A finished paper out-of-warehouse dynamic tolerance foolproof system and method applied to a weighbridge

Abstract

The application relates to a finished paper out-of-warehouse dynamic tolerance fool-proof system and method applied to a weighbridge, which comprises a SKU risk grading modeling module, a weight fingerprint library module, a dynamic tolerance calculation engine and a weighbridge checking and intercepting execution module. The SKU risk grading modeling module is used for reading SKU master data and multidimensional data, performing risk scoring on the SKU based on the multidimensional data to obtain a corresponding SKU risk grade, and determining a single-piece standard weight of the SKU based on a packaging process BOM table. The dynamic tolerance calculation engine is used for determining a tolerance coefficient according to the SKU risk grade, and calculating a dynamic tolerance limit value according to the tolerance coefficient. The weighbridge checking and intercepting execution module is used for comparing a weighing result of a current train with a calculated release area, generating corresponding early warning information and executing a corresponding release strategy according to a comparison result. Frequent false interception of low-risk SKUs is avoided, and abnormality of high-risk SKUs can be effectively detected.

Description

Technical Field

[0001] This application relates to the field of papermaking technology, specifically to a dynamic tolerance and error-proofing system and method for finished paper leaving the warehouse, applied to weighbridges. Background Technology

[0002] Currently, the finished product warehouse logistics and delivery management in the paper industry mainly relies on the following methods:

[0003] (1) Outbound verification based on fixed theoretical weight: The existing system uses the theoretical weight value of the order for outbound weight comparison, without considering the weight deviation differences between different SKUs (Stock Keeping Units) caused by factors such as packaging method, product specifications, and measurement method. When outbound, only "theoretical value of the order ± fixed percentage" is used as the release basis.

[0004] (2) Manual verification and document check: After loading is completed, the warehouse manager manually checks each item against the shipping list or weighs the items on a weighbridge and then manually compares the order weight. This method relies on the experience and sense of responsibility of the personnel, and is inefficient and prone to errors.

[0005] (3) Single threshold weighbridge error prevention: Some companies have set up simple upper and lower limit alarm functions for the weighbridge, but the threshold is set to a fixed value and dynamic factors such as the historical deviation pattern of different SKUs and the weight fluctuation of the packaging are not considered.

[0006] The existing finished product warehouse logistics and delivery management methods in the paper manufacturing industry mainly have the following problems:

[0007] Disadvantage 1: Lack of SKU-differentiated management, leading to a high false positive rate. The existing system uses a uniform fixed weight threshold for all products, failing to differentiate the weight characteristics of different SKUs. For example, the weight fluctuation rate of packaging materials for white cardboard sheets (theoretical weight measurement) differs significantly from that of pulp boards (actual weight measurement). Using the same threshold results in low-risk SKUs being frequently and falsely intercepted, while anomalies in high-risk SKUs cannot be effectively detected. Statistics show that customer complaints from paper manufacturing companies due to discrepancies in finished product logistics and delivery, as well as differences in finished product weight, account for approximately 15% to 30% of all customer complaints.

[0008] Disadvantage 2: Insufficient interception accuracy due to failure to consider packaging weight fluctuations. The existing system directly compares the "gross weight of the whole vehicle with the theoretical total weight of the order," without taking into account the weight fluctuations of packaging materials (pallets, stretch film, corner protectors, cardboard, etc.). Since the weight of packaging materials itself varies from batch to batch, direct comparison will misjudge packaging material fluctuations as abnormal product quantities, resulting in a high false alarm rate and poor system usability.

[0009] Disadvantage 3: Static tolerance settings prevent adaptive optimization. The existing system's weight threshold remains unchanged once set, lacking a closed-loop feedback mechanism based on historical deviation data, customer complaint records, and actual weighing data. The system cannot dynamically adjust the tolerance coefficient based on actual operating data, leading to a gradual decrease in interception accuracy over time.

[0010] Disadvantage 4: Lack of preventative mechanisms against human intervention. The system lacks effective access control and auditing mechanisms for operations such as "human-caused deletion of abnormal records." If anomalies such as over-installation, under-installation, or missed scanning occur, the system cannot trace them, leading to an inability to effectively manage operational losses and customer complaints. Summary of the Invention

[0011] In view of the above problems, this application provides a dynamic tolerance error prevention system and method for finished paper outbound shipment applied to weighbridges, which solves the problem of lack of SKU differentiation management and high misjudgment rate in the existing finished paper warehouse logistics and shipment management methods in the paper industry.

[0012] To achieve the above objectives, the inventors provide a dynamic tolerance and error-proofing system for finished paper outbound shipment applied to weighbridges, comprising:

[0013] The SKU risk grading modeling module is used to read SKU master data and multi-dimensional data, and to score the risk of SKU based on the multi-dimensional data to obtain the corresponding SKU risk level. The multi-dimensional data includes product specifications, packaging method data, historical weight deviation rate and customer complaint records.

[0014] The weight fingerprint database module is used to determine the standard weight of a single SKU based on the packaging process BOM, and to determine the type of the SKU according to the standard weight of a single SKU and the packaging tare weight waviness, and to establish the SKU weight fingerprint database.

[0015] The dynamic tolerance calculation engine is used to determine the tolerance coefficient based on the SKU risk level and calculate the dynamic tolerance limit based on the tolerance coefficient. When the trigger condition is met, the current tolerance coefficient is updated.

[0016] The weighbridge verification and interception execution module is used to obtain the standard weight of a single piece and the dynamic tolerance limit of the corresponding SKU of the current vehicle, calculate the release area, compare the weighing result of the current vehicle with the calculated release area, generate corresponding warning information and execute corresponding release strategies based on the comparison result.

[0017] In some embodiments, the SKU risk classification modeling module is specifically used to normalize the original scores corresponding to the current SKU dimension data to obtain a normalized score, and then perform weighted summation according to the corresponding weights to obtain a comprehensive score, and determine the corresponding SKU risk level based on the comprehensive score.

[0018] In some embodiments, the weight fingerprint library module is used to determine the packaging weight of the SKU based on the packaging process BOM, obtain the standard weight of the SKU per unit based on the paper system weight and the packaging, calculate the actual weight based on the effective sampling amount of the SKU within a preset time, calculate the volatility based on the actual weight of the SKU, the standard weight per unit, and the inverse calculation, and determine the type corresponding to the SKU based on the calculated volatility.

[0019] In some embodiments, the dynamic tolerance calculation engine is used to determine the corresponding base coefficient based on the risk level of the current SKU, determine the adjustment coefficient based on the historical average deviation rate, and determine the total tolerance coefficient based on the base coefficient and the adjustment coefficient.

[0020] In some embodiments, the dynamic tolerance calculation engine is further used to calculate a new tolerance coefficient based on the historical tolerance coefficient and the deviation coefficient of the current batch using a weighted average method.

[0021] In some embodiments, the dynamic tolerance calculation engine is also used to determine whether the difference between the new tolerance coefficient and the historical tolerance coefficient exceeds the adjustment step size. If it does, the historical tolerance coefficient is adjusted according to the adjustment step size to obtain the new tolerance coefficient.

[0022] In some embodiments, the dynamic tolerance calculation engine obtains the final tolerance coefficient by taking the square root of the sum of squares of the newly calculated tolerance coefficient and the weighing instrument error coefficient.

[0023] In some embodiments, the dynamic tolerance calculation engine is also used to determine whether the obtained final tolerance coefficient exceeds the limit corresponding to the current risk level. If it does, the limit value is used as the final tolerance coefficient.

[0024] In some embodiments, it also includes:

[0025] The audit and anti-human intervention module is used to control access to the clearing of abnormal records, record the operator, operation time, operation reason and trigger a secondary confirmation mechanism;

[0026] The closed-loop feedback optimization module is used to send the cause of deviation, processing result and corrected weight back to the model database after each anomaly handling is completed.

[0027] Another technical solution is also provided: a dynamic tolerance error-proofing method for finished paper outbound shipment applied to weighbridges. This method is applied to the aforementioned dynamic tolerance error-proofing system for finished paper outbound shipment applied to weighbridges, and includes the following steps:

[0028] When a vehicle enters the weighbridge, the SKU master data and multi-dimensional data of the current vehicle's SKU are obtained, and the risk level of the corresponding SKU is calculated based on the multi-dimensional data.

[0029] The standard weight of a single unit is determined based on the packaging process BOM based on the SKU, and the type of the corresponding SKU is determined based on the standard weight of the single unit and the tare weight waviness of the packaging.

[0030] Tolerance coefficients are determined based on SKU risk levels, and dynamic tolerance limits are calculated based on these tolerance coefficients.

[0031] Calculate the release area based on the standard weight of a single piece and the dynamic tolerance limit of the corresponding SKU for the current train service;

[0032] The weighing result of the current train is compared with the calculated release area, and corresponding warning information is generated and the corresponding release strategy is executed based on the comparison result.

[0033] Unlike existing technologies, the above technical solution reads SKU master data and multi-dimensional data to perform risk scoring on SKUs, obtaining corresponding SKU risk levels. It determines the standard weight of each SKU based on the packaging process BOM, then identifies the SKU type based on the standard weight and packaging tare weight waviness, establishing an SKU weight fingerprint database. Based on the SKU risk level, it determines the corresponding tolerance coefficient, calculates dynamic tolerance limits based on the tolerance coefficient, and updates the current tolerance coefficient when trigger conditions are met. It obtains the standard weight and dynamic tolerance limit of the SKU corresponding to the current train, calculates the release area, compares the current train's weighing result with the calculated release area, generates corresponding warning information, and executes corresponding release strategies based on the comparison results. Through SKU risk classification and differentiated tolerance settings, it avoids frequent false interceptions of low-risk SKUs by applying a "one-size-fits-all" threshold, while ensuring that anomalies in high-risk SKUs can be effectively detected. A differentiated dynamic tolerance coefficient based on SKU risk level was set, with a wider tolerance range for high-risk SKUs (providing reasonable room for error) and a narrower tolerance range for low-risk SKUs (achieving precise control).

[0034] The above description of the invention is merely an overview of the technical solution of this application. In order to enable those skilled in the art to better understand the technical solution of this application and to implement it based on the description and drawings, and to make the above-mentioned objectives and other objectives, features and advantages of this application easier to understand, the following description is provided in conjunction with the specific embodiments and drawings of this application. Attached Figure Description

[0035] The accompanying drawings are only used to illustrate the principles, implementation methods, applications, features, and effects of specific embodiments of this application and other related content, and should not be considered as limitations on this application.

[0036] In the accompanying drawings of the instruction manual:

[0037] Figure 1 This is a schematic diagram of a dynamic tolerance and error-proofing system for finished paper outbound from a weighbridge, as described in a specific implementation method.

[0038] Figure 2 This is a schematic diagram of the outlier removal rule described in a specific implementation method;

[0039] Figure 3 This is a schematic diagram of the data processing flow of the dynamic tolerance calculation engine described in a specific implementation method;

[0040] Figure 4 A schematic diagram of the system pop-up window content in Embodiment 1 described in the specific implementation method;

[0041] Figure 5 This is a schematic diagram of the system pop-up window content in Embodiment 2 described in the specific implementation method;

[0042] Figure 6 This is a flowchart illustrating a specific implementation of the dynamic tolerance and error-proofing method for finished paper leaving the warehouse, applied to weighbridges. Detailed Implementation

[0043] To illustrate the possible application scenarios, technical principles, implementable specific solutions, and achievable objectives and effects of this application in detail, the following description, in conjunction with the listed specific embodiments and accompanying drawings, provides a detailed explanation. The embodiments described herein are merely illustrative of the technical solutions of this application and are therefore intended to limit the scope of protection of this application.

[0044] In this document, the term "embodiment" means that a specific feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The term "embodiment" appearing in various places throughout the specification does not necessarily refer to the same embodiment, nor does it specifically limit its independence or connection with other embodiments. In principle, in this application, as long as there are no technical contradictions or conflicts, the technical features mentioned in each embodiment can be combined in any way to form corresponding implementable technical solutions.

[0045] Unless otherwise defined, the technical terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains; the use of related terms herein is merely for the purpose of describing particular embodiments and is not intended to limit this application.

[0046] In the description of this application, the term "and / or" is used to describe the logical relationship between objects, indicating that three relationships can exist. For example, A and / or B means: A exists, B exists, and A and B exist simultaneously. Additionally, the character " / " in this document generally indicates that the preceding and following objects have an "or" logical relationship.

[0047] In this application, terms such as “first” and “second” are used only to distinguish one entity or operation from another, and do not necessarily require or imply any actual quantity, hierarchy or order relationship between these entities or operations.

[0048] Unless otherwise specified, the use of terms such as “comprising,” “including,” “having,” or other similar expressions in this application is intended to cover non-exclusive inclusion, which does not exclude the presence of additional elements in a process, method, or product that includes the stated elements, such that a process, method, or product that includes a list of elements may include not only those defined elements but also other elements not expressly listed, or elements inherent to such a process, method, or product.

[0049] As understood in the Examination Guidelines, in this application, expressions such as "greater than," "less than," and "exceeding" are understood to exclude the stated number; expressions such as "above," "below," and "within" are understood to include the stated number. Furthermore, in the description of the embodiments in this application, "multiple" means two or more (including two), and similar expressions related to "multiple" are also understood in this way, such as "multiple groups" and "multiple times," unless otherwise explicitly specified.

[0050] In the description of the embodiments of this application, the space-related expressions used, such as "center," "longitudinal," "lateral," "length," "width," "thickness," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "vertical," "top," "bottom," "inner," "outer," "clockwise," "counterclockwise," "axial," "radial," and "circumferential," indicate the orientation or positional relationship based on the orientation or positional relationship shown in the specific embodiments or drawings. They are only for the purpose of describing the specific embodiments of this application or for the reader's understanding, and do not indicate or imply that the device or component referred to must have a specific position, a specific orientation, or be constructed or operated in a specific orientation. Therefore, they should not be construed as limitations on the embodiments of this application.

[0051] Unless otherwise expressly specified or limited, the terms "installation," "connection," "linking," "fixing," and "setting," as used in the description of the embodiments of this application, should be interpreted broadly. For example, "connection" can be a fixed connection, a detachable connection, or an integral setting; it can be a mechanical connection, an electrical connection, or a communication connection; it can be a direct connection or an indirect connection through an intermediate medium; it can be the internal connection of two components or the interaction between two components. For those skilled in the art to which this application pertains, the specific meaning of the above terms in the embodiments of this application can be understood according to the specific circumstances.

[0052] Please see Figure 1 This embodiment provides a dynamic tolerance and error-proofing system for finished paper outbound shipment applied to weighbridges, including:

[0053] The SKU risk grading modeling module 110 is used to read SKU master data and multi-dimensional data, and to perform risk scoring on SKUs based on the multi-dimensional data to obtain the corresponding SKU risk level. The multi-dimensional data includes product specifications, packaging method data, historical weight deviation rate, and customer complaint records. Specifically, it reads SKU master data from the ERP system, packaging method data from the packaging process BOM, weight deviation rate from historical shipment data of the TMS (Transportation Management System), and abnormal records from the customer complaint ledger. Based on the multi-dimensional data, each SKU is risk-scored and divided into three levels: A (high risk), B (medium risk), and C (low risk).

[0054] The weight fingerprint database module 120 is used to determine the standard weight of a single SKU based on the packaging process BOM, and to determine the type of the SKU according to the standard weight of the single SKU and the packaging tare weight waviness, and to establish an SKU weight fingerprint database; the standard weight of a single SKU (base weight) is determined based on the packaging process BOM and includes the standard weight of packaging materials such as cardboard, stretch film, corner protectors, and pallets.

[0055] The dynamic tolerance calculation engine 130 is used to determine the tolerance coefficient based on the SKU risk level and calculate the dynamic tolerance limit based on the tolerance coefficient. When the trigger condition is met, the current tolerance coefficient is updated. The dynamic tolerance calculation formula is: dynamic tolerance = base weight × (1 ± tolerance coefficient), where the tolerance coefficient = base coefficient + adjustment coefficient.

[0056] The weighbridge verification and interception execution module 140 is used to obtain the single standard weight and dynamic tolerance limit of the corresponding SKU of the current train, calculate the release area, compare the weighing result of the current train with the calculated release area, generate corresponding warning information and execute corresponding release strategies based on the comparison result.

[0057] By reading SKU master data and multi-dimensional data, risk scoring is performed on SKUs to obtain their corresponding risk levels. The standard weight per unit of each SKU is determined based on the packaging process BOM (Bill of Materials). Then, the SKU type is determined based on the standard weight per unit and the packaging tare weight waviness, establishing an SKU weight fingerprint database. A tolerance coefficient is determined based on the SKU risk level, and a dynamic tolerance limit is calculated based on this coefficient. When a trigger condition is met, the current tolerance coefficient is updated. The standard weight per unit and dynamic tolerance limit of the SKU corresponding to the current train are obtained, and the release area is calculated. The weighing result of the current train is compared with the calculated release area, and corresponding warning information is generated and the corresponding release strategy is executed based on the comparison result. Through SKU risk classification and differentiated tolerance settings, frequent false interceptions of low-risk SKUs by a "one-size-fits-all" threshold are avoided, while ensuring that anomalies of high-risk SKUs can be effectively detected. A differentiated dynamic tolerance coefficient based on SKU risk level was set, with a wider tolerance range for high-risk SKUs (providing reasonable room for error) and a narrower tolerance range for low-risk SKUs (achieving precise control).

[0058] In some embodiments, the SKU risk classification modeling module 110 is specifically used to normalize the original scores corresponding to the current SKU dimension data to obtain a normalized score, and then perform weighted summation according to the corresponding weights to obtain a comprehensive score, and determine the corresponding SKU risk level based on the comprehensive score.

[0059] Five-dimensional subdivision of scoring intervals and quantitative scoring rules

[0060] 1. Product Specifications (Weight 30%):

[0061]

[0062] Quantification rules: Score = Base score × Specification coefficient (Special specification 1.2, Regular 1.0, Standard 0.8).

[0063] 2. Packaging method (weight 25%):

[0064]

[0065] Quantification rules: Score = base score + packaging layer coefficient (5 points for each additional layer, up to 100 points).

[0066] 3. Historical weight deviation rate (weight 20%):

[0067]

[0068] Sample collection rules:

[0069] Data collection period: Shipment data from the past 6 months;

[0070] Removal rules: Remove the following outliers: rework and re-weighing data (records marked as "rework and re-weighing" by the system);

[0071] Data during weighbridge calibration (records marked as "weighbridge calibration" in the system); records cleared by human intervention (records marked as "abnormal clearing" in the system log); extreme outliers (records with deviation rate > mean + 3 standard deviations);

[0072] Sample size requirement: At least 30 valid samples must be collected for each SKU. If there are fewer than 30 valid samples, the actual number of samples will be used but marked as "insufficient samples".

[0073] Deviation rate calculation formula: Deviation rate = |Actual weight - Theoretical weight| ÷ Theoretical weight × 100%;

[0074] Overall deviation rate calculation: Overall deviation rate = weighted average of deviation rates of each batch.

[0075] Weighting: Data from the past month accounts for 40%, data from the past 2-3 months accounts for 35%, and data from the past 4-6 months accounts for 25%.

[0076] 4. Customer complaints / abnormal records (weight 15%):

[0077]

[0078] Single customer complaint point deduction rules:

[0079]

[0080] Weighted deduction rules for continuous customer complaints:

[0081]

[0082] Note: If a certain SKU receives a Category B customer complaint for 3 consecutive months: 20 points will be deducted for the first time, 30 points for the second time (20 x 1.5), and 40 points for the third time (20 x 2.0), for a total of 90 points. If there is no customer complaint for 1 month in between, the points will be deducted again from the first time.

[0083] Customer complaint statute of limitations rules:

[0084] Customer complaints are valid for 12 months; complaints older than 12 months will not be included in the rating.

[0085] Customer complaints in the past 6 months account for 100% of the weighting, while customer complaints in the past 6-12 months account for 50%.

[0086] 5. Measurement method (weight 10%)

[0087]

[0088] Quantification rule: Score = 100 - (Actual weight percentage × 70), that is, the higher the actual weight percentage, the lower the score.

[0089] II. Total Score Calculation Method

[0090] Comprehensive Risk Score Formula: Comprehensive Risk Score = Product Specification Score × 30% + Packaging Method Score × 25% + Historical Deviation Rate Score × 20% + Customer Complaint Record Score × 15% + Measurement Method Score × 10%;

[0091] The scores for each dimension must be normalized to ensure that they are on the same scale.

[0092] Normalized score = (actual score - lowest score in this dimension) ÷ (highest score in this dimension - lowest score in this dimension) × 100;

[0093] Example: A certain SKU has a historical deviation rate score of 80 points (maximum 100 points, minimum 10 points).

[0094] Normalized score = (80-10) ÷ (100-10) × 100 = 77.78 points;

[0095] Final overall score: Final overall score = Σ(normalized score of each dimension × corresponding weight).

[0096] III. Thresholds for Determining the Total Score in the A / B / C Levels

[0097]

[0098] Threshold adjustment rules:

[0099] If a SKU's rating remains stable within a certain range for three consecutive months, it can maintain its current rating.

[0100] If the score exceeds the current level range for two consecutive months, a level adjustment review will be triggered;

[0101] If a major customer complaint (Category A) occurs, it will be immediately upgraded to Level A, without being subject to scoring restrictions.

[0102] IV. Complete Scoring Example (White Cardboard Flat Sheet):

[0103]

[0104] Judgment result: Grade A (high risk), overall score 88.95 points ≥ 70 points.

[0105] V. Supporting Management Rules:

[0106] 1. Rating update frequency.

[0107] Daily updates: Historical deviation rate data is automatically updated after each batch is shipped;

[0108] Monthly Update: Customer complaint record ratings are updated at the beginning of each month;

[0109] Quarterly review: A comprehensive review of the SKU classification results is conducted every quarter.

[0110] 2. Rules for handling outliers.

[0111] Data collected during weighbridge calibration: Marked as "calibration data" and not included in the scoring;

[0112] Rework and re-weighing data: Marked as "rework and re-weighing" and removed from the sample;

[0113] Manual intervention to clear records: Marked as "abnormal clearing", triggering the audit process.

[0114] 3. Initial scoring rules for new SKUs.

[0115] When there is no historical data for a new SKU, it is temporarily set according to the default level of similar products (recommended level B);

[0116] After a new SKU has been shipped 30 times consecutively, it will be re-evaluated and graded based on actual data.

[0117] In some embodiments, the weight fingerprint library module 120 is used to determine the weight of the packaging of the SKU according to the packaging process BOM, obtain the standard weight of the SKU per piece according to the weight of the paper system and the packaging, calculate the actual weight according to the effective sampling amount of the SKU within a preset time, calculate the volatility according to the actual weight of the SKU, the standard weight per piece and the reverse calculation, and determine the type corresponding to the SKU according to the calculated volatility.

[0118] Formula for calculating tare weight volatility:

[0119] 1. Basic Definitions:

[0120] Tare weight: refers to the total weight of a single piece of finished paper including packaging materials, i.e., "paper system weight + packaging material weight".

[0121] Tare weight fluctuation rate: refers to the relative deviation between the actual tare weight and the standard tare weight of the same SKU in consecutive batches.

[0122] 2. Calculation formula:

[0123] Single-batch tare weight fluctuation rate: |Actual tare weight - Standard tare weight| ÷ Standard tare weight × 100%;

[0124] Overall tare weight volatility (used for SKU classification): ;

[0125] Where: n = number of valid sampling batches, volatility i = tare weight volatility of the i-th batch;

[0126] 3. Sampling batch requirements:

[0127]

[0128] Insufficient sample handling rules:

[0129] If the sample size for a certain SKU in the past 3 months is less than 30 batches, the calculation will be based on the actual sample size, but it will be marked "Insufficient sample size, results are for reference only";

[0130] If a SKU has a sample size of less than 10 batches in the past 3 months, it will not be included in the volatility classification and will be classified as "volatility category" by default.

[0131] Outlier removal rules:

[0132] Remove extreme outliers with a deviation rate greater than the mean plus three standard deviations;

[0133] Remove records marked as "rework and reweighing", "weighbridge verification", or "abnormality removal" from the system;

[0134] Abnormal weighing data caused by reasons such as the vehicle not coming to a complete stop or speeding during weighing are removed.

[0135] Specific outlier removal rules are as follows: Figure 2 As shown;

[0136] II. Thresholds for classifying SKUs into three categories: constant, fluctuating, and bulk.

[0137] 1. Classification criteria:

[0138]

[0139] 2. Dynamic adjustment rules for categories:

[0140]

[0141] 3. Correspondence between classification and tolerance coefficient:

[0142]

[0143] III. Sampling, mean, and standard deviation of standard weights for various types of packaging materials and their warehousing process;

[0144] 1. Packaging material classification and sampling rules:

[0145] Based on the packaging material data in the search results, the main packaging material categories are as follows:

[0146]

[0147] 2. Standard weight calculation process:

[0148] Step 1: Raw data collection. After each batch of sampling, record the following data: sampling date, time, name of packaging material, specifications, actual weight of a single item (accurate to 0.001kg), and signature of the sampling personnel.

[0149] Step 2: Calculate the mean: Where: i = the measured weight of the i-th sample, n = the number of valid samples.

[0150] Step 3: Standard Deviation Calculation: .

[0151] Step 4: Data fields for data entry:

[0152]

[0153] 3. Data Updates and Version Management:

[0154]

[0155] 4. Warehousing process illustration:

[0156] Sampling trigger -- On-site weighing record (accurate to 0.001kg) -- Data entry system -- Automatic verification (outlier removal) -- Calculate mean, standard deviation, and coefficient of variation -- Generate version number -- Write to packaging standard weight database -- Update TMS system baseline weight parameters -- Trigger tolerance coefficient recalculation.

[0157] IV. Complete Example (White Cardboard Roll Packaging):

[0158]

[0159] Tare weight volatility calculation:

[0160] Standard tare weight = Paper system weight (555kg) + Total weight of packaging materials (18.131kg) = 573.131kg;

[0161] If the actual tare weight of a certain batch is 574.200 kg, then the single-batch volatility = |574.200 - 573.131| ÷ 573.131 × 100% = 0.186%;

[0162] Overall volatility (30 lots in the last 3 months) = 0.21%;

[0163] Judgment result: Constant category (overall volatility 0.21% ≤ 0.5%).

[0164] V. Supporting Management Rules:

[0165] Sampling cycle requirements: Constant SKUs should be sampled once a month, fluctuating SKUs should be sampled at least once a week, and bulk SKUs should be sampled at least once a week.

[0166] Outlier handling: If the coefficient of variation of the weight of a certain batch of packaging materials is >5%, an anomaly warning will be triggered, and resampling and verification are required.

[0167] Data archiving: All sampled data are retained for at least 2 years to facilitate auditing and traceability.

[0168] System integration: The standard weight database for packaging materials is synchronized with the TMS system in real time, and the dynamic tolerance coefficient is automatically recalculated after each update.

[0169] In some embodiments, the dynamic tolerance calculation engine 130 is used to determine the corresponding base coefficient based on the risk level of the current SKU, determine the adjustment coefficient based on the historical average deviation rate, and determine the total tolerance coefficient based on the base coefficient and the adjustment coefficient.

[0170] Differentiated tolerance coefficients are set based on SKU risk levels. The calculation formula is: Dynamic Tolerance = Base Weight × (1 ± Tolerance Coefficient), where Tolerance Coefficient = Base Coefficient + Adjustment Coefficient. Specific tiered settings are as follows:

[0171]

[0172] In some embodiments, the dynamic tolerance calculation engine 130 is further configured to calculate a new tolerance coefficient based on the historical tolerance coefficient and the current batch deviation coefficient using a weighted average method.

[0173] I. Formula and triggering conditions for automatic coefficient update:

[0174] 1. Adjust coefficients to automatically update the formula.

[0175] Basic formula: ;

[0176] in: =Updated tolerance factor; =Current tolerance factor (base factor + adjustment factor); = Actual deviation rate (%) of the i-th batch; n = Number of valid batches within the statistical period; =Learning rate factor (default 0.3, configurable).

[0177] Simplified version (moving weighted average method): ;

[0178] in, =The moving average of the deviation rate of this batch (the average of the most recent N batches).

[0179] 2. Update trigger conditions:

[0180]

[0181] Example of cumulative deviation:

[0182] A certain Grade A SKU currently has a tolerance factor of ±0.28%, and has exceeded the tolerance for 3 consecutive batches.

[0183] First instance of exceeding tolerance: Record and mark as "Abnormal Batch 1";

[0184] • Second instance of exceeding tolerance: Recorded, marked as "Abnormal Batch 2", triggering an alert;

[0185] • Third instance of exceeding tolerance: Record it, mark it as "Abnormal Batch 3", and trigger automatic adjustment;

[0186] • New tolerance factor = 0.28% + 0.05% = 0.33%.

[0187] II. Tolerance coefficient upper and lower limit locking rules:

[0188] 1. Lock-in range of tolerance coefficients at each level:

[0189]

[0190] 2. Lockout rules:

[0191] Rule 1: Absolute value is locked. The absolute value of the tolerance coefficient must not exceed the upper limit of the level, nor fall below the lower limit of the level.

[0192] IF Cnew>Cmax THEN Cnew=Cmax,IF Cnew <Cmin THEN Cnew=Cmin。

[0193] Rule 2: The rate of change is locked, and the magnitude of a single update must not exceed the step size limit to prevent large fluctuations in the range due to extreme data in a single batch.

[0194] IF|Cnew-Cold|>step;THEN Cnew=Cold+sign(Cnew-Cold)×step.

[0195] Rule 3: Level linkage lock-in. If the tolerance coefficient reaches the upper limit three times in a row, the SKU level upgrade review will be triggered; if it reaches the lower limit three times in a row, the level downgrade review will be triggered.

[0196] If Cnew = Cmax for three consecutive times, then an A-level upgrade review will be triggered.

[0197] If Cnew = Cmin for three consecutive times, then THEN will trigger a C-level downgrade review.

[0198] Rule 4: Manual intervention coverage. The system administrator can manually lock the tolerance coefficient. Once locked, the system will no longer update automatically until it is manually unlocked.

[0199] III. Dynamic Calculation Logic for Vehicle Release Range Wtarerange Tare Weight Threshold:

[0200] 1. Basic Definitions:

[0201] Vehicle release range: refers to the weight range within which the system determines whether to release the vehicle (including cargo) after it has passed through the weighbridge.

[0202] Wtare_range=[Wbase(1-Clower),Wbase(1+Cupper)].

[0203] Where: Wbase = baseline vehicle weight = baseline cargo weight + vehicle tare weight; Clower = lower limit of tolerance coefficient; Cupper = upper limit of tolerance coefficient.

[0204] 2. Dynamic calculation of vehicle tare weight:

[0205] Vehicle tare weight standard value acquisition: The empty tare weight is recorded as the initial standard value for each vehicle's first weighing. Subsequent weighings are updated using a moving weighted average method.

[0206] Wvehicle_std=Wvehicle_old×0.8+Wvehicle_current×0.2.

[0207] Vehicle tare weight standard deviation calculation: .

[0208] 3. Dynamic fluctuation rules for tare weight threshold:

[0209]

[0210] Formula for calculating dynamic tare weight threshold:

[0211] Wvehicle_min=Wvehicle_std-k×σvehicle,

[0212] Wvehicle_max=Wvehicle_std}+k×σvehicle.

[0213] 2. Where k is a floating coefficient, which is determined based on the amount of data (see the table above).

[0214] 4. Complete calculation of the entire vehicle release range:

[0215] Wtare_min=Wgoods_base×(1-Clower)+Wvehicle_min;

[0216] Wtare_max=(Wgoods_base×(1+Cupper)+Wvehicle_max.

[0217] Decision logic: IF Wactual>=Wtaremin AND Wactual<=Wtaremax; THEN allow; ELSE block and alert; END IF.

[0218] IV. Dynamic Tolerance Calculation Engine 130 – Data Processing Flow as follows Figure 3 As shown.

[0219] Judgment Logic Flow: Start → Receive Weighbridge Data Packet → Parse Data Packet (License Plate Number, Gross Weight, Time, Weighbridge Number) → Data Validity Verification → Invalid / Valid, where: Invalid → Return Error and Log; where: Valid → Match Order Based on License Plate Number → Unmatched / Matched, where: Unmatched → Intercept (No Order Information); where: Successful Match → Obtain SKU Code from Order → Query SKU Master Data (Product Specifications, Packaging Method, Measurement Method) → Query SKU Risk Level (A / B / C) → Obtain Current Tolerance Factor (Base + Adjustment) → Apply Lock-in Rules (Absolute Value Lock-in, Rate Lock-in) → Query Packaging Standard Weight (from Weight Fingerprint Database) → Calculate Baseline Cargo Weight Wgoodsbase → Query Vehicle Tare Weight Record → New Vehicle / Historical, where: New Vehicle → Create Establish a file, with a 5% fluctuation range; The process includes: Historical data available → Obtain the standard value and standard deviation of vehicle tare weight → Determine the fluctuation coefficient k based on the amount of historical data → Calculate the upper and lower limits of vehicle tare weight → Calculate the vehicle release range Wtaremin~Wtaremax → Compare the actual gross weight with the release range → Within / Outside the range: Within the range → Release (open gate, issue document) → Update vehicle tare weight file → Record weighing log; Outside the range → Intercept (lock document, close gate, alarm) → Record anomaly log → Check if the tolerance coefficient update is triggered → 3 consecutive exceedances / not reaching the threshold → 3 consecutive exceedances → Automatically increase the tolerance coefficient; Not reaching the threshold → Only record; → Check if continuous abnormal alarms are triggered; 2 consecutive exceedances → Upgrade the alarm level; Only 1 → Regular alarm → End.

[0220] In some embodiments, the dynamic tolerance calculation engine 130 is further used to determine whether the difference between the new tolerance coefficient and the historical tolerance coefficient exceeds the adjustment step size. If it does, the historical tolerance coefficient is adjusted according to the adjustment step size to obtain the new tolerance coefficient.

[0221] In some embodiments, the dynamic tolerance calculation engine 130 obtains the final tolerance coefficient by taking the square root of the sum of squares of the newly calculated tolerance coefficient and the weighing instrument error coefficient.

[0222] I. Sources of error and the principle of superposition.

[0223] 1. Two types of error sources:

[0224]

[0225] 2. Superposition Principle: The two types of errors are independent and are superimposed using the sum of squares and root of square (RSS) method.

[0226] .

[0227] II. Permissible error values ​​for Class III weighing instruments:

[0228] 1. Weighing instrument grades and verification scale divisions:

[0229] According to JJG 539-2016, the relationship between the verification scale division value e and the number of verification scale divisions n for Class III weighing instruments is as follows:

[0230]

[0231] 2. Allowable error in segmented values

[0232]

[0233] III. Formula for calculating the Level 3 alarm threshold:

[0234] 1. Basic parameter definition:

[0235] The baseline vehicle weight is calculated by adding the vehicle tare weight to the baseline cargo weight.

[0236] Business dynamic tolerance coefficient, dynamically calculated according to SKU risk level (Level A: ±0.25%~±0.30%, Level B: ±0.18%~±0.20%, Level C: ±0.11%~±0.13%);

[0237] The legally permissible error coefficient for weighing instruments is ±0.10% (a conservative value).

[0238] : Prompt value coefficient, 0.5 (only 50% of the business tolerance is enabled);

[0239] Warning value coefficient, 0.8 (only 80% of business tolerance is enabled);

[0240] Forced interception value coefficient, 1.0 (only 100% of business tolerance is enabled).

[0241] 2. Three-level threshold calculation formula:

[0242] (1) Alert Level 1, trigger condition: the actual weight deviates from the benchmark value, but is still within the controllable range, prompting the operator to pay attention.

[0243] ;

[0244] ;

[0245] .

[0246] System action: A pop-up message "Please note the weight deviation" is displayed, the message is logged, and no action is taken to block it.

[0247] (2) Warning Level 2, trigger condition: the actual weight deviates significantly and requires manual verification.

[0248] ;

[0249] ;

[0250] .

[0251] System action: A pop-up warning "Weight deviation requires manual verification" is displayed, and the warehouse manager is notified to conduct on-site verification. The system will not be blocked at this time.

[0252] (3) Forced interception value (Intercept Level 3), trigger condition: the actual weight exceeds the superposition range of business tolerance and weighing instrument error, and must be intercepted.

[0253] ;

[0254] ;

[0255] .

[0256] System actions: gate interception, order locking, pop-up alarm, triggering exception handling process.

[0257] In some embodiments, the dynamic tolerance calculation engine 130 is further used to determine whether the obtained final tolerance coefficient exceeds the limit corresponding to the current risk level. If it does, the limit value is used as the final tolerance coefficient.

[0258] In some embodiments, it also includes:

[0259] The audit and anti-human intervention module is used to control access to the clearing of abnormal records, record the operator, operation time, operation reason and trigger a secondary confirmation mechanism;

[0260] The closed-loop feedback optimization module is used to send the cause of deviation, processing result and corrected weight back to the model database after each anomaly handling is completed.

[0261] Audit and anti-human intervention module: Access control is implemented for clearing abnormal records, limiting access to supervisors and above. The system automatically records the operator, time, and reason for the operation and triggers a secondary confirmation mechanism. All operation logs are tamper-proof, and audit reports are pushed periodically. When an operator is detected clearing an abnormal record, the algorithm automatically triggers the audit process, generates an audit work order, and pushes it to the audit supervisor and the corporate management manager, driving management-level accountability actions.

[0262] Closed-loop feedback optimization module: After each anomaly handling, data such as the cause of the deviation, the handling result, and the corrected weight are fed back to the model database for dynamic adjustment of the tolerance coefficient and model optimization. After each anomaly handling, the algorithm automatically adjusts the model parameters (tolerance coefficient, SKU level, etc.) based on the handling result, forming a complete closed loop of "perception-decision-execution-feedback-optimization".

[0263] This system constructs a multi-dimensional SKU risk grading model, incorporating five dimensions—product specifications, packaging method, historical weight deviation rate, customer complaint records, and measurement method—into a scoring system to achieve differentiated management. The historical weight deviation rate is calculated as: Deviation Rate = |Actual Weight - Theoretical Weight| ÷ Theoretical Weight × 100%. An SKU weight fingerprint database based on "standard weight per unit + packaging tare weight fluctuation rate" is established to eliminate the interference of packaging weight fluctuations on outbound verification and improve interception accuracy. A differentiated dynamic tolerance coefficient based on SKU risk level is set, with a wider tolerance range for high-risk SKUs (providing reasonable error tolerance) and a narrower tolerance range for low-risk SKUs (achieving precise management). A complete closed-loop feedback mechanism is established from "system interception → on-site verification → root cause analysis → data feedback → tolerance coefficient update," and the system possesses self-learning and adaptive optimization capabilities. Built-in audit tracking and anti-human intervention mechanisms are implemented, with access control, log recording, and secondary confirmation for clearing abnormal records, preventing human intervention to cover up anomalies. It has the following advantages:

[0264] Advantage 1: Reduced false positive rate and improved outbound efficiency. By using SKU risk grading and differentiated tolerance settings, frequent false interceptions of low-risk SKUs by a "one-size-fits-all" threshold are avoided, while ensuring that anomalies in high-risk SKUs can be effectively detected. It is estimated that compared with the existing fixed threshold scheme, the false interception rate can be reduced by 40% to 60%.

[0265] Advantage 2: Reduced customer complaints and economic losses. By intercepting anomalies such as overloading, underloading, mixed loading, and concealed items at the weighbridge stage, discrepancies in shipments are prevented from reaching the customer at the source. According to industry data, customer complaints caused by discrepancies in finished product logistics account for 15% to 30% of all customer complaints in the paper industry, resulting in direct economic losses of several million yuan annually. This system can significantly reduce the incidence of such complaints.

[0266] Advantage 3: The system has self-learning capabilities and continuous optimization. Through a closed-loop feedback mechanism, the system can dynamically adjust the tolerance coefficient and SKU risk level based on actual operating data, eliminating the need for regular manual maintenance. As data accumulates, the interception accuracy continues to improve.

[0267] Advantage 4: Effectively prevents the risk of human intervention. Through a triple mechanism of access control, log recording, and secondary confirmation, the system effectively constrains the deletion of abnormal records, solving the pain point of existing systems where "abnormalities cannot be traced after being covered up by human intervention".

[0268] Advantage 5: Reduced reliance on human experience. By systematizing and automating rules such as SKU classification, tolerance calculation, and anomaly detection, the reliance on the personal experience of warehouse staff in the shipping process is reduced, thus mitigating the impact of staff turnover on business stability.

[0269] Example 1: Interception due to low weight:

[0270] Scene description:

[0271] • Time: 14:32 on June 15, 2026;

[0272] •SKU: White cardboard sheet (BK-PL-787-1092, Grade A high risk);

[0273] • Order: 20 packages, base cargo weight 3,345.120 kg;

[0274] • License plate: Min E·85708, standard tare weight of vehicle: 8,520 kg, standard deviation: 32.5 kg;

[0275] • Vehicle release range: 11,790.064kg~11,939.936kg;

[0276] • Actual gross weight: 11,700.000 kg (below the lower limit of 90.064 kg).

[0277] The system pop-up window content is as follows Figure 4 As shown.

[0278] Approval process:

[0279]

[0280] Deviation Reason Archive:

[0281] Anomaly Log ID: ABN-20260615-0032

[0282] Type: Low weight (small quantity)

[0283] SKU: BK-PL-787-1092

[0284] Order Number: ORD-20260615-0289

[0285] License plate: Min E·85708

[0286] Actual gross weight: 11,700.000 kg

[0287] Release limit: 11,790.064 kg

[0288] Deviation: -90.064kg

[0289] Deviation rate: -0.76%

[0290] Cause classification: Insufficient filling

[0291] Detailed description: The warehouse clerk found that 18 bags were actually loaded onto the truck, while the order required 20 bags; 2 bags were missing. The reason was that the forklift driver missed loading them.

[0292] Responsible Department: Finished Goods Warehouse

[0293] Responsible person: Forklift driver

[0294] Handling measures: After refilling with 2 more packages, weigh them again and release them.

[0295] Archived at: 2026-06-15 15:10:00.

[0296] Model parameters are updated automatically:

[0297] Step 1: Check the consecutive out-of-tolerance count; Current consecutive out-of-tolerance count: 0 (previous batch was normal), current out-of-tolerance count → consecutive out-of-tolerance count = 1.

[0298] Step 2: Calculate the new tolerance factor (moving weighted average method); Historical tolerance factor (before adjustment): ±0.28%; Deviation rate of this batch: 0.76%; New tolerance factor = 0.28%×0.7+0.76%×0.3=0.196%+0.228%=0.424%.

[0299] Step 3: Apply the lockout rules; Level A upper limit: ±0.50%, 0.424% < 0.50%, upper limit lockout not triggered; Check the rate of change: step size 0.05%, |0.424% - 0.28%| = 0.144% > 0.05%, rate lockout triggered; New tolerance coefficient = 0.28% + 0.05% = 0.33%.

[0300] Step 4: Update the database;

[0301] Step 5: Record the update history.

[0302] Example 2: Interception due to excessive weight:

[0303] Scene description:

[0304] • Time: 09:15, June 18, 2026;

[0305] •SKU: Cultural Paper Rolls (CP-RL-635-1200, Class B Medium Risk);

[0306] • Order: 12 rolls, base cargo weight 5,875.836 kg;

[0307] • License plate: Min D J5238, standard value of vehicle tare weight: 12,350 kg, standard deviation: 28.6 kg;

[0308] • Vehicle release range: 18,155.165kg~18,303.835kg;

[0309] • Actual gross weight: 18,350,000 kg (higher than the upper limit of 46,165 kg).

[0310] System pop-up window such as Figure 5 As shown.

[0311] Approval process:

[0312]

[0313] Deviation Reason Archive:

[0314] Exception record ID: ABN-20260618-0015; Type: Excessive weight (mixed loading); SKU: CP-RL-635-1200; Order number: ORD-20260618-0056; License plate: Min D·J5238; Actual gross weight: 18,350.000kg; Release upper limit: 18,303.835kg; Deviation: +46.165kg; Deviation rate: +0.25%; Cause category: Mixed loading; Detailed description: Security personnel inspection found 2 rolls of cultural paper reels with specification CP-RL-635-1100 (non-products of this order) mixed-loaded in the carriage, which was verified to be left over from the unloading of the previous trip. The mixed-loaded paper rolls have been unloaded and reweighed. Responsible department: Finished product warehouse; Responsible person: Loader of the previous trip; Treatment measure: Unload the mixed-loaded paper rolls, reweigh and release, and return the mixed-loaded paper rolls to the warehouse; Archiving time: 2026-06-18 10:00:00.

[0315] Automatic update of model parameters:

[0316] Step 1: Check the consecutive out-of-tolerance count; Current consecutive out-of-tolerance count: 0 (the previous batch is normal), this out-of-tolerance → consecutive out-of-tolerance count = 1;

[0317] Step 2: Calculate a new tolerance coefficient; Historical tolerance coefficient (before adjustment): ±0.19%, deviation rate of this batch: 0.25%; New tolerance coefficient = 0.19%×0.7+0.25%×0.3=0.133%+0.075%=0.208%;

[0318] Step 3: Apply the locking rule; Class B upper limit: ±0.35%, 0.208%<0.35%, upper limit locking is not triggered; Change rate: |0.208%-0.19%|=0.018%<step size 0.03%, rate locking is not triggered; Final new tolerance coefficient: ±0.208%.

[0319] Step 4: Update the database.

[0320] In other embodiments, there are also functions such as upgrading alarms for consecutive 2 trips of exceptions, and triggering audit and accountability when an operator deletes an exception record.

[0321] Please refer to Figure 6 , a dynamic tolerance error-proof method for finished paper outbound applied to a weighbridge, the method is applied to the above-mentioned embodiment of the dynamic tolerance error-proof system for finished paper outbound applied to a weighbridge, comprising the following steps:

[0322] Step S610: When a vehicle drives onto the weighbridge for weighing, obtain the SKU master data and multi-dimensional data of the SKU of the current trip, and calculate the risk level of the corresponding SKU based on the multi-dimensional data;

[0323] Step S620: Determine the standard weight of a single unit based on the packaging process BOM of the SKU, and determine the type of the SKU based on the standard weight of the single unit and the tare weight waviness of the packaging.

[0324] Step S630: Determine the tolerance coefficient based on the SKU risk level, and calculate the dynamic tolerance limit based on the tolerance coefficient;

[0325] Step S640: Calculate the release area based on the standard weight of a single piece and the dynamic tolerance limit of the corresponding SKU for the current train;

[0326] Step S650: Compare the weighing result of the current train with the calculated release area, generate corresponding warning information and execute corresponding release strategies based on the comparison result.

[0327] By reading SKU master data and multi-dimensional data, risk scoring is performed on SKUs to obtain their corresponding risk levels. The standard weight per unit of each SKU is determined based on the packaging process BOM (Bill of Materials). Then, the SKU type is determined based on the standard weight per unit and the packaging tare weight waviness, establishing an SKU weight fingerprint database. A tolerance coefficient is determined based on the SKU risk level, and a dynamic tolerance limit is calculated based on this coefficient. When a trigger condition is met, the current tolerance coefficient is updated. The standard weight per unit and dynamic tolerance limit of the SKU corresponding to the current train are obtained, and the release area is calculated. The weighing result of the current train is compared with the calculated release area, and corresponding warning information is generated and the corresponding release strategy is executed based on the comparison result. Through SKU risk classification and differentiated tolerance settings, frequent false interceptions of low-risk SKUs by a "one-size-fits-all" threshold are avoided, while ensuring that anomalies of high-risk SKUs can be effectively detected. A differentiated dynamic tolerance coefficient based on SKU risk level was set, with a wider tolerance range for high-risk SKUs (providing reasonable room for error) and a narrower tolerance range for low-risk SKUs (achieving precise control).

[0328] Finally, it should be noted that although the above embodiments have been described in the text and drawings of this application, this should not limit the scope of patent protection of this application. Any technical solutions that are based on the essential concept of this application and utilize the content described in the text and drawings of this application, resulting in equivalent structural or procedural substitutions or modifications, as well as the direct or indirect application of the technical solutions of the above embodiments to other related technical fields, are all included within the scope of patent protection of this application.

Claims

1. A dynamic tolerance and error-proofing system for finished paper outbound shipment applied to weighbridges, characterized in that, include: The SKU risk grading modeling module is used to read SKU master data and multi-dimensional data, and to score the risk of SKU based on the multi-dimensional data to obtain the corresponding SKU risk level. The multi-dimensional data includes product specifications, packaging method data, historical weight deviation rate and customer complaint records. The weight fingerprint database module is used to determine the standard weight of a single SKU based on the packaging process BOM, and to determine the type of the SKU according to the standard weight of a single SKU and the packaging tare weight waviness, and to establish the SKU weight fingerprint database. The dynamic tolerance calculation engine is used to determine the tolerance coefficient based on the SKU risk level and calculate the dynamic tolerance limit based on the tolerance coefficient. When the trigger condition is met, the current tolerance coefficient is updated. The weighbridge verification and interception execution module is used to obtain the standard weight of a single piece and the dynamic tolerance limit of the corresponding SKU of the current vehicle, calculate the release area, compare the weighing result of the current vehicle with the calculated release area, generate corresponding warning information and execute corresponding release strategies based on the comparison result.

2. The dynamic tolerance and error-proofing system for finished paper outbound shipment applied to weighbridges according to claim 1, characterized in that, The SKU risk grading modeling module is specifically used to normalize the original scores corresponding to the current SKU dimension data to obtain a normalized score, and then perform weighted summation according to the corresponding weights to obtain a comprehensive score. The corresponding SKU risk level is determined based on the comprehensive score.

3. The dynamic tolerance and error-proofing system for finished paper outbound shipment applied to weighbridges according to claim 1, characterized in that, The weight fingerprint database module is used to determine the packaging weight of the SKU based on the packaging process BOM, obtain the standard weight of the SKU per unit based on the paper system weight and the packaging, calculate the actual weight based on the effective sampling amount of the SKU within a preset time, calculate the volatility based on the actual weight of the SKU, the standard weight per unit, and the inverse calculation, and determine the type of the SKU based on the calculated volatility.

4. The dynamic tolerance and error-proofing system for finished paper outbound shipment applied to weighbridges according to claim 1, characterized in that, The dynamic tolerance calculation engine is used to determine the corresponding base coefficient based on the risk level of the current SKU, determine the adjustment coefficient based on the historical average deviation rate, and determine the total tolerance coefficient based on the base coefficient and the adjustment coefficient.

5. The dynamic tolerance and error-proofing system for finished paper outbound shipment applied to weighbridges according to claim 4, characterized in that, The dynamic tolerance calculation engine is also used to calculate a new tolerance coefficient based on the historical tolerance coefficient and the deviation coefficient of the current batch using a weighted average method.

6. The dynamic tolerance and error-proofing system for finished paper outbound from weighbridges as described in claim 5, characterized in that, The dynamic tolerance calculation engine is also used to determine whether the difference between the new tolerance coefficient and the historical tolerance coefficient exceeds the adjustment step size. If it does, the historical tolerance coefficient is adjusted according to the adjustment step size to obtain the new tolerance coefficient.

7. The dynamic tolerance and error-proofing system for finished paper outbound shipment applied to weighbridges according to claim 5, characterized in that, The dynamic tolerance calculation engine also calculates the square root of the sum of squares of the newly calculated tolerance coefficient and the weighing instrument error coefficient to obtain the final tolerance coefficient.

8. The dynamic tolerance and error-proofing system for finished paper outbound shipment applied to weighbridges according to claim 7, characterized in that, The dynamic tolerance calculation engine is also used to determine whether the final tolerance coefficient exceeds the limit corresponding to the current risk level. If it does, the limit value is used as the final tolerance coefficient.

9. The dynamic tolerance and error-proofing system for finished paper outbound shipment applied to weighbridges according to claim 7, characterized in that, Also includes: The audit and anti-human intervention module is used to control access to the clearing of abnormal records, record the operator, operation time, operation reason and trigger a secondary confirmation mechanism; The closed-loop feedback optimization module is used to send back the cause of deviation, processing result and corrected weight to the model database after each anomaly handling is completed.

10. A method for dynamic tolerance error prevention of finished paper outbound shipment applied to weighbridges, wherein the method is applied to the dynamic tolerance error prevention system for finished paper outbound shipment applied to weighbridges as described in any one of claims 1-10, comprising the following steps: When a vehicle enters the weighbridge, the SKU master data and multi-dimensional data of the current vehicle's SKU are obtained, and the risk level of the corresponding SKU is calculated based on the multi-dimensional data. The standard weight of a single unit is determined based on the packaging process BOM based on the SKU, and the type of the corresponding SKU is determined based on the standard weight of the single unit and the tare weight waviness of the packaging. Tolerance coefficients are determined based on SKU risk levels, and dynamic tolerance limits are calculated based on these tolerance coefficients. Calculate the release area based on the standard weight of a single piece and the dynamic tolerance limit of the corresponding SKU for the current train service; The weighing result of the current train is compared with the calculated release area, and corresponding warning information is generated and the corresponding release strategy is executed based on the comparison result.

Images