Batch cargo proofreading and early warning method and system for portal channel machines

Abstract

The present invention belongs to the field of channel machine inventory, and in particular, relates to a batch cargo proofreading and early warning method and system for a gate-type channel machine. The method generates a user batch demand order table by using a demand matching prediction model through user historical orders and limited demand data. Secondly, a unique radio frequency identification code is configured for the batch cargo, and the radio frequency information and seasonal curve of the cargo are obtained; thirdly, a matching judgment threshold is set, and the matching degree of the initial information of the cargo is calculated by a matching judgment algorithm according to the demand order table and radio frequency information; for cargo whose matching degree does not meet the threshold, a secondary function replacement matching judgment is performed using functional attribute information and random sampling information; finally, whether the cargo meets the user's needs is determined based on the secondary matching results and the seasonal function threshold, and the user batch demand order table is updated or the early warning device is triggered for early warning; this method effectively improves the accuracy and efficiency of batch cargo proofreading, and reduces the risk of wrong shipments and delays.

Description

Technical Field

[0001] The invention belongs to the field of inventory counting of channel machines, and in particular relates to a batch cargo proofreading and early warning method and system for a gate-type channel machine. Background Art

[0002] With the rapid development of the logistics and warehousing industries, efficient cargo management and accurate proofreading have become essential. Traditional door-type channel machines mainly rely on barcode scanning and RFID tag reading to identify and proofread cargo information. However, these methods have significant shortcomings: barcode scanning is inefficient, scanning one by one leads to slow processing speed, and is easily affected by contamination; RFID technology is fast but costly, and its performance is unstable in complex electromagnetic environments. In addition, the existing system lacks an effective early warning mechanism and cannot promptly detect and deal with problems such as inconsistent cargo quantities or damaged labels. The functions are also relatively simple, usually only simple identification or counting can be performed, and information from multiple data sources cannot be comprehensively processed.

[0003] For example, a Chinese patent with authorization announcement number CN118396527B discloses a method and system for counting goods in and out of the warehouse based on target tracking. The method includes the following steps: collecting goods video; obtaining a matching template, obtaining all particles in the current goods image, and calculating the matching deviation between the particles and the matching template; if the minimum matching deviation is less than a preset deviation threshold, obtaining the target goods area in the currently processed goods image; if the minimum matching deviation is greater than the preset deviation threshold, calculating the necessity of updating the template; if the necessity of updating the template is greater than the preset necessity threshold, updating the matching template, and performing positioning analysis using the updated matching template; if the necessity of updating the template is less than the preset update necessity threshold, restarting the positioning analysis from the current goods image; completing the positioning analysis of the target goods area of ​​each frame of the goods image to assist in the early warning of the goods in and out of the warehouse inventory.

[0004] The above existing technologies have the following problems: the existing portal channel machines only obtain inventory and proofreading to improve the accuracy of recognition, which requires a complete and accurate demand list to be given in advance, but in many cases the user can only give a rough demand, which requires the portal channel machine to be able to intelligently determine whether the inventory-counted goods are what the user needs. In addition, sometimes the goods required by the user are missing, and the merchant will use other goods as substitutes. The existing portal channel machine inventory and proofreading method cannot intelligently determine whether the substitute is feasible. For this reason, the present invention provides a batch goods proofreading early warning method and system for the portal channel machine. Summary of the invention

[0005] In view of the shortcomings of the prior art, the present invention proposes a batch cargo proofreading and early warning method and system for a gate-type channel machine. The method generates a user batch demand order table by using a demand matching prediction model through user historical orders and limited demand data. Secondly, a unique radio frequency identification code is configured for the batch cargo, and the radio frequency information of the cargo is obtained; thirdly, a matching judgment threshold is set, and the matching degree of the initial information of the cargo is calculated by a matching judgment algorithm according to the demand order table and radio frequency information; for cargo whose matching degree does not meet the threshold, a secondary function replacement matching judgment is performed using functional attribute information and random sampling information; finally, whether the cargo meets the user's needs is determined based on the secondary matching results, and the user batch demand order table is updated or the early warning device is triggered for early warning; this method effectively improves the accuracy and efficiency of batch cargo proofreading and reduces the risk of wrong shipments and delays.

[0006] To achieve the above object, the present invention provides the following technical solutions:

[0007] A batch cargo proofreading and early warning method for a portal channel machine comprises the following steps:

[0008] S1. Obtain the user's historical batch order data and limited demand data, and obtain the user's batch demand order table through the configured demand matching prediction model;

[0009] S2. Obtain the user's batch demand goods information and configure a unique radio frequency identification code, obtain the radio frequency information of the batch goods through the configured radio frequency identification algorithm, and construct a seasonal change curve of the corresponding type of batch goods according to the radio frequency information of the batch goods;

[0010] S3. Set a matching discrimination threshold, and obtain the matching degree of the initial information of each batch type of goods through a matching discrimination algorithm according to the user batch demand order table and the batch goods radio frequency information;

[0011] S4. When the initial information matching degree of a certain batch type of goods is greater than the matching judgment threshold, the corresponding type of batch goods passes the verification. If it is less than or equal to the matching judgment threshold, the functional attribute information of the randomly sampled goods in the batch goods of the type that does not meet the matching judgment threshold and the functional attribute information corresponding to the user batch demand order table are used to perform secondary function replacement matching judgment through the matching judgment algorithm to obtain the secondary function matching degree of the corresponding goods;

[0012] S5. Configure the seasonal function matching threshold according to the seasonal change curve. If the secondary function matching of the corresponding goods is greater than the seasonal function matching threshold, the corresponding goods are judged to meet the user's needs. If it is less than or equal to the seasonal function matching threshold, the corresponding goods are judged to not meet the user's needs. The information of the goods that are secondarily judged to meet the user's needs is fed back to the user batch demand order table for information update. The information of the goods that are secondarily judged not to meet the user's needs is fed back to the early warning device for early warning.

[0013] The steps for configuring the seasonal function matching threshold according to the seasonal change curve include:

[0014] S501. According to the corresponding goods type in the radio frequency information of the batch goods, the sales volume data and purchase volume data of the corresponding type of goods in different historical time periods are obtained, and the double-axis seasonal change curves of different types of goods and the corresponding curve functions are constructed with the time axis as the horizontal axis and the sales volume data and purchase volume data of different time periods as the double vertical axes.

[0015] Specifically, the steps for building the demand matching prediction model include:

[0016] S101. Construct a limited demand information input sequence based on the user's historical batch order data and limited demand data and cargo information search space ,in, represents the type of the i-th commodity, The table represents the color, size, style and quality grade information of the i-th type of goods. represents the seasonal timestamp information corresponding to the i-th commodity, represents the keyword text information in the limited demand data corresponding to the jth user, represents the probability that the i-th product purchased by the user is a seasonal product, represents the historical purchase frequency of the i-th commodity, It represents the predicted probability that the i-th product meets the user's needs;

[0017] S102: Input limited demand information into sequence and cargo information search space Input into the model information extraction layer of the demand matching prediction model to obtain the limited demand fuzzy features and the cargo information search feature set;

[0018] Specifically, the steps of constructing the demand matching prediction model also include:

[0019] S103. According to the fuzzy characteristics of limited demand, the perceived utility matrix and disappointment preference function of user j for the i-th and k-th goods under the evaluation indicators of goods color, size, style, quality grade, probability of meeting user demand and probability of seasonal goods are constructed through disappointment theory and historical goods selection results, and the constructed perceived utility matrix and disappointment preference function are built into the demand discrimination search layer;

[0020] S104, inputting the limited demand fuzzy features and the goods information search feature set into the demand discrimination search layer, using the configured disappointment preference function to perform demand goods type matching search, and obtaining the user's disappointment value corresponding to each type of goods and the corresponding search matching error probability loss between the user's demand and the goods;

[0021] S105, constructing a training loss function according to the user's disappointment value corresponding to each type of goods and the search matching error loss corresponding to the user's demand and the goods, and embedding the training loss function into the demand matching prediction model for training, and outputting the probability of each type of goods meeting the user's demand and the search matching accuracy rate of all demand types of goods according to the user's disappointment value corresponding to each type of goods through the output layer;

[0022] Specifically, the steps of constructing the demand matching prediction model also include:

[0023] S106, setting a search matching accuracy threshold and a user demand satisfaction rate threshold, when the probability that each type of goods meets the user's demand is greater than the user demand satisfaction rate threshold and the search matching accuracy of all demand types of goods is greater than the search matching accuracy threshold, obtaining a trained demand matching prediction model and obtaining the user batch demand goods type predicted by the search;

[0024] S107. Based on the search-predicted user batch demand goods types and the corresponding goods information table, a user batch demand order table is constructed.

[0025] Specifically, the step of configuring the seasonal function matching threshold according to the seasonal change curve also includes:

[0026] S501. According to the corresponding goods type in the radio frequency information of the batch goods, the sales volume data and the purchase volume data of the corresponding goods in different time periods in history are obtained, and the double-axis seasonal change curves of different types of goods and the corresponding curve functions are constructed with the time axis as the horizontal axis and the sales volume data and the purchase volume data of different time periods as the double vertical axes;

[0027] S502, predicting sales volume and purchase volume according to the sales volume change curve function and the purchase volume change curve function, and optimizing the dual-axis seasonal change curve and the sales volume change curve function and the purchase volume change curve function by using the difference between the predicted value and the actual value;

[0028] S503, obtaining the sales volume and purchase volume at the next moment according to the optimized sales volume change curve function and purchase volume change curve function;

[0029] S504. Obtain the sales volume ratio at the current moment and the sales volume ratio at the next moment according to the sales volume and the purchase volume at the current moment and the sales volume and the purchase volume at the next moment.

[0030] Specifically, the step of configuring the seasonal function matching threshold according to the seasonal change curve also includes:

[0031] S505, setting a benchmark function matching threshold for the i-th type of goods, and constructing a genetic algorithm input sequence according to the benchmark function matching threshold, the type of goods that do not meet the matching judgment threshold, the corresponding sales ratio at the current moment and the sales ratio at the next moment, the disappointment value of the corresponding goods, and the probability of meeting user needs;

[0032] S506. Constructing a seasonal function matching threshold function and corresponding constraint conditions corresponding to the i-th type of goods according to the genetic algorithm input sequence;

[0033] S507, setting a training cycle threshold, inputting the genetic algorithm input sequence, the seasonal function matching threshold function and the corresponding constraint conditions into the genetic algorithm for training, and when the training cycle is greater than the training cycle threshold, stopping the iteration, and obtaining the seasonal function matching threshold of the corresponding goods at the corresponding time.

[0034] Specifically, the steps of obtaining the randomly sampled goods in S4 include:

[0035] According to the obtained batch goods that do not meet the matching judgment threshold type, the corresponding batch goods are sampled in layers to obtain random sampling goods.

[0036] The batch cargo proofreading and early warning system for the gate-type channel machine includes: demand prediction module, radio frequency information module, matching recognition module and abnormality discrimination module;

[0037] The demand forecasting module is used to obtain the user batch demand order table by using the user's historical batch order data and limited demand data through the configured demand matching forecasting model;

[0038] The radio frequency information module is used to configure a unique radio frequency identification code using the user's batch demand goods information, obtain the radio frequency information of the batch goods through the configured radio frequency identification algorithm, and construct a seasonal change curve of the corresponding type of batch goods based on the radio frequency information of the batch goods;

[0039] The matching and identification module includes a matching unit and an initial discrimination unit; the matching unit is used to obtain the matching degree of the initial information of each batch type of goods through a matching discrimination algorithm according to the user batch demand order table and the batch goods radio frequency information;

[0040] The primary identification unit is used to verify the corresponding type of batch goods according to the initial information matching degree and matching identification threshold of each batch type of goods. If it is greater than the matching identification threshold, the corresponding type of batch goods passes the verification; if it is less than or equal to the matching identification threshold, a secondary functional verification is performed.

[0041] Specifically, the abnormality discrimination module includes a secondary matching unit, a secondary discrimination unit and an updating unit;

[0042] A secondary matching unit is used to use the functional attribute information of the batch goods that do not meet the condition type and the functional attribute information corresponding to the random sampling information and the user batch demand order table to perform secondary function replacement matching judgment through a matching judgment algorithm to obtain the secondary function matching degree of the corresponding goods that do not meet the initial matching;

[0043] The secondary discrimination unit is used to determine whether the corresponding goods meet the user's needs based on the secondary function matching degree of the corresponding goods and the seasonal function matching degree threshold. If it is greater than the function matching degree threshold, it is determined that the corresponding goods meet the user's needs; if it is less than or equal to, it is determined that the user's needs are not met.

[0044] A computer-readable storage medium stores computer instructions, which, when executed, execute a batch cargo proofreading and early warning method for a gate-type channel machine.

[0045] Compared with the prior art, the present invention has the following beneficial effects:

[0046] In view of the deficiencies in the prior art, the present invention integrates user historical orders with limited demand data, builds a demand forecasting model, and accurately generates batch demand order tables without the need for a complete and accurate demand table; at the same time, the use of radio frequency identification technology and matching judgment algorithm not only improves the accuracy of goods identification, but also can intelligently judge whether the goods meet user needs. Even in the face of ambiguity in user needs or commodity substitution, it can also make intelligent judgments through secondary function substitution matching judgments; in addition, the present invention configures seasonal function matching thresholds, and adjusts the seasonal function matching thresholds according to whether the corresponding functions of the replacement products meet user needs and the seasonal characteristics of the replacement products, so that the corresponding replacement products can not only make up for the shortage of merchants' goods, but also better meet the corresponding user's goods needs when the goods are insufficient, thereby improving user satisfaction; this method effectively improves the proofreading efficiency and accuracy of the gate channel machine, reduces the risk of misjudgment, and ensures accurate matching and sufficient distribution of goods demand. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 This is a flow chart of a batch cargo proofreading and early warning method for a portal channel machine according to Embodiment 1 of the present invention;

[0048] Figure 2 This is an application scenario architecture diagram corresponding to the batch cargo proofreading and early warning method for a portal access machine according to Embodiment 1 of the present invention;

[0049] Figure 3 This is a module diagram of a batch cargo proofreading and warning system for a portal access machine according to Example 2 of the present invention.

[0050] Explanation of the accompanying reference numerals: 001, bulk goods to be inspected; 002, portal channel machine body; 003, radio frequency detection device; 004, processing output terminal; 005, early warning device. DETAILED DESCRIPTION

[0051] Example 1

[0052] Figure 2 The present invention is a diagram of an application scenario architecture corresponding to the batch cargo proofreading and early warning method for a portal access machine, which is applied to scenes such as warehouses and logistics centers where frequent cargo in-and-out management is required, wherein 001 represents the batch cargo to be detected, 002 represents the portal access machine body, 003 represents the radio frequency detection device configured on the portal access machine, 004 is the portal access machine detection result processing and output terminal, and 005 represents the early warning device configured on the portal access machine; in this embodiment, the batch cargo 001 to be detected is detected by the radio frequency detection device 003 configured on the portal access machine body 002, and the detection information is input into the processing and output terminal 004, and the cargo matching judgment and inventory are performed, and the judgment result is displayed through the processing and output terminal 004, and the identified abnormal cargo information is transmitted back to the early warning device 005 on the portal access machine for early warning;

[0053] See also Figure 1 The present invention provides an embodiment: a batch cargo proofreading and early warning method for a portal channel machine, the steps comprising:

[0054] S1. Obtain the user's historical batch order data and limited demand data, and obtain the user's batch demand order table through the configured demand matching prediction model;

[0055] Furthermore, the steps of constructing the demand matching prediction model in this embodiment include:

[0056] S101. Construct a limited demand information input sequence based on the user's historical batch order data and limited demand data and cargo information search space ,in, represents the type of the i-th commodity, The table represents the color, size, style and quality grade information of the i-th type of goods. represents the seasonal timestamp information corresponding to the i-th commodity, represents the keyword text information in the limited demand data corresponding to the jth user, represents the probability that the i-th product purchased by the user is a seasonal product, represents the historical purchase frequency of the i-th commodity, It represents the predicted probability that the i-th product meets the user's needs;

[0057] Furthermore, in this embodiment, the goods information search space is constructed by the user's historical purchase of goods types and corresponding basic information and the purchase merchant's corresponding goods types and basic information. The corresponding basic information includes goods type, color, size, style and quality grade and corresponding functional information;

[0058] S102: Input limited demand information into sequence and cargo information search space Input into the model information extraction layer of the demand matching prediction model to obtain the limited demand fuzzy features and the cargo information search feature set;

[0059] Furthermore, the model information extraction layer in this embodiment is constructed through the resnet-18 network;

[0060] S103. According to the fuzzy characteristics of limited demand, the perceived utility matrix and disappointment preference function of user j for the i-th and k-th goods under the evaluation indicators of goods color, size, style, quality grade, probability of meeting user demand and probability of seasonal goods are constructed through disappointment theory and historical goods selection results, and the constructed perceived utility matrix and disappointment preference function are built into the demand discrimination search layer;

[0061] S104, inputting the limited demand fuzzy features and the goods information search feature set into the demand discrimination search layer, using the configured disappointment preference function to perform demand goods type matching search, and obtaining the user's disappointment value corresponding to each type of goods and the corresponding search matching error probability loss between the user's demand and the goods;

[0062] Furthermore, the demand discrimination search layer in this embodiment is constructed by the SENET attention network;

[0063] S105, constructing a training loss function according to the user's disappointment value corresponding to each type of goods and the search matching error loss corresponding to the user's demand and the goods, and embedding the training loss function into the demand matching prediction model for training, and outputting the probability of each type of goods meeting the user's demand and the search matching accuracy rate of all demand types of goods according to the user's disappointment value corresponding to each type of goods through the output layer in the demand matching prediction model; further, in this embodiment, the output layer is constructed by a fully connected network;

[0064] Furthermore, the search matching accuracy rate of all demand type goods in this embodiment is the ratio of the number of searched goods types that are the same as the real goods types to the number of real goods types;

[0065] S106, setting a search matching accuracy threshold and a user demand satisfaction rate threshold, when the probability that each type of goods meets the user's demand is greater than the user demand satisfaction rate threshold and the search matching accuracy of all demand types of goods is greater than the search matching accuracy threshold, obtaining a trained demand matching prediction model and obtaining the user batch demand goods type predicted by the search;

[0066] S107. Based on the search-predicted user batch demand goods types and the corresponding goods information table, a user batch demand order table is constructed.

[0067] Through the construction and application of the above-mentioned demand matching prediction model, this process can significantly improve the accuracy and personalized service level of the user's batch demand order table; first, the model integrates the user's limited demand data with historical batch order information, not only considering the specific attributes of the goods such as color, size, style and quality grade, but also combining seasonal timestamp information and seasonal goods probability, so as to more accurately capture the user's potential demand preferences; secondly, the introduction of disappointment theory and perceived utility matrix enables the model to fully consider the user's subjective feelings and possible disappointment when evaluating whether the goods meet the user's needs, thereby enhancing the consistency and satisfaction of the user experience; in addition, the built-in disappointment preference function and training loss function in the model ensure the effective control of the error probability in the search and matching process, and improve the accuracy of search matching; in particular, setting the search matching accuracy threshold and the user demand satisfaction rate threshold further guarantees the quality of the output results. Only when the model reaches the preset standard will it be confirmed that the training is completed, ensuring the stability and reliability of the model; finally, the user batch demand order table generated based on the optimized model can better guide inventory management and supply chain decisions, reduce the occurrence of excess or shortage, and improve overall operational efficiency.

[0068] S2. Obtain the user's batch demand goods information and configure a unique radio frequency identification code, obtain the radio frequency information of the batch goods through the configured radio frequency identification algorithm, and construct a seasonal change curve of the corresponding type of batch goods according to the radio frequency information of the batch goods;

[0069] S3. Set a matching discrimination threshold, and obtain the matching degree of the initial information of each batch type of goods through a matching discrimination algorithm according to the user batch demand order table and the batch goods radio frequency information;

[0070] S4. When the initial information matching degree of a certain batch type of goods is greater than the matching judgment threshold, the corresponding type of batch goods passes the verification. If it is less than or equal to the matching judgment threshold, the functional attribute information of the randomly sampled goods in the batch goods of the type that does not meet the matching judgment threshold and the functional attribute information corresponding to the user batch demand order table are used to perform secondary function replacement matching judgment through the matching judgment algorithm to obtain the secondary function matching degree of the corresponding goods;

[0071] Furthermore, in this embodiment, the step of obtaining the randomly sampled goods includes:

[0072] According to the obtained batch of goods that do not meet the matching judgment threshold type, the corresponding batch of goods is subjected to a stratified extraction algorithm to obtain randomly sampled goods. Further, the stratified extraction algorithm in this embodiment is a prior art and will not be described in detail.

[0073] S5. Configure the seasonal function matching threshold according to the seasonal change curve. If the secondary function matching of the corresponding goods is greater than the seasonal function matching threshold, the corresponding goods are judged to meet the user's needs. If it is less than or equal to the threshold, it is judged that the corresponding goods do not meet the user's needs. The information of the goods that are secondarily judged to meet the user's needs is fed back to the user batch demand order table for information update, and the information of the goods that are secondarily judged not to meet the user's needs is fed back to the early warning device for early warning.

[0074] Furthermore, in this embodiment, the step of configuring the seasonal function matching degree threshold according to the seasonal change curve includes:

[0075] S501. According to the corresponding goods type in the radio frequency information of the batch goods, the sales volume data and the purchase volume data of the corresponding goods in different time periods in history are obtained, and the double-axis seasonal change curves of different types of goods and the corresponding curve functions are constructed with the time axis as the horizontal axis and the sales volume data and the purchase volume data of different time periods as the double vertical axes;

[0076] Further, the dual-axis seasonal variation curve in this embodiment is the seasonal variation curve;

[0077] S502, predicting sales volume and purchase volume according to the sales volume change curve function and the purchase volume change curve function, and optimizing the dual-axis seasonal change curve and the sales volume change curve function and the purchase volume change curve function by using the difference between the predicted value and the actual value;

[0078] S503, obtaining the sales volume and purchase volume at the next moment according to the optimized sales volume change curve function and purchase volume change curve function;

[0079] S504. According to the current sales volume and the purchase volume and the next sales volume and the purchase volume, obtain the current sales volume ratio and the next sales volume ratio. Further, the current sales volume ratio in this embodiment is the ratio of the current sales volume to the purchase volume; similarly, the next sales volume ratio is the ratio of the next sales volume to the purchase volume.

[0080] S505, setting a benchmark function matching threshold for the i-th type of goods, and constructing a genetic algorithm input sequence according to the benchmark function matching threshold, the type of goods that do not meet the matching judgment threshold, the corresponding sales ratio at the current moment and the sales ratio at the next moment, the disappointment value of the corresponding goods, and the probability of meeting user needs;

[0081] S506. Constructing a seasonal function matching threshold function and corresponding constraint conditions corresponding to the i-th type of goods according to the genetic algorithm input sequence;

[0082] Further, in this embodiment, when i Type of goods currently t The difference between the sales volume share at the moment and the sales volume share at the next moment is less than or equal to 0, indicating that the corresponding goods are in the growth sales stage and the corresponding market prospects are good. Therefore, the corresponding seasonal function matching threshold is smaller. i Type of goods currently t If the difference between the sales ratio at a certain moment and the sales ratio at the next moment is greater than 0, it means that the corresponding product sales period is in a declining sales stage, and the corresponding future market sales prospects are not good. Therefore, the corresponding seasonal function matching threshold should be larger;

[0083] Furthermore, in this embodiment, when the corresponding goods are in the growth sales stage, the corresponding seasonal function matching degree threshold is smaller, so that more types of goods can meet the seasonal function matching degree threshold of the corresponding goods, so that there is more room for selection of other goods to make up for the shortage of corresponding types of goods; on the contrary, when the corresponding goods are in the decline sales stage, the corresponding seasonal function matching degree threshold is larger, so that many non-seasonal functional replacement goods can be filtered out, reducing the selection of corresponding non-seasonal goods;

[0084] Furthermore, the matching degree threshold function and corresponding constraint conditions in this embodiment are constructed by those skilled in the art based on the above-mentioned goods sales and seasonal change process.

[0085] Furthermore, the benchmark function matching threshold in this embodiment is constructed by the average value of the historical function matching of the same goods;

[0086] S507, setting a training cycle threshold, inputting the genetic algorithm input sequence, the seasonal function matching threshold function and the corresponding constraint conditions into the genetic algorithm for training, and when the training cycle is greater than the training cycle threshold, stopping the iteration, and obtaining the seasonal function matching threshold of the corresponding goods at the corresponding time.

[0087] This process configures a unique RFID code and uses an RFID algorithm to obtain RFID information of batch goods, which can track the location and status of goods in real time and accurately, thereby improving the accuracy of inventory management. Secondly, by constructing a seasonal change curve and setting a matching judgment threshold, the matching degree of goods can be dynamically adjusted to ensure that the supply and demand matching of goods in different seasons is more accurate. Specifically, by constructing a seasonal change curve through historical data on sales and purchase volume, and using a genetic algorithm to optimize the seasonal function matching threshold, future sales trends can be effectively predicted, thereby adjusting inventory and production plans in advance to avoid inventory backlogs or shortages. In addition, by replacing the matching judgment with a secondary function, even if the initial Goods that initially have a low degree of matching can also be further improved in accuracy and flexibility through random sampling and functional attribute matching; this process not only reduces the losses caused by supply and demand mismatch, but also enhances user satisfaction with the goods; ultimately, by feeding back information about goods that meet user needs to the user's batch demand order table and issuing early warnings for goods that do not meet demand, closed-loop management of the supply chain can be achieved, improving overall operational efficiency; in summary, this process not only improves the accuracy and flexibility of goods matching, but also achieves efficient management of the supply chain and accurate satisfaction of user needs through dynamic adjustment and predictive optimization, thereby significantly enhancing the company's market competitiveness and user satisfaction.

[0088] Example 2

[0089] See also Figure 3 , another embodiment provided by the present invention: a batch cargo proofreading and early warning system for a gate-type channel machine, comprising: a demand prediction module, a radio frequency information module, a matching recognition module and an abnormality discrimination module;

[0090] The demand forecasting module is used to obtain the user batch demand order table by using the user's historical batch order data and limited demand data through the configured demand matching forecasting model;

[0091] The radio frequency information module is used to configure a unique radio frequency identification code using the user's batch demand goods information, obtain the radio frequency information of the batch goods through the configured radio frequency identification algorithm, and construct a seasonal change curve of the corresponding type of batch goods based on the radio frequency information of the batch goods;

[0092] A matching and identification module is used to match and identify the radio frequency information of batch goods; the matching and identification module includes a matching unit and a primary identification unit;

[0093] The matching unit is used to obtain the matching degree of the initial information of each batch type of goods through a matching discrimination algorithm according to the user batch demand order table and the batch goods radio frequency information; further, the matching discrimination algorithm in this embodiment is obtained by constructing a matching sub-model based on the Chinese pre-trained Bert model and a discrimination sub-model constructed by a logical discrimination algorithm;

[0094] For example, the type of goods is discriminated by matching the logical discrimination model in the discrimination algorithm, and the initial information of the corresponding type of goods is matched by the matching sub-model to obtain the matching degree of the initial information of each batch of goods;

[0095] The primary identification unit is used to verify the corresponding type of batch goods according to the matching degree of the initial information of each batch of goods and the matching identification threshold. If it is greater than the matching identification threshold, the corresponding type of batch goods passes the verification; if it is less than or equal to the matching identification threshold, a secondary function verification is performed;

[0096] An abnormality discrimination module is used to perform secondary abnormal matching discrimination and update on the corresponding type of batch goods that do not meet the matching threshold; the abnormality discrimination module includes a secondary matching unit, a secondary discrimination unit and an update unit;

[0097] A secondary matching unit is used to use the functional attribute information of the batch goods that do not meet the condition type and the functional attribute information corresponding to the random sampling information and the user batch demand order table to perform secondary function replacement matching judgment through a matching judgment algorithm to obtain the secondary function matching degree of the corresponding goods that do not meet the initial matching;

[0098] The secondary discrimination unit is used to determine whether the corresponding goods meet the user's needs according to the secondary function matching degree of the corresponding goods and the seasonal function matching degree threshold. If it is greater than the function matching degree threshold, it is determined that the corresponding goods meet the user's needs; if it is less than or equal to the function matching degree threshold, it is determined that the corresponding goods do not meet the user's needs;

[0099] The updating unit is used to feed back the information of the goods that are secondarily determined to meet the user's needs to the user's batch demand order table for information update, and feed back the information of the goods that are secondarily determined not to meet the user's needs to the early warning device for early warning.

[0100] Example 3

[0101] A computer-readable storage medium stores computer instructions, which, when executed, execute a batch cargo proofreading and early warning method for a gate-type channel machine.

[0102] The embodiments of the present invention are described above in conjunction with the accompanying drawings, but the present invention is not limited to the above-mentioned specific implementation modes, which are merely illustrative rather than restrictive. Under the guidance of the present invention, ordinary technicians in the field may also change, modify, replace and modify the above-mentioned embodiments without departing from the scope of protection of the purpose of the present invention and the claims, and all of these are within the protection of the present invention.

[0103] If the disclosed technical solution involves personal information, the product using the disclosed technical solution has clearly informed the personal information processing rules and obtained the individual's voluntary consent before processing the personal information. If the disclosed technical solution involves sensitive personal information, the product using the disclosed technical solution has obtained the individual's separate consent before processing the sensitive personal information, and at the same time meets the "explicit consent" requirement. For example, on personal information collection devices such as cameras, clear and prominent signs are set to inform that the personal information collection scope has been entered and personal information will be collected. If the individual voluntarily enters the collection scope, it is deemed that he or she agrees to the collection of his or her personal information; or on the device that processes personal information, when the personal information processing rules are notified by obvious signs / information, the individual's authorization is obtained through pop-up information or by asking the individual to upload his or her personal information; among them, the personal information processing rules may include information such as the personal information processor, the purpose of personal information processing, the processing method, and the type of personal information processed.

Claims

1. A batch cargo proofreading and early warning method for a portal channel machine, characterized in that the steps include: S1. Obtain the user's historical batch order data and limited demand data, and obtain the user's batch demand order table through the configured demand matching prediction model; S2. Obtain the user's batch demand goods information and configure a unique radio frequency identification code, obtain the radio frequency information of the batch goods through the configured radio frequency identification algorithm, and construct a seasonal change curve of the corresponding type of batch goods according to the radio frequency information of the batch goods; S3. Set a matching discrimination threshold, and obtain the matching degree of the initial information of each batch type of goods through a matching discrimination algorithm according to the user batch demand order table and the batch goods radio frequency information; S4. When the initial information matching degree of a certain batch type of goods is greater than the matching judgment threshold, the corresponding type of batch goods passes the verification. If it is less than or equal to the matching judgment threshold, the functional attribute information of the randomly sampled goods in the batch goods of the type that does not meet the matching judgment threshold and the functional attribute information corresponding to the user batch demand order table are used to perform secondary function replacement matching judgment through the matching judgment algorithm to obtain the secondary function matching degree of the corresponding goods; S5. Configure the seasonal function matching threshold according to the seasonal change curve. If the secondary function matching of the corresponding goods is greater than the seasonal function matching threshold, the corresponding goods are judged to meet the user's needs. If it is less than or equal to the seasonal function matching threshold, the corresponding goods are judged to not meet the user's needs. The information of the goods that are secondarily judged to meet the user's needs is fed back to the user batch demand order table for information update. The information of the goods that are secondarily judged not to meet the user's needs is fed back to the early warning device for early warning. The step of configuring the seasonal function matching threshold according to the seasonal change curve includes: S501. According to the corresponding goods type in the radio frequency information of the batch goods, the sales volume data and the purchase volume data of the corresponding goods in different time periods in history are obtained, and the double-axis seasonal change curves of different types of goods and the corresponding curve functions are constructed with the time axis as the horizontal axis and the sales volume data and the purchase volume data of different time periods as the double vertical axes; The step of configuring the seasonal function matching threshold according to the seasonal change curve also includes: S505, setting a benchmark function matching threshold for the i-th type of goods, and constructing a genetic algorithm input sequence according to the benchmark function matching threshold, the type of goods that do not meet the matching judgment threshold, the corresponding sales ratio at the current moment and the sales ratio at the next moment, the disappointment value of the corresponding goods, and the probability of meeting user needs; S506, constructing a seasonal function matching threshold function and corresponding constraint conditions corresponding to the i-th type of goods according to the genetic algorithm input sequence; S507, setting a training cycle threshold, inputting the genetic algorithm input sequence, the seasonal function matching threshold function and the corresponding constraint conditions into the genetic algorithm for training, and when the training cycle is greater than the training cycle threshold, stopping the iteration, and obtaining the seasonal function matching threshold of the corresponding goods at the corresponding time.

2. The batch cargo proofreading and early warning method for a portal channel machine according to claim 1, characterized in that: The steps of constructing the demand matching prediction model include: S101. Construct a limited demand information input sequence based on the user's historical batch order data and limited demand data and cargo information search space ,in, represents the type of the i-th commodity, The table represents the color, size, style and quality grade information of the i-th type of goods. represents the seasonal timestamp information corresponding to the i-th commodity, represents the keyword text information in the limited demand data corresponding to the jth user, represents the probability that the i-th product purchased by the user is a seasonal product, represents the historical purchase frequency of the i-th commodity, It represents the predicted probability that the i-th product meets the user's needs; S102: Input limited demand information into sequence and cargo information search space The model information extraction layer is input into the demand matching prediction model to obtain the limited demand fuzzy features and cargo information search feature set.

3. The batch cargo proofreading and early warning method for a portal channel machine as claimed in claim 2, characterized in that: The step of constructing the demand matching prediction model also includes: S103. According to the fuzzy characteristics of limited demand, the perceived utility matrix and disappointment preference function of user j for the i-th and k-th goods under the evaluation indicators of goods color, size, style, quality grade, probability of meeting user demand and probability of seasonal goods are constructed through disappointment theory and historical goods selection results, and the constructed perceived utility matrix and disappointment preference function are built into the demand discrimination search layer; S104, inputting the limited demand fuzzy features and the goods information search feature set into the demand discrimination search layer, using the configured disappointment preference function to perform demand goods type matching search, and obtaining the user's disappointment value corresponding to each type of goods and the corresponding search matching error probability loss between the user's demand and the goods; S105. Construct a training loss function based on the user's disappointment value corresponding to each type of goods and the search matching error loss between the user's demand and the goods, and build the training loss function into the demand matching prediction model for training. Then, through the output layer, output the probability that each type of goods meets the user's demand and the search matching accuracy of all demand types of goods based on the user's disappointment value corresponding to each type of goods.

4. The batch cargo proofreading and early warning method for a portal channel machine as claimed in claim 3, characterized in that: The step of constructing the demand matching prediction model also includes: S106, setting a search matching accuracy threshold and a user demand satisfaction rate threshold, when the probability that each type of goods meets the user's demand is greater than the user demand satisfaction rate threshold and the search matching accuracy of all demand types of goods is greater than the search matching accuracy threshold, obtaining a trained demand matching prediction model and obtaining the user batch demand goods type predicted by the search; S107. Based on the search-predicted user batch demand goods types and the corresponding goods information table, a user batch demand order table is constructed.

5. The batch cargo proofreading and early warning method for a portal channel machine as claimed in claim 4, characterized in that: The step of configuring the seasonal function matching threshold according to the seasonal change curve also includes: S502, predicting sales volume and purchase volume according to the sales volume change curve function and the purchase volume change curve function, and optimizing the dual-axis seasonal change curve and the sales volume change curve function and the purchase volume change curve function by using the difference between the predicted value and the actual value; S503, obtaining the sales volume and purchase volume at the next moment according to the optimized sales volume change curve function and purchase volume change curve function; S504. Obtain the sales volume ratio at the current moment and the sales volume ratio at the next moment according to the sales volume and the purchase volume at the current moment and the sales volume and the purchase volume at the next moment.

6. The batch cargo proofreading and early warning method for a portal channel machine according to claim 1, characterized in that: The step of obtaining the randomly sampled goods in S4 includes: According to the obtained batch of goods that do not meet the matching judgment threshold type, the corresponding batch of goods is subjected to a stratified sampling algorithm to obtain randomly sampled goods.

7. A batch cargo proofreading and warning system for a portal-type channel machine, which is used to implement the batch cargo proofreading and warning method for a portal-type channel machine as claimed in any one of claims 1 to 6, characterized in that: include: Demand prediction module, radio frequency information module, matching recognition module and abnormality identification module; The demand prediction module is used to obtain a user batch demand order table by using the user's historical batch order data and limited demand data through a configured demand matching prediction model; The radio frequency information module is used to configure a unique radio frequency identification code using the user's batch demand goods information, and obtain the radio frequency information of the batch goods through the configured radio frequency identification algorithm, and construct a seasonal change curve of the corresponding type of batch goods according to the radio frequency information of the batch goods; The matching identification module includes a matching unit and an initial discrimination unit; the matching unit is used to obtain the matching degree of the initial information of each batch type of goods through a matching discrimination algorithm according to the user batch demand order table and the batch goods radio frequency information; The primary identification unit is used to verify the corresponding type of batch goods according to the matching degree of the initial information of each batch type of goods and the matching identification threshold. If it is greater than the matching identification threshold, the corresponding type of batch goods passes the verification; if it is less than or equal to the matching identification threshold, a secondary function verification is performed.

8. The batch cargo proofreading and warning system for a portal channel machine as claimed in claim 7, characterized in that: The abnormality discrimination module includes a secondary matching unit, a secondary discrimination unit and an updating unit; The secondary matching unit is used to use the functional attribute information of the batch goods that do not meet the condition type and the functional attribute information corresponding to the random sampling information and the user batch demand order table to perform secondary function replacement matching judgment through a matching judgment algorithm to obtain the secondary function matching degree of the corresponding goods that do not meet the initial matching; The secondary discrimination unit is used to determine whether the corresponding goods meet the user's needs based on the secondary function matching degree of the corresponding goods and the seasonal function matching degree threshold. If it is greater than the function matching degree threshold, it is determined that the corresponding goods meet the user's needs; if it is less than or equal to, it is determined that the corresponding goods do not meet the user's needs.

9. A computer-readable storage medium, characterized in that: Computer instructions are stored thereon, and when the computer instructions are executed, the batch cargo proofreading and early warning method for a portal channel machine as described in any one of claims 1 to 6 is executed.

Images