, the material change segment is marked as a valid playback segment. If either the start or end time is outside the vehicle dwell window, the material change segment is marked as a boundary abnormal segment.

[0124] The weighbridge edge terminal sequentially assigns the trigger time intervals of each adjacent axle during the initial weighing phase to a pre-set continuous time interval according to the order in which the axles pass the detection position, and arranges them in the trigger order as the first axle sequence fingerprint; the trigger time intervals of each adjacent axle during the return weighing phase are arranged in the same way as the tail axle sequence fingerprint.

[0125] The weighbridge edge terminal determines the axle sequence fingerprint comparison method based on the entry direction markings in the initial weighing stabilization segment and the return weighing stabilization segment. When the entry direction is opposite to that of the initial weighing stage and the return weighing stage, the tail axle sequence fingerprint is reversed and compared item by item with the head axle sequence fingerprint; when the two entry directions are the same, they are directly compared item by item in the original order. If the number of axles is different, or the trigger time interval between inconsistent adjacent axles reaches a preset number, it is determined that the head and tail axle sequence fingerprints cannot match; otherwise, it is determined that the head and tail axle sequence fingerprints can match.

[0126] For example, the initial axle sequence fingerprint formed during the first weighing stage includes intervals 1, 2, 3, and 1 in sequence. When the direction of entry during the return weighing stage is opposite to that of the first weighing stage, the tail axle sequence fingerprint is reversed. If the reversed tail axle sequence fingerprint differs from the aforementioned initial axle sequence fingerprint by only one interval, and the difference does not reach a preset number, then the initial and tail axle sequence fingerprints are considered to correspond.

[0127] The weighbridge edge terminal also determines the net weighing difference direction based on the initial stable weighing segment and the return stable weighing segment. For inbound unloading operations, when the stable weight value in the initial stable weighing segment is greater than the stable weight value in the return stable weighing segment, the net weighing difference direction is determined as the unloading direction; when the stable weight value in the initial stable weighing segment is not greater than the stable weight value in the return stable weighing segment, the corresponding candidate closed chain is marked as an abnormal weight direction. For outbound loading operations, the determination direction is reversed.

[0128] The edge terminal at the unloading point verifies the consistency between the direction of material change in the material change segment and the direction of the net weighing difference. For inbound unloading operations, when the net weighing difference direction is the unloading direction and the direction of change of the bearing surface at the unloading position is the increasing direction, the two are confirmed to be consistent. If the conveyor connected to the unloading position is in operation within the vehicle parking window, the edge terminal at the unloading point reads the conveyor operation mark and determines whether there is a record of a continuous local increase in the bearing surface overlapping with the continuous vibration segment of the hopper. In this case, a continuous increase in the bearing surface as a whole within the vehicle parking window is not required.

[0129] After completing the local playback, the weighbridge edge terminal and the unloading point edge terminal respectively return the playback results to the cloud. The playback results include at least the candidate closed chain identifier, edge terminal number, unified playback sequence, weighing net difference direction, material change direction, first and last axle sequence fingerprint verification results, vehicle parking window boundary verification results, conveying device operation mark, anomaly type, and playback completion mark.

[0130] The cloud performs cross-elimination on each candidate closed chain based on the replay results.

[0131] If a candidate closed chain corresponds to a boundary anomalous segment, then the candidate closed chain is removed.

[0132] If the direction of the net weighing difference is inconsistent with the direction of material change, the candidate closed chain is eliminated.

[0133] If the first and last axis fingerprints do not match, the candidate closed chain is removed. The removed candidate closed chain is written into the abnormal closed chain record table, and the corresponding abnormal type is recorded.

[0134] For candidate closed chains that are not eliminated, the cloud further verifies them according to the unique occupancy rule and the sequential continuity rule of the unloading witness segment. The sequential continuity means that in the local write records of the same weighbridge edge terminal, the first weighing stable segment is formed before the return weighing stable segment; the vehicle dwell window corresponding to the unloading witness segment is located between the two; and there are no other pending closure entry tokens that have occupied the same unloading witness segment. The local incremental sequence number of different edge terminals is only used within the scope of their respective terminals and no cross-terminal numerical comparison is performed.

[0135] When only one candidate closed chain is retained, the cloud identifies that candidate closed chain as the unique closed chain and writes the unique closed chain identifier into the conversion record field of the corresponding shadow inventory unit. When more than two candidate closed chains are retained, the cloud maintains the isolation state of the corresponding shadow inventory unit and generates a mark pending manual review. When all candidate closed chains are removed, the cloud generates a closure failure mark and prohibits the corresponding shadow inventory unit from being converted into a formal inbound record.

[0136] It should be noted that the cloud utilizes locally cached data on edge terminals to recover the vehicle weighing, unloading, and tare processes, directly determining inventory changes without relying on the upload order. This reduces the possibility of incorrect conversion of shadow inventory units caused by weak network re-uploads, cross-weighing, and incorrect splicing across vehicles.

[0137] Based on the unique closed chain, the cloud platform converts the corresponding shadow inventory unit into actual inventory changes and synchronously updates the vehicle release status and available quantity for outbound shipment.

[0138] In this embodiment, after the unique closed chain is determined, the cloud reads the corresponding shadow inventory unit identifier, vehicle identifier, storage location identifier, initial weighing stability segment, and return weighing stability segment. The cloud first calculates the actual inbound weight. The actual inbound weight can be calculated according to the following expression: ;in, Indicates the actual weight received into the warehouse; This indicates the stable weight value in the initial weighing stabilization segment, which is the weight of the loaded vehicle obtained when it is fully loaded and enters the yard. This represents the stable weight value in the stable segment of the weighing return, which is the empty vehicle weight obtained when the vehicle finishes unloading and returns to the weighbridge.

[0139] For the warehousing and unloading operation, when When the value is greater than 0, the calculation result will be used as the actual inbound weight when the corresponding shadow inventory unit is converted into a formal inbound record;

[0140] when When the value is ≤0, the corresponding unique closed chain is marked as a weight abnormal chain, and the isolation state of the corresponding shadow inventory unit is maintained, and no inventory increase is performed.

[0141] For example, if the stable weight value in the initial weighing stabilization segment is 48.6 tons, and the stable weight value in the return weighing stabilization segment is 16.2 tons, then... =48.6-16.2=32.4; Therefore, the actual weight of the goods put into storage this time is 32.4 tons.

[0142] For inbound and unloading operations, if the calculation result is greater than 0, the inventory conversion will continue.

[0143] When the calculation result is less than or equal to 0, the unique closed chain is marked as a weight anomalous chain, and the isolation state of the corresponding shadow inventory unit is maintained.

[0144] The cloud generates inventory conversion tokens based on a unique closed-chain identifier. Each inventory conversion token includes a unique closed-chain identifier, a shadow inventory unit identifier, a vehicle identifier, a storage location identifier, a conversion version, and a token digest. The conversion version is the previously recorded conversion version for the corresponding shadow inventory unit plus one. The token digest is generated using hashing: the unique closed-chain identifier, shadow inventory unit identifier, vehicle identifier, storage location identifier, and conversion version are concatenated in a fixed order, then hashed to obtain a fixed-length digest. The cloud uses the token digest as a unique search key to check if the same search key already exists in the inventory conversion record table. If no matching search key exists, the inventory conversion token is recorded as unused; if a matching search key already exists, the inventory increase operation is not performed again, and the current request is written to the pending compensation queue.

[0145] When the inventory conversion token is in an unused state, the cloud creates a formal inbound record during the same write process and changes the inventory conversion token to a used state. The formal inbound record includes at least a unique closed-chain identifier, a shadow inventory unit identifier, a storage location identifier, the actual inbound weight, and a write sequence number. The actual storage location inventory is updated as follows: the updated actual storage location inventory equals the original actual storage location inventory plus the actual inbound weight. After the write is completed, the corresponding shadow inventory unit is changed from an isolated state to a converted state. A converted shadow inventory unit cannot generate a formal inbound record again.

[0146] The cloud establishes available credit segments within the managed credit pool, each corresponding one-to-one with a formal inbound record. Each available credit segment includes at least a credit segment identifier, a formal inbound record identifier, a storage location identifier, credit weight, remaining credit weight, occupancy status, and a conversion version. The credit weight equals the actual inbound weight; the initial remaining credit weight equals the credit weight; and the initial occupancy status is unoccupied. Outbound availability is updated as follows: the updated outbound availability equals the original outbound availability plus the credit weight of the newly established available credit segment.

[0147] To prevent the same available quota segment from being reused for multiple outbound transactions, the cloud first reads the occupancy status of the available quota segment when it receives an outbound occupancy request. If the outbound request weight equals the remaining quota weight of the available quota segment, the occupancy status is changed to occupied, and the remaining quota weight is set to 0. If the outbound request weight is less than the remaining quota weight, the original available quota segment is split into an occupied quota segment and a remaining quota segment. The quota weight of the occupied quota segment equals the outbound request weight, and the quota weight of the remaining quota segment equals the remaining quota weight before splitting minus the outbound request weight. Each split quota segment is only allowed to complete one occupancy write. Therefore, the available outbound quota is always equal to the sum of the remaining quota weights of all unoccupied quota segments in the managed quota pool.

[0148] The cloud-based system establishes causal barriers around events such as writing official database records, creating available quota fragments, and releasing vehicles. Each event records an event identifier, preceding event identifiers, completion status, and conversion version.

[0149] The completion status of the formal entry record writing event is recorded as the first status value, and the completion status of the available quota fragment creation event is recorded as the second status value. Both the first and second status values ​​are represented by 0 or 1, where 0 indicates incomplete and 1 indicates complete. Vehicle release conditions are determined as follows: the release judgment value equals the first status value multiplied by the second status value. When the release judgment value is 1, the vehicle release status is changed to "allowed to release"; when the release judgment value is 0, the vehicle release status remains "pending verification".

[0150] For example, a vehicle's initial stable weight in the weighing stability segment is 48.6 tons, and its stable weight in the return weighing stability segment is 16.2 tons, with an actual inbound weight of 32.4 tons. The corresponding actual inventory at the storage location before the update is 810.0 tons, and the available outbound quantity before the update is 742.0 tons. After the inventory conversion token arrives at the cloud for the first time, the actual inventory at the storage location is updated to 842.4 tons, and an available quota segment with a quota weight of 32.4 tons is created in the managed quota pool, updating the available outbound quantity to 774.4 tons. After both the formal inbound record writing event and the available quota segment creation event are completed, the release judgment value is 1, and the vehicle release status is changed from pending verification to allowed release.

[0151] If the same inventory conversion token arrives at the cloud again due to a weak network connection, and the cloud queries that the same token digest is already in a used state, it will not add the actual inventory of the storage location again, will not create a new available quota segment, and will not modify the vehicle release status again. The corresponding request will be written to the compensation queue. The compensation queue records at least the inventory conversion token digest, the exception type, the most recent arrival time, the number of retries, and the corresponding shadow inventory unit identifier. When the preceding event is missing, the cloud will query the corresponding write result again according to the preceding event identifier; if it finds that the missing event has been completed, it will recalculate the release judgment value; if it is still not completed, the vehicle release status will remain in the pending verification state.

[0152] It should be noted that the shadow inventory unit is converted into actual inventory changes only after the unique closed-loop verification is passed. The formal inbound records, outbound available quantity and vehicle release status are updated synchronously in a clear causal order, which can reduce the repeated increase of inventory, repeated release of quota and premature release of vehicles caused by duplicate transmission, out-of-order arrival and concurrent occupation.

[0153] Example 2, please refer to Figure 2 As shown in this embodiment, a bulk material inbound / outbound and inventory collaborative management system based on edge-cloud collaboration includes:

[0154] The warehouse entry token generation module obtains the first weighing stability segment and vehicle axle sequence segment cached by the weighbridge edge terminal in local incremental order, and generates a warehouse entry token to be closed.

[0155] The shadow inventory isolation establishment module uploads the pending closure entry token to the cloud and establishes a shadow inventory unit isolated from the available inventory for the pending closure entry token;

[0156] The unloading witness segment formation module acquires the material change segment formed at the edge terminal of the unloading point during the vehicle's stay, and attaches the material change segment to the warehousing token to be closed to form the unloading witness segment.

[0157] The candidate closed chain generation module obtains the stable back-taper weighing segments cached by the weighbridge edge terminal. When a stable back-taper weighing segment can be associated with multiple pending closure entry tokens, or when the upload order is inconsistent with the local incremental order, the cloud generates multiple candidate closed chains based on the spatial order of access control, first weighing, unloading, and back-taper weighing, and according to the rule that a unloading witness segment can only be occupied by one pending closure entry token.

[0158] The unique closed chain determination module sends the candidate closed chains to the corresponding edge terminals. The corresponding edge terminals replay the locally cached weighing stability segments, vehicle axle sequence segments, and material change segments, and eliminate candidate closed chains that are inconsistent with the direction of material change or whose vehicle axle sequence cannot be matched end to end, thus obtaining the unique closed chain.

[0159] The collaborative status update module converts the corresponding shadow inventory unit into actual inventory changes based on the unique closed chain in the cloud, and synchronously updates the vehicle release status and available quantity for outbound shipment.

[0160] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

Claims

1. A method for collaborative management of bulk material entry and exit and inventory based on edge-cloud collaboration, characterized in that, include: Obtain the first weighing stability segment and vehicle axle sequence segment cached by the weighbridge edge terminal in local incremental order, and generate a token to be closed for storage. Upload the pending closure entry token to the cloud, and establish a shadow inventory unit isolated from the available inventory for the pending closure entry token; Obtain material change segments formed at the edge terminal of the unloading point during the vehicle's stay, and attach the material change segments to the warehousing token to be closed to form an unloading witness segment. Obtain stable back-weighing segments cached at the edge terminal of the weighbridge; when a stable back-weighing segment can be associated with multiple pending closure entry tokens, or when the upload order is inconsistent with the local incremental order, the cloud generates multiple candidate closure chains based on the spatial order of access control, first weighing, unloading and back-weighing, and according to the rule that a unloading witness segment can only be occupied by one pending closure entry token. The candidate closed chain is sent to the corresponding edge terminal, and the corresponding edge terminal replays the locally cached weighing stability segment, vehicle axle sequence segment and material change segment. Candidate closed chains that are inconsistent with the direction of material change or whose vehicle axle sequence cannot be matched end to end are eliminated to obtain a unique closed chain. Based on the unique closed chain, the cloud platform converts the corresponding shadow inventory unit into actual inventory changes and synchronously updates the vehicle release status and available quantity for outbound shipment.

2. The method for collaborative management of bulk material entry and exit and inventory based on end-to-end cloud collaboration as described in claim 1, characterized in that, The token to be closed for entry includes: local incrementing sequence number, stable weight value, vehicle axle sequence summary, and the time of the first weighing.

3. The method for collaborative management of bulk material entry and exit and inventory based on end-to-end cloud collaboration as described in claim 1, characterized in that, Establish shadow inventory units that are isolated from available inventory, including: The cloud uses the edge terminal number and the local incrementing sequence number to form a joint verification condition to perform duplicate transmission identification on the received token to be closed in the database; when a corresponding temporary storage entry already exists, only the most recent reception time and reception count are updated. When there is no corresponding temporary storage entry, the cloud creates a unique shadow inventory unit and writes the shadow inventory unit into an isolated storage area that is independent of the formal inventory ledger, so that the weight snapshot in it does not participate in the real-time available inventory calculation. The cloud establishes the first index layer based on the edge terminal number, the second index layer based on the local incremental sequence number, and the third index layer based on the axle sequence retrieval tags formed by arranging the vehicle's direction of entry, the number of axles, and the trigger time interval between adjacent axles. The cloud enables the same shadow inventory unit to be simultaneously written into a sequential chain arranged according to the local incremental sequence number and the corresponding axis sequence candidate retrieval group, which is used to restore the initial weighing formation order when the offline retransmission order is disordered.

4. The method for collaborative management of bulk material entry and exit and inventory based on end-to-end cloud collaboration as described in claim 1, characterized in that, The unloading witness segment includes: Based on the change in geomagnetic occupancy edge and axle triggering sequence formed when the vehicle enters the unloading position, the edge terminal of the unloading point determines the start time when the vehicle fully occupies the unloading position and the end time when the vehicle leaves the unloading position, and the time interval between the two is determined as the vehicle dwell window. Within the vehicle parking window, the edge terminal of the unloading point continuously acquires the changes in the bearing surface of the unloading position, the continuous period of hopper vibration, and the continuous period of dust obstruction, and deletes the change records that occur before the vehicle fully occupies the unloading position or after the vehicle leaves the unloading position. At the edge of the unloading point, the records of changes in the bearing surface that are consistent in direction, overlap with the duration of the hopper vibration, and are covered by the duration of dust obstruction are merged into a material change segment.

5. The method for collaborative management of bulk material entry and exit and inventory based on end-to-end cloud collaboration according to claim 4, characterized in that, Generate multiple candidate closed chains, including: The cloud receives stable back-to-tape weighing segments cached by the edge terminal of the weighbridge and establishes a back-to-tape weighing record pool according to the edge terminal number, upload batch and local incremental sequence number; Each stable segment of weighing return is associated with multiple pending closure tokens in the axle sequence candidate retrieval group. When the same stable segment of weighing return can be associated with multiple pending closure tokens, or when its upload order is inconsistent with the local increment order of the corresponding weighbridge edge terminal, the stable segment of weighing return is marked as a conflict segment. Based on the spatial sequence of access control records, initial weighing stabilization segment, vehicle parking window, attached unloading witness segment, and conflict segment, a preliminary closed chain is established, consisting of the initial weighing stabilization segment, the unloading witness segment, and the return weighing stabilization segment in sequence. According to the rule that a single unloading witness segment can only be occupied by one preliminary closed chain, preliminary closed chains that satisfy spatial order but have different matching combinations are retained as multiple candidate closed chains.

6. The method for collaborative management of bulk material entry and exit and inventory based on edge-cloud collaboration according to claim 1, characterized in that, The candidate closed chain is sent to the corresponding edge terminal, including: The cloud reads the local occurrence time and the most recent clock calibration offset recorded by the edge terminal of the weighbridge and the edge terminal of the unloading point, and converts the local occurrence time of the corresponding segments of each candidate closed chain into a unified playback timing sequence. When no new clock calibration offset is obtained during a network outage, the cloud uses the most recent valid clock calibration offset before the network outage and adds a mark indicating that it needs to be re-verified to the corresponding candidate closed chain. The weighbridge edge terminal uses the stable start edge of the first weighing stable segment as the front playback anchor point and the stable end edge of the return weighing stable segment as the rear playback anchor point. It reads the weighing stable segment and vehicle axle sequence segment located between the two playback anchor points according to the local writing order. The edge terminal at the unloading point reads the corresponding material change segment and vehicle parking window based on the unloading witness segment identifier, and marks the material change segment whose start time or end time is outside the vehicle parking window as a boundary abnormal segment.

7. The method for collaborative management of bulk material entry and exit and inventory based on edge-cloud collaboration according to claim 1, characterized in that, Eliminate candidate closed chains, including: The weighbridge edge terminal assigns the trigger time intervals of adjacent axles in the first weighing phase to a pre-set continuous time interval according to the order in which the axles pass the detection position, and arranges them sequentially to form the first axle sequence fingerprint; the trigger time intervals of adjacent axles in the return weighing phase are formed into the tail axle sequence fingerprint in the same way. When the edge terminal of the weighbridge determines that the direction of travel for the first weighing stage is opposite to that for the return weighing stage based on the direction of travel marking, the rear axle sequence fingerprint is reversed and compared with the front axle sequence fingerprint item by item; when the directions of travel are the same, the fingerprints are compared item by item in the original order. For inbound unloading operations, when the stable weight value of the first stable weighing segment is greater than the stable weight value of the return weighing segment, the weighing net difference direction is determined as the unloading direction; when the bearing surface change direction is the increasing direction, the material change direction is determined as the unloading direction. Candidate closed chains are removed from the cloud if the direction of net weighing difference does not correspond to the direction of material change, the first and last axis sequence fingerprints cannot be matched, or the material change segment is marked as a boundary abnormal segment.

8. The method for collaborative management of bulk material entry and exit and inventory based on end-to-end cloud collaboration according to claim 1, characterized in that, A unique closed chain is obtained, including: The cloud performs sequential continuous verification on the candidate closed chains that have not been eliminated. The sequential continuous verification includes: in the local write records of the same weighbridge edge terminal, the first weighing stable segment is formed before the return weighing stable segment; the vehicle dwell window corresponding to the unloading witness segment is located between the first weighing stable segment and the return weighing stable segment; there are no other pending closure entry tokens that have occupied the same unloading witness segment. When only one candidate closed chain that satisfies the sequential continuous verification is retained, the cloud determines it as the unique closed chain and writes the unique closed chain identifier into the conversion record field of the corresponding shadow inventory unit; When more than two candidate closed chains are retained, the cloud maintains the isolation status of the corresponding shadow inventory unit and marks it as pending manual review; When all candidate closed chains are eliminated, a closure failure flag is generated in the cloud, and the corresponding shadow inventory unit is prohibited from being converted into a formal inbound record.

9. The method for collaborative management of bulk material entry and exit and inventory based on end-to-end cloud collaboration according to claim 1, characterized in that, Convert the corresponding shadow inventory units to actual inventory changes and synchronously update vehicle release status and available outbound quantity, including: The cloud generates an inventory conversion token that can only be used once based on a unique closed chain, and writes the unique closed chain identifier, shadow inventory unit identifier, vehicle identifier, storage location identifier, and conversion version into the inventory conversion token in a fixed order; After confirming that the inventory conversion token has not been used, the cloud will convert the corresponding shadow inventory unit into a formal inbound record and create an available quota fragment in the managed quota pool that corresponds one-to-one with the formal inbound record. The available quota fragment carries the quota weight, remaining quota weight, occupancy status and conversion version. The cloud establishes a causal barrier based on the sequential dependency of the formal entry record writing event, the available quota fragment creation event, and the vehicle release event. Only after both the formal entry record writing event and the available quota fragment creation event are completed will the corresponding vehicle release status be updated to allow release. When a duplicate inventory conversion token is detected, an available quota segment is already occupied, or any preceding event is not completed, the cloud will not repeatedly increase the actual inventory, will not repeatedly release the available outbound quantity, will keep the vehicle release status unchanged, and will write the corresponding update request to the compensation queue.

10. A bulk material inbound / outbound and inventory collaborative management system based on edge-cloud collaboration, used to implement the bulk material inbound / outbound and inventory collaborative management method based on edge-cloud collaboration as described in any one of claims 1-9, characterized in that, include: The warehouse entry token generation module obtains the first weighing stability segment and vehicle axle sequence segment cached by the weighbridge edge terminal in local incremental order, and generates a warehouse entry token to be closed. The shadow inventory isolation establishment module uploads the pending closure entry token to the cloud and establishes a shadow inventory unit isolated from the available inventory for the pending closure entry token; The unloading witness segment formation module acquires the material change segment formed at the edge terminal of the unloading point during the vehicle's stay, and attaches the material change segment to the warehousing token to be closed to form the unloading witness segment. The candidate closed chain generation module obtains the stable back-taper weighing segments cached by the weighbridge edge terminal. When a stable back-taper weighing segment can be associated with multiple pending closure entry tokens, or when the upload order is inconsistent with the local incremental order, the cloud generates multiple candidate closed chains based on the spatial order of access control, first weighing, unloading, and back-taper weighing, and according to the rule that a unloading witness segment can only be occupied by one pending closure entry token. The unique closed chain determination module sends the candidate closed chains to the corresponding edge terminals. The corresponding edge terminals replay the locally cached weighing stability segments, vehicle axle sequence segments, and material change segments, and eliminate candidate closed chains that are inconsistent with the direction of material change or whose vehicle axle sequence cannot be matched end to end, thus obtaining the unique closed chain. The collaborative status update module converts the corresponding shadow inventory unit into actual inventory changes based on the unique closed chain in the cloud, and synchronously updates the vehicle release status and available quantity for outbound shipment.

Images