Batch material transfer method, system and equipment based on workbin robot and medium

Abstract

The invention relates to a batch material transfer method, system and equipment based on a workbin robot and a medium, and the method comprises the steps: obtaining the original code scanning data of a plurality of workbin robots for a target workbin, updating the original code scanning data based on a verification screening algorithm, and obtaining the target code scanning data; acquiring transfer sensing data of a target material box in a transfer time period based on the material box robot, determining a first transfer parameter of the target material box, and determining a second transfer parameter of the target material box according to the target code scanning data; and according to the first transfer parameter and the second transfer parameter, based on an accuracy prediction neural network, an accuracy parameter of the material transfer information in the target material box is determined, and the accuracy parameter is used for representing the accuracy degree of the material transfer process in the production plan.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data processing, in particular to a batch material transfer method, system, device and medium based on a bin robot. BACKGROUND

[0002] As an automatic operation tool for transporting materials through bins, the bin robot is widely used in manufacturing, logistics and food and drug production, and can replace manual material transportation, saving time and effort, thereby improving material transportation efficiency.

[0003] In the prior art, an invention patent with publication number CN118485365B discloses a batch material transfer method, system and storage medium. In the transfer process, the material information is obtained by the identification of the code of the material bin by the scanning code device of the handling robot and the intelligent shelf. The operator only needs to place the material bin on the intelligent shelf. The system terminal autonomously selects the delivery address according to the production plan and the material condition of each work station, thereby improving the transfer efficiency.

[0004] However, in the prior art, the material information is often determined by scanning code to obtain data. The scanning code data may have data deviation, operation error or scanning device failure, thereby reducing the accuracy of material transfer management and the reliability of decision-making. Therefore, in view of the deficiencies of the prior art, the problem of insufficient management accuracy and decision-making reliability in material transfer needs to be solved. SUMMARY

[0005] The primary purpose of the present application is to solve at least one of the above problems and provide a batch material transfer method, system, device and medium based on a bin robot.

[0006] To meet the various purposes of the present application, the present application adopts the following technical solutions: A batch material transfer method based on a bin robot is provided to adapt to one of the purposes of the present application, which includes the following steps: Obtain the original scanning code data of a target bin of a plurality of bin robots, update the original scanning code data based on a verification screening algorithm, and obtain target scanning code data; Collect the transfer sensing data of the target bin within the transfer period based on the bin robot, determine the first transfer parameter of the target bin, and determine the second transfer parameter of the target bin according to the target scanning code data; According to the first transfer parameter and the second transfer parameter, determine the accuracy parameter of the material transfer information in the target bin based on an accuracy prediction neural network, and the accuracy parameter is used to represent the accuracy of the material transfer process in the production plan.

[0007] In an optional embodiment, the original scanning code data is updated based on the verification screening algorithm to obtain target scanning code data, including the following steps: Obtain the original scanning code data of the target bin scanned by the bin robot and the operation parameters corresponding to the scanning operation; For each original scanning code data, calculate the similarity between the original scanning code data and the material object data to determine the material information corresponding to the original scanning code data; According to the material information and the operation parameters corresponding to the scanning operation, the original scanning code data is screened to determine the abnormal scanning code data of the bin robot, and the secondary scanning operation instruction is sent to the bin robot for the abnormal scanning code data in the original scanning code data. Delete abnormal scanning code data to update original scanning code data and obtain target scanning code data.

[0008] In an optional embodiment, the similarity between the original scanning code data and the material object data is calculated to determine the material information corresponding to the original scanning code data, including the following steps: Based on the historical material transfer information, the object database of material transfer is pre-constructed to obtain the historical transfer time and historical transfer location of the material object; Calculate the time difference between the historical transfer time and each scanning operation time, and calculate the distance difference between the historical transfer location and each scanning operation location; According to the object database, the description text content of the material object data is obtained, the text difference between the text content and the text content of the original scanning code data is calculated, and the time difference and the distance difference are set as the weight of the text difference. Calculate the similarity threshold of the material object data; According to the similarity threshold, the original scanning code data is judged, if it meets the similarity threshold, the material object data corresponding to the scanning code data is obtained to represent the material information corresponding to the original scanning code data, if it does not meet the similarity threshold, the material object data with the highest similarity is selected as the material information corresponding to the original scanning code data. Represent.

[0009] In an optional embodiment, the original scanning code data is screened according to the material information and the operation parameters corresponding to the scanning operation, including the following steps: According to the scanning operation time, the original scanning code data is sorted to obtain the scanning time sequence data set, and the historical scanning data and the abnormal annotation data are used to train the neural network to obtain the scanning abnormality prediction neural network; The scanning time sequence data set is input into the scanning abnormality prediction neural network to obtain the abnormal parameters and the abnormal type corresponding to the scanning time sequence data set, and the abnormal type includes transfer distance abnormality, material batch abnormality, material type abnormality and operation parameter abnormality; Based on the degree of impact of the anomaly type, transfer distance anomaly, material batch anomaly, material type anomaly, and operating parameter anomaly are set as weight factors for the anomaly parameters. The weighted average of all anomaly parameters and weight factors is calculated to obtain the screening threshold. Using abnormal parameters greater than the filtering threshold as the filtering condition for the original barcode data, abnormal barcode data is obtained from the original barcode data. The abnormal operation correction model is input into the material information corresponding to the abnormal scanning data and the operation parameters corresponding to the scanning operation to obtain the secondary scanning operation instruction and send it to the bin robot. The abnormal operation correction model is obtained by training multiple training scanning data, material information data, abnormal type data and operation parameter data. The secondary scanning instruction determines the target bin corresponding to the secondary scanning operation based on the abnormal scanning data.

[0010] In an optional embodiment, the first transfer parameter of the target bin is determined based on the transfer sensor data of the target bin collected by the bin robot during the transfer period, including the following steps: Calculate the difference in transfer sensor data between two historical time points within the transfer period for the target bin. The transfer sensor data includes the transfer path, transfer distance, transfer midpoint location, and transfer target point location. Based on the transfer path, transfer distance, transfer intermediate point location, and transfer target point location, a first linear relationship is established between the difference in transfer sensor data and the displacement of the target bins. Based on the first linear relationship, the total displacement of all target bins during the transfer period is determined. Acquire multiple historical barcode scanning data collected by the bin robot during historical barcode scanning operations, filter the historical barcode scanning data within the transfer period based on the scanning time, and obtain the transfer barcode scanning data; Based on the object database of material transfer, determine the material information in the transfer barcode scanning data, and obtain multiple transfer material information corresponding to the transfer period; All material transfer information within the transfer period is deduplicated and merged according to transfer distance, material batch, and material type to obtain the total material quantity of the same batch within the transfer period. The first transfer parameter of the target material box is obtained by calculating the ratio of the total displacement of the target material box to the total amount of material in the same batch during the transfer period.

[0011] In an optional embodiment, determining the second transfer parameters of the target bin based on the target barcode scanning data includes the following steps: Calculate the difference between each target barcode scan data and other target barcode scan data, calculate the average difference of the target barcode scan data, and obtain the average difference of each target barcode scan data. The target barcode scan data includes material information and material-related production plan transfer information. The target scanning code data is filtered according to the preset difference threshold, and difference scanning code data with an average difference greater than the difference threshold is obtained. The difference between the two difference scanning code data is calculated, and the average value of the difference between the difference scanning code data is calculated to obtain the average difference of all difference scanning code data. According to the material information and the transfer information of the target scanning code data, a second linear relationship between the average difference of the difference scanning code data and the transfer parameter is set, and the second transfer parameter of the target bin is determined based on the second linear relationship.

[0012] In an optional embodiment, according to the first transfer parameter and the second transfer parameter, the accuracy parameter of the material transfer information in the target bin is determined based on the accuracy prediction neural network, including the following steps: The first transfer parameter and the second transfer parameter are regularized based on a preset regularization rule, and the regularized first transfer parameter and the second transfer parameter are input into the accuracy prediction neural network to obtain the accuracy parameter of the material transfer information in the target bin. The accuracy prediction neural network is trained by a plurality of training scanning code data, training sensing data, training transfer parameters, and material transfer information training data with accuracy labels. The accuracy parameter is also used to determine the material information corresponding to the target bin for batch material transfer in the production plan, to eliminate the target bin that does not meet the accuracy threshold and to recheck all materials of the transfer batch corresponding to the target bin, and to determine whether there is an abnormal situation in the batch material transfer, if so, to redevelop the transfer decision for the material transfer in the production plan, and to add the abnormal situation to the object database of the material transfer, and to update the threshold value corresponding to the threshold parameter.

[0013] On the other hand, a batch material transfer system based on a bin robot is provided to adapt to one of the purposes of the present application, which comprises: The data acquisition module is configured to acquire original scanning code data of the target bin by a plurality of bin robots, and update the original scanning code data based on a verification filtering algorithm to obtain target scanning code data. The data analysis module is configured to determine the first transfer parameter of the target bin based on the transfer sensing data of the target bin collected by the bin robot during the transfer period, and determine the second transfer parameter of the target bin based on the target scanning code data. The data confirmation module is configured to determine the accuracy parameter of the material transfer information in the target bin based on the first transfer parameter and the second transfer parameter and the accuracy prediction neural network, and the accuracy parameter is used to represent the accuracy of the material transfer process in the production plan.

[0014] In yet another aspect, provided for the purpose of one of the objects of the present application is a batch material transfer device based on a bin robot, comprising a central processing unit and a memory, the central processing unit being configured to invoke a computer program stored in the memory to execute the steps of the batch material transfer method based on a bin robot as described in the present application.

[0015] In yet another aspect, provided for the purpose of one of the objects of the present application is a computer readable storage medium storing computer executable instructions for causing a computer to execute the batch material transfer method based on a bin robot as described in any one of the first aspect of the present application.

[0016] The technical solutions of the present application have multiple advantages, including but not limited to the following aspects: The present application first collects original scan code data through multiple bin robots, identifies abnormal scan code data using a verification and screening algorithm, and generates secondary scan code operation instructions required for correcting abnormal scan code data and sends them to the bin robot to re-collect scan code data, thereby reducing the impact of abnormal scan code data on batch material transfer, improving the matching accuracy of material information and transfer information, and further improving the accuracy of material transfer management in production planning and the reliability of subsequent production decisions, providing accurate data support for material scheduling and management in production planning.

[0017] Secondly, the first transfer parameter and the second transfer parameter representing abnormal operation, abnormal material information or abnormal material transfer information are determined by the transfer sensing data and the target scan code data collected by the bin robot during the transfer process, and then the first transfer parameter and the second transfer parameter after regularization processing are input into the prediction neural network to obtain the accuracy parameter representing the accuracy of the material transfer process in the production plan, thereby realizing accurate evaluation of the accuracy of batch material transfer. The accuracy parameter not only can eliminate abnormal target bins for tracking and tracing in the material loading stage, but also can identify abnormal conditions in the material transfer process according to the material information and the material transfer information, improve the accuracy of the whole process evaluation of the material in the transfer state, and further improve the accuracy of batch material transfer management and the reliability of decision-making, and optimize the efficiency of material transfer tracking and material management. BRIEF DESCRIPTION OF DRAWINGS

[0018] The above and / or additional aspects and advantages of the present application will become apparent and more readily appreciated from the following description of the embodiments, taken in conjunction with the accompanying drawings, in which: Figure 1 Flowchart of an embodiment of the batch material transfer method based on a bin robot of the present application; Figure 2 Schematic diagram of the batch material transfer system based on a bin robot employed in the present application; Figure 3A schematic diagram of an illustrative bin robot of the present application; Figure 4 A schematic diagram of a framework for connecting a bin robot-based batch material transfer system to a cloud production decision platform. DETAILED DESCRIPTION

[0019] The technical solution of the present application is applicable to the field of data processing, particularly to the scenario of batch material transfer by a bin robot. In this context, the technical solution of the present application can be applied in a typical bin robot transfer scenario.

[0020] As shown in Figure 3 , a bin robot, as an automated work tool for transporting materials through bins, is widely used in manufacturing, logistics and food and drug production, and can replace manual material transportation to save time and effort, thereby improving material transportation efficiency. The bin robot 100 is provided with a plurality of bin storage positions, and the actuator of the bin robot 100 places the bin on the storage position by clamping for transportation.

[0021] To solve the problem of data deviation, operation error or scanning equipment failure in the material transfer management of the prior art, which reduces the accuracy of material transfer management and the reliability of decision-making, the technical solution provided by the embodiments of the present application has the following general idea: The original scanning data is collected by a plurality of bin robots, abnormal scanning data is identified by a verification and screening algorithm, and secondary scanning operation instructions required for correcting abnormal scanning data are generated and sent to the bin robot to re-collect scanning data, thereby reducing the influence of abnormal scanning data on batch material transfer. The transfer sensing data and target scanning data collected by the bin robot during the transfer process are used to determine the first transfer parameter and the second transfer parameter representing abnormal operation, abnormal material information or abnormal material transfer information. Then, the first transfer parameter and the second transfer parameter subjected to regularization processing are input into a prediction neural network to obtain an accuracy parameter representing the accuracy of the material transfer process in the production plan, thereby realizing accurate evaluation of the accuracy of batch material transfer.

[0022] In order to better understand the above technical solution, the above technical solution will be described in detail in combination with the drawings and specific embodiments.

[0023] The following specific embodiments can be combined with each other, and the same or similar concepts or processes will not be described again in some embodiments. The embodiments of the present application will be described in combination with the drawings.

[0024] Referring to Figure 1 , the present application discloses a batch material transfer method based on a bin robot, comprising the following steps: 110. Obtain original scan code data of the target bin by the plurality of bin robots, and update the original scan code data based on the verification screening algorithm to obtain target scan code data; The bin robot is provided with a scan code device, and the target bin is labeled with code information. The code information is usually based on the production plan to prepare the target material corresponding to the production plan in the target bin and the transfer scheme corresponding to the target material, and then the code information is set for the bin robot, the target transfer point and the manual scan code recognition code information. It can be understood that the setting of the code information can be manually binding the material and the bin, but the manual setting will also have operation errors, device precision limitations or device failures, etc. Factors that cause errors between the code information and the actual information, so the accuracy of the scan code data needs to be verified.

[0025] Optionally, the scan code data includes scan code operation data, material information and material transfer information. The scan code operation data includes but is not limited to recording scan code image, scan code angle, scan code position, scan code device information, scan code bin robot information and scan code times. The material information includes but is not limited to material type, material position, material name, material quantity, material volume, material batch, and material production link. The material transfer information includes material transfer path, material initial transfer position, material transfer position, and material transfer target position. The material transfer position can be an intelligent shelf or a temporary feeding task position. The material transfer target position can be an intelligent shelf at the transfer end or a position corresponding to the target feeding task.

[0026] As can be seen, through the above optional embodiments, the material association information in the scan code data and the content of the scan code operation are limited to comprehensively reflect the relevant features of the batch material transfer process, improve the completeness and accuracy of the material transfer data, and provide accurate data support for the subsequent verification screening algorithm. Ensure the material transfer process, especially the correct matching of material information and transfer information in batch material transfer, thereby improving the accuracy of material transfer management in the production plan and the reliability of subsequent production decisions, providing accurate data support for material scheduling and management in the production plan.

[0027] In specific implementation, updating the original scan code data based on the verification screening algorithm to obtain the target scan code data includes the following steps: Obtain the original scan code data of the target bin by the plurality of bin robots, and update the original scan code data based on the verification screening algorithm to obtain target scan code data; For each original scan code data, calculate the similarity between the original scan code data and the material object data to determine the material information corresponding to the original scan code data; The original scan code data is filtered according to the operation parameters corresponding to the material information and the scan code operation, and the abnormal scan code data of the bin robot is determined. A secondary scan code operation instruction is sent to the bin robot for the abnormal scan code data in the original scan code data, the abnormal scan code data is deleted to update the original scan code data, and target scan code data is obtained.

[0028] It can be seen that, through the above specific embodiments, based on the similarity between the original scan code data and the material object data, the material information corresponding to the original scan code data is first determined, and then the abnormal scan code data in the original scan code data is screened according to the determined material information and the operation parameters corresponding to the scan code operation, and the target bin corresponding to the abnormal scan code data is re-executed. The scan code operation, thereby reducing the problem of inaccurate scan code matching caused by factors such as scan code operation or scan code equipment failure, ensuring the accuracy of the target scan code data obtained finally, and being able to be used to provide data support for subsequent batch material transfer and management.

[0029] In specific implementation, the similarity between the original scan code data and the material object data is calculated to determine the material information corresponding to the original scan code data, including the following steps: The object database for material transfer is pre-constructed based on historical material transfer information, and the historical transfer time and the historical transfer location of the material object are obtained; The time difference between the historical transfer time and each scan code operation time is calculated, and the distance difference between the historical transfer location and each scan code operation location is calculated; The description text content of the material object data is obtained based on the object database, the text difference between the text content of the original scan code data and the text content of the original scan code data is calculated, the time difference and the distance difference are set as the weight of the text difference, and the similarity threshold of the material object data is calculated; The original scan code data is judged according to the similarity threshold. If the similarity threshold is met, the material object data corresponding to the scan code data is obtained to represent the material information corresponding to the original scan code data. If the similarity threshold is not met, the material object data with the highest similarity is selected as the material information corresponding to the original scan code data.

[0030] Optionally, the object database can be constructed in the form of a knowledge graph with materials as the main entity and the three tuple relationship of material associated information as other entities. It can also be an object-oriented database, a relational database supporting object storage, or a hybrid object database. The type of the object database is not limited in the present application.

[0031] It can be seen that through the above embodiments, the transfer process of the material in the historical period can be recorded through the object database, and the similarity threshold of the material object data is set according to the distance difference and the time difference as the weight, so that the screening of the scan code data is more in line with the actual situation of the material in the transfer process. The similarity threshold is used to screen the original scan code data, which can be used as a screening index for screening effective material information in the scan code data, reduces the material information matching failure caused by the scan code process, improves the accuracy of the scan code data, and further improves the accuracy of the obtained target scan code data. It can be used to provide data support for subsequent batch material transfer and management.

[0032] In specific implementation, the original scan code data is screened according to the operation parameters corresponding to the material information and the scan code operation, including the following steps: The original scan code data is sorted according to the scan code operation time to obtain a scan code time sequence data set, and the neural network is trained using historical scan code data and abnormal annotation data to obtain a scan code abnormality prediction neural network; The scan code time sequence data set is input into the scan code abnormality prediction neural network to obtain abnormal parameters and abnormal types corresponding to the scan code time sequence data set, the abnormal types including transfer distance abnormality, material batch abnormality, material type abnormality and operation parameter abnormality; According to the influence degree of the abnormal type, the transfer distance abnormality, the material batch abnormality, the material type abnormality and the operation parameter abnormality are set as weight factors of the abnormal parameters, and the weighted mean of all abnormal parameters and weight factors is calculated to obtain a screening threshold; The abnormal parameters greater than the screening threshold are used as the screening condition of the original scan code data, and the abnormal scan code data in the original scan code data is screened; The abnormal operation correction model is input with the material information corresponding to the abnormal scan code data and the operation parameters corresponding to the scan code operation, to obtain a secondary scan code operation instruction and send it to the bin robot. The abnormal operation correction model is trained by a plurality of training scan code data, material information data, abnormal type data and operation parameter data. The secondary scan code instruction determines the target bin corresponding to the secondary scan code operation according to the abnormal scan code data.

[0033] It can be seen that, by the above embodiment, the scan code timing data set is constructed and the scan code abnormality prediction neural network is trained, the abnormal parameters and the abnormal types in the scan code data are identified by the scan code abnormality prediction neural network, the weight of the abnormal parameters is set according to the influence degree of different abnormal types on the transfer process, and thus the screening threshold of the scan code data is determined, and the abnormal scan code data is further judged by the screening threshold, which can effectively identify the repeated scan code, the mis-scan code or the abnormal data input in the scan code process, and assist in realizing the accuracy of the scan code result. For the abnormal scan code data, the secondary scan code operation instruction is generated by the abnormal operation correction model, and the target bin is re-scanned by the bin robot according to the secondary scan code operation instruction, so as to obtain more accurate scan code data, ensure the correct matching of the material information, and further improve the accuracy of the material transfer management in the production plan and the reliability of the subsequent production decision, thereby providing accurate data support for material scheduling and management of the production plan.

[0034] 210、based on the transfer sensing data of the target bin collected by the bin robot in the transfer period, determining the first transfer parameter of the target bin, and determining the second transfer parameter of the target bin according to the target scan code data; Optionally, the transfer sensing data includes the transfer path, the transfer distance, the transfer intermediate point position and the transfer target point position, and whether the corresponding batch of materials in the production plan is accurately transferred according to the production plan decision can be determined according to the transfer sensing data. The transfer sensing data can also be cross-verified with the scan code data, thereby ensuring the accuracy of the material transfer management and the reliability of the production decision.

[0035] In specific implementation, based on the transfer sensing data of the target bin collected by the bin robot in the transfer period, determining the first transfer parameter of the target bin, includes the following steps: Calculate the difference degree of the transfer sensing data corresponding to the two historical time points of the target bin in the transfer period, the transfer sensing data including the transfer path, the transfer distance, the transfer intermediate point position and the transfer target point position; According to the first linear relationship between the transfer sensing data difference degree and the displacement amount of the target bin, the total displacement amount of all target bins in the transfer period is determined; Obtain a plurality of historical scan code data collected by the bin robot in the historical scan code operation, and filter out the historical scan code data in the transfer period according to the scan code time, to obtain the transfer scan code data; Determine the material information in the transfer scan code data based on the object database of material transfer, to obtain a plurality of transfer material information corresponding to the transfer period; The transfer material information in all transfer periods is de-duplicated and combined according to the transfer distance, the material batch and the material type, to obtain the total material amount of the same batch in the transfer period. The ratio of the total displacement amount of the target bin in the transfer period to the total material amount of the same batch is calculated to obtain a first transfer parameter of the target bin.

[0036] Optionally, the difference degree calculation of the transfer sensing data can be realized by vector distance of each sensing data or existing difference degree algorithm of data type corresponding to the sensing data, and the manner of difference degree calculation is not limited in the application.

[0037] Optionally, the first linear relationship between the difference degree of the transfer sensing data and the displacement amount of the target bin can be determined by a linear relationship model trained by historical data, or can be determined by fitting the linear relationship based on the difference degree-displacement related test data, and the manner of determining the first linear relationship is not limited in the application.

[0038] As can be seen, by the above embodiments, the linear relationship between the total displacement of the bin and the difference degree can be constructed based on the difference degree of the transfer sensing data, and then the total displacement amount of the same batch of materials is determined, and the total material amount of the same batch of materials is determined according to the time matching corresponding transfer scanning data, and the transfer situation of the same batch of materials in the transfer period is labeled by combining the total displacement amount and the total material amount, which can comprehensively consider the change of the sensing data and the material characteristics, accurately evaluate the transfer situation of the batch materials, and then reduce the material transfer information error caused by data deviation, equipment failure or operation error, improve the accuracy of batch material transfer management and the reliability of decision-making, and optimize the efficiency of material transfer tracking and material management.

[0039] In specific implementation, according to the target scanning data, a second transfer parameter of the target bin is determined, including the following steps: The difference degree between each target scanning data and other target scanning data is calculated, and the average value of the difference degree of the target scanning data is calculated to obtain the average difference degree of each target scanning data, the target scanning data including material information and transfer information of material associated production plan; The target scanning data is filtered according to a preset difference degree threshold to obtain difference scanning data with an average difference degree greater than the difference degree threshold; The difference degree between two difference scanning data is calculated, and the average value of the difference degree between the difference scanning data is calculated to obtain the average difference degree of all difference scanning data; According to the material information and the transfer information of the target scanning data, a second linear relationship between the average difference degree of the difference scanning data and the transfer parameter is set, and the second linear relationship is used to determine the second transfer parameter of the target bin.

[0040] Optionally, the second linear relationship between the average difference degree of the difference scanning code data and the transfer parameter can be determined by a linear relationship model trained by historical data, or can be determined by fitting the linear relationship based on the average difference degree-transfer parameter correlation test data. The manner of determining the second linear relationship is not limited in the present application.

[0041] As can be seen, by calculating the average difference degree between the target scanning code data containing material information and transfer information, screening out the difference scanning code data through the difference degree threshold, and determining the second transfer parameter of the target bin according to the second linear relationship between the average difference degree of the difference scanning code data and the transfer parameter, the transfer change of the material in the scanning code record can be accurately reflected, the evaluation accuracy of the material in the transfer state can be improved, and the accuracy of the batch material transfer management and the reliability of the decision-making can be improved, thereby optimizing the efficiency of the material transfer tracking and the material management.

[0042] 310. According to the first transfer parameter and the second transfer parameter, the accuracy parameter of the material transfer information in the target bin is determined based on an accuracy prediction neural network, and the accuracy parameter is used to represent the accuracy degree of the material transfer process in the production plan.

[0043] Optionally, the prediction neural network can be a CNN network, an RNN network, a GNN network, a Transformer network, or a composite network composed of multiple neural networks. The type of the prediction neural network is not limited in the present application.

[0044] In a specific implementation, according to the first transfer parameter and the second transfer parameter, the accuracy parameter of the material transfer information in the target bin is determined based on an accuracy prediction neural network, including the following steps: The first transfer parameter and the second transfer parameter are regularized based on a preset regularization rule, and the regularized first transfer parameter and the second transfer parameter are input into the accuracy prediction neural network to obtain the accuracy parameter of the material transfer information in the target bin; The accuracy prediction neural network is trained by a plurality of training scanning code data, training sensing data, training transfer parameters, and material transfer information training data with accuracy labels; The accuracy parameter is also used to determine the material information corresponding to the target bin of the batch material transfer in the production plan, to eliminate the target bin that does not meet the accuracy threshold and to recheck all materials of the transfer batch corresponding to the target bin, and to determine whether there is an abnormal situation in the material transfer information of the batch material transfer. If so, the material transfer in the production plan is re-formulated, the abnormal situation is added to the object database of the material transfer, and the threshold value corresponding to the threshold parameter is updated.

[0045] Optionally, the preset regularization rule is determined based on the fitting balance requirements of the prediction neural network for the historical first and second transport parameters input accuracy. This can be a regularization rule determined from existing model fitting experiments, or a regularization rule selected according to the type of prediction neural network. For example, when using a graph neural network as the prediction neural network, an orthogonal regularization rule can be selected to constrain the orthogonality of the first and second transport parameters, reducing redundancy and improving the stability of the prediction results. When the historical sample size of the first and second transport parameters is small, data augmentation can be used to expand data diversity and prioritize improving the generalization ability of the prediction neural network.

[0046] It should be noted that the first transfer parameter characterizes the actual material transfer situation caused by abnormal operations and abnormal material transfer information, while the second transfer parameter characterizes the actual material transfer situation caused by abnormal operations and abnormal material information. For target bins corresponding to target materials without abnormal operations, abnormal material information, or abnormal material transfer information, the prediction accuracy obtained by cross-validation of the first and second transfer parameters is relatively close to the target transfer requirements of the production plan. However, in cases where abnormal operations, abnormal material information, or abnormal material transfer exist, the prediction accuracy obtained by cross-validation of the first and second transfer parameters will differ significantly from the target transfer requirements of the production plan. Therefore, in addition to secondary scanning by the bin robot, when the accuracy parameter does not meet the accuracy threshold, manual intervention can be used to handle abnormal situations in the batch material transfer process and add the data corresponding to the abnormal situations as abnormal samples to the object database. This facilitates the identification of abnormalities in subsequent batch material transfer processes, thereby improving the accuracy of batch material transfer management and the reliability of decision-making, and optimizing the efficiency of material transfer tracking and material management.

[0047] As can be seen, through the above embodiments, regularization can be applied to the first and second transfer parameters to eliminate overfitting and underfitting problems, reduce redundancy to improve the stability of prediction results, and output the accuracy parameter of the target bin for transferring the same batch of materials through the accuracy prediction neural network. This enables accurate evaluation of the accuracy of batch material transfer. The accuracy parameter can not only eliminate abnormal target bins and trace the material loading stage, but also identify abnormal situations in the material transfer process based on material information and material transfer information, thereby improving the accuracy of the whole process evaluation of the material in the transfer state, and thus improving the accuracy of batch material transfer management and the reliability of decision-making, and optimizing the efficiency of material transfer tracking and material management.

[0048] Finally, the unique technical advantage of the present application is that the original scanning code data is collected by the plurality of bin robots, the abnormal scanning code data is identified by the verification screening algorithm, and the secondary scanning operation instruction required for correcting the abnormal scanning code data is generated and sent to the bin robot to re-collect the scanning code data, thereby reducing the influence of abnormal scanning code data on batch material transfer, improving the matching accuracy of material information and transfer information, and further improving the accuracy of material transfer management in production planning and the reliability of subsequent production decision-making, providing accurate data support for material scheduling and management in production planning; the transfer sensing data and target scanning code data collected by the bin robot during the transfer process are used to determine the first transfer parameter and the second transfer parameter representing abnormal operation, material information abnormality or material transfer information abnormality, and then the regularized first transfer parameter and the second transfer parameter are input into the prediction neural network to obtain the accuracy parameter representing the accuracy of the material transfer process in the production plan, thereby realizing accurate evaluation of the accuracy of batch material transfer. The accuracy parameter can not only eliminate abnormal target bins for tracking and tracing in the material loading stage, but also identify abnormal conditions in the material transfer process according to the material information and the material transfer information, improve the accuracy of the whole process evaluation of the material in the transfer state, and further improve the accuracy of batch material transfer management and the reliability of decision-making, optimize the efficiency of material transfer tracking and material management, and has high application prospect in material management in manufacturing, logistics and food and drug production.

[0049] Please refer to Figure 2 According to one aspect of the present application, a batch material transfer system based on a bin robot is provided, the system comprising: A data acquisition module is configured to acquire original scanning code data of a target bin collected by a plurality of bin robots, and update the original scanning code data based on a verification screening algorithm to obtain target scanning code data. A data analysis module is configured to collect transfer sensing data of the target bin within a transfer period based on the bin robot, determine a first transfer parameter of the target bin, and determine a second transfer parameter of the target bin according to the target scanning code data. A data confirmation module is configured to determine an accuracy parameter of material transfer information in the target bin based on an accuracy prediction neural network according to the first transfer parameter and the second transfer parameter, wherein the accuracy parameter is used to represent the accuracy of the material transfer process in the production plan.

[0050] On the basis of any embodiment of the system of the application, the system of the application further comprises: a verification screening module configured to obtain original scanning data of the target bin by the bin robot and operation parameters corresponding to the scanning operation; for each original scanning data, calculate the similarity between the original scanning data and the material object data to determine the material information corresponding to the original scanning data; screen the original scanning data according to the material information and the operation parameters corresponding to the scanning operation to determine the abnormal scanning data of the bin robot, send a secondary scanning operation instruction to the bin robot for the abnormal scanning data in the original scanning data, delete the abnormal scanning data to update the original scanning data, and obtain target scanning data.

[0051] On the basis of any embodiment of the system of the application, the system of the application further comprises: a material information confirmation module configured to pre-construct an object database of material transfer based on historical material transfer information, obtain historical transfer time and historical transfer location of the material object; calculate the time difference between the historical transfer time and each scanning operation time, and calculate the distance difference between the historical transfer location and each scanning operation location; obtain the description text content of the material object data from the object database, calculate the text difference between the text content and the text content of the original scanning data, set the time difference and the distance difference as the weight of the text difference to calculate the similarity threshold of the material object data; judge the original scanning data according to the similarity threshold, if the similarity threshold is met, obtain the material object data corresponding to the scanning data to represent the material information corresponding to the original scanning data, if the similarity threshold is not met, select the material object data with the highest similarity as the material information corresponding to the original scanning data.

[0052] On the basis of any embodiment of the system of the application, the system of the application further comprises: a secondary scanning module configured to sort the original scanning data according to the scanning operation time to obtain a scanning time sequence data set, train a neural network using historical scanning data and abnormal annotation data to obtain a scanning abnormality prediction neural network; input the scanning time sequence data set into the scanning abnormality prediction neural network to obtain abnormal parameters and abnormal types corresponding to the scanning time sequence data set, the abnormal types including transfer distance abnormality, material batch abnormality, material type abnormality and operation parameter abnormality; set the transfer distance abnormality, the material batch abnormality, the material type abnormality and the operation parameter abnormality as weight factors of the abnormal parameters according to the influence degree of the abnormal types, calculate the weighted mean of all abnormal parameters and weight factors to obtain a screening threshold; set the abnormal parameters greater than the screening threshold as the screening condition of the original scanning data, and screen the abnormal scanning data in the original scanning data. The abnormal operation correction model is trained by a plurality of training scan code data, material information data, abnormal type data and operation parameter data, and the secondary scan code instruction is determined according to the abnormal scan code data.

[0053] On the basis of any embodiment of the system of the application, the system of the application further comprises: a first transfer module configured to calculate a difference degree of transfer sensing data corresponding to two historical time points of a target bin in a transfer period, the transfer sensing data including a transfer path, a transfer distance, a transfer intermediate point position and a transfer target point position; set a first linear relationship between the difference degree of the transfer sensing data and a displacement amount of the target bin according to the transfer path, the transfer distance, the transfer intermediate point position and the transfer target point position, determine a total displacement amount of all target bins in the transfer period according to the first linear relationship; obtain a plurality of historical scan code data collected by the bin robot in the historical scan code operation, filter out the historical scan code data in the transfer period according to the scan code time, and obtain transfer scan code data; determine the material information in the transfer scan code data based on an object database of material transfer, and obtain a plurality of transfer material information corresponding to the transfer period; combine and calculate the transfer material information in all transfer periods according to the transfer distance, the material batch and the material type to obtain a total material amount of the same batch in the transfer period; calculate a ratio of the total displacement amount of the target bin in the transfer period to the total material amount of the same batch to obtain a first transfer parameter of the target bin.

[0054] On the basis of any embodiment of the system of the application, the system of the application further comprises: a second transfer module configured to calculate a difference degree between each target scan code data and other target scan code data, calculate an average value of the difference degree of the target scan code data to obtain an average difference degree of each target scan code data, the target scan code data including material information and transfer information of the material associated production plan; filter the target scan code data according to a preset difference degree threshold to obtain difference scan code data with an average difference degree greater than the difference degree threshold; calculate a difference degree between two difference scan code data, calculate an average value of the difference degree between the difference scan code data to obtain an average difference degree of all difference scan code data; set a second linear relationship between the average difference degree of the difference scan code data and a transfer parameter according to the material information and the transfer information of the target scan code data, and determine a second transfer parameter of the target bin based on the second linear relationship.

[0055] On the basis of any embodiment of the system of the application, the system of the application further comprises: an accuracy prediction module configured to perform regularization processing on the first transfer parameter and the second transfer parameter based on a preset regularization rule, input the first transfer parameter and the second transfer parameter after the regularization processing into the accuracy prediction neural network, and obtain the accuracy parameter of the material transfer information in the target bin; the accuracy prediction neural network is obtained by training a plurality of training code scanning data, training sensing data, training transfer parameters, and material transfer information training data with an accuracy label; the accuracy parameter is further used to determine the material information corresponding to the target bin in the batch material transfer in the production plan, eliminate the target bin that does not meet the accuracy threshold, and recheck all materials of the transfer batch corresponding to the target bin, and determine whether there is an abnormal situation in the batch material transfer information, if yes, reestablish the transfer decision of the material transfer in the production plan, and add the abnormal situation to the object database of the material transfer, and update the threshold value corresponding to the threshold parameter.

[0056] Another embodiment of the application also provides a batch material transfer equipment based on a bin robot, which comprises a processor, a computer readable storage medium, a memory and a network interface connected through a system bus. Wherein the computer readable non-volatile storage medium of the batch material transfer equipment based on the bin robot stores an operating system, a database and computer readable instructions, the database can store information sequences, and the computer readable instructions can make the processor realize a batch material transfer method based on a bin robot when executed by the processor.

[0057] As shown in Figure 4 The batch material transfer system 10 based on the bin robot is connected to the cloud production decision platform, the batch material transfer system 10 comprises a data acquisition module 20, a data analysis model 30 and a data confirmation module 40, the data acquisition module 10 acquires data of the transfer process from a plurality of bin robots 100, the data analysis module 20 calculates and analyzes the data of the transfer process through the cloud computing server 50 connected by the batch material transfer system 10, and stores the data of the calculation process and the analysis result data in the object database 60 connected by the batch material transfer system 10, wherein the cloud production decision platform 70 issues a control instruction of material transfer to the bin robot 100 according to a production plan, the bin robot 100 performs material transfer work according to the instruction, the batch material transfer system 10 calculates and analyzes the data uploaded by the bin robot 100 in real time, evaluates the state of the batch material transfer through the output accuracy parameter, and sends the state to the cloud production decision platform 70 to provide reference data of the real-time transfer state, and provides data support for production decision.

[0058] The processor of the bin-based robotic bulk material transfer apparatus is configured to provide computing and control capabilities to support operation of the bin-based robotic bulk material transfer apparatus. The memory of the bin-based robotic bulk material transfer apparatus can store computer readable instructions that, when executed by the processor, cause the processor to perform the bin-based robotic bulk material transfer method of the present application. The network interface of the bin-based robotic bulk material transfer apparatus is configured to enable communication with a terminal.

[0059] The processor in the embodiment is configured to perform the specific functions of each module in Figure 2 The memory stores the program codes and various data required to execute the above-mentioned modules or sub-modules. The network interface is configured to enable data transmission between the terminal and the server.

[0060] The non-volatile readable storage medium in the embodiment stores the program codes and data required to execute all modules in the bin-based robotic bulk material transfer system of the present application. The server can call the program codes and data of the server to execute the functions of all modules.

[0061] The present application also provides a non-volatile readable storage medium storing computer readable instructions that, when executed by one or more processors, cause the one or more processors to perform the steps of the bin-based robotic bulk material transfer method of any embodiment of the present application.

[0062] The present application also provides a computer program product comprising computer programs / instructions that, when executed by one or more processors, implement the steps of the method of any embodiment of the present application.

Claims

1. A method for batch material transfer based on a bin robot, characterized in that, Includes the following steps: The system acquires the original barcode scanning data of multiple bin robots on the target bin, updates the original barcode scanning data based on a verification and filtering algorithm, and obtains the target barcode scanning data. Based on the transfer sensor data of the target bin collected by the bin robot during the transfer period, the first transfer parameter of the target bin is determined, and the second transfer parameter of the target bin is determined based on the target barcode scanning data. Based on the first transfer parameter and the second transfer parameter, and using an accuracy prediction neural network, an accuracy parameter for the material transfer information in the target bin is determined. This accuracy parameter is used to characterize the accuracy of the material transfer process in the production plan.

2. The batch material transfer method based on a bin robot according to claim 1, characterized in that, The original QR code data is updated based on the verification and filtering algorithm to obtain the target QR code data, including the following steps: Obtain the original barcode scanning data and corresponding operation parameters of the barcode scanning operation performed by the barcode robot on the target barcode; For each original barcode scan data, calculate the similarity between the original barcode scan data and the material object data to determine the material information corresponding to the original barcode scan data; The original barcode scanning data is filtered based on the material information and the corresponding operation parameters of the barcode scanning operation. Abnormal barcode scanning data of the bin robot is identified. A second barcode scanning operation instruction is sent to the bin robot for the abnormal barcode scanning data in the original barcode scanning data. The abnormal barcode scanning data is deleted to update the original barcode scanning data and obtain the target barcode scanning data.

3. The batch material transfer method based on a bin robot according to claim 2, characterized in that, Calculating the similarity between the original barcode scan data and the material object data to determine the material information corresponding to the original barcode scan data includes the following steps: A database of material transfer objects is pre-built based on historical material transfer information to obtain the historical transfer time and historical transfer location of material objects. Calculate the time difference between historical transit time and the time of each scanning operation, and calculate the distance difference between historical transit location and each scanning operation location; Based on the descriptive text content of the material object data obtained from the object database, the text difference between the text content and the text content of the original barcode data is calculated. The time difference and distance difference are set as the weights of the text difference to calculate the similarity threshold of the material object data. The original barcode scan data is judged based on a similarity threshold. If the similarity threshold is met, the material object data corresponding to the scan data is obtained to represent the material information corresponding to the original barcode scan data. If the similarity threshold is not met, the material object data with the highest similarity is selected as the material information representation corresponding to the original barcode scan data.

4. The batch material transfer method based on a bin robot according to claim 3, characterized in that, The raw barcode scanning data is filtered based on the material information and the corresponding operation parameters, including the following steps: The original scanning data is sorted according to the scanning operation time to obtain a scanning time series dataset. The neural network is trained using historical scanning data and anomaly labeled data to obtain a scanning anomaly prediction neural network. Input the barcode scanning time series dataset into the barcode scanning anomaly prediction neural network to obtain the anomaly parameters and anomaly types corresponding to the barcode scanning time series dataset. The anomaly types include abnormal transfer distance, abnormal material batch, abnormal material type, and abnormal operation parameters. Based on the degree of impact of the anomaly type, transfer distance anomaly, material batch anomaly, material type anomaly, and operating parameter anomaly are set as weight factors for the anomaly parameters. The weighted average of all anomaly parameters and weight factors is calculated to obtain the screening threshold. Using abnormal parameters greater than the filtering threshold as the filtering condition for the original barcode data, abnormal barcode data is obtained from the original barcode data. The abnormal operation correction model is input into the material information corresponding to the abnormal scanning data and the operation parameters corresponding to the scanning operation to obtain the secondary scanning operation instruction and send it to the bin robot. The abnormal operation correction model is obtained by training multiple training scanning data, material information data, abnormal type data and operation parameter data. The secondary scanning instruction determines the target bin corresponding to the secondary scanning operation based on the abnormal scanning data.

5. The batch material transfer method based on a bin robot according to claim 1, characterized in that, Based on the transfer sensor data of the target bin collected by the bin robot during the transfer period, the first transfer parameter of the target bin is determined, including the following steps: Calculate the difference in transfer sensor data between two historical time points within the transfer period for the target bin. The transfer sensor data includes the transfer path, transfer distance, transfer midpoint location, and transfer target point location. Based on the transfer path, transfer distance, transfer midpoint location, and transfer target point location, establish a first linear relationship between the difference in transfer sensor data and the displacement of the target bins, and determine the total displacement of all target bins during the transfer period based on the first linear relationship. Acquire multiple historical barcode scanning data collected by the bin robot during historical barcode scanning operations, filter the historical barcode scanning data within the transfer period based on the scanning time, and obtain the transfer barcode scanning data; Based on the object database of material transfer, determine the material information in the transfer barcode scanning data, and obtain multiple transfer material information corresponding to the transfer period; All material transfer information within the transfer period is deduplicated and merged according to transfer distance, material batch, and material type to obtain the total material quantity of the same batch within the transfer period. The first transfer parameter of the target material box is obtained by calculating the ratio of the total displacement of the target material box to the total amount of material in the same batch during the transfer period.

6. The batch material transfer method based on a bin robot according to claim 1, characterized in that, Based on the target barcode scanning data, the second transfer parameters of the target bin are determined, including the following steps: Calculate the difference between each target barcode scan data and other target barcode scan data, calculate the average difference of the target barcode scan data, and obtain the average difference of each target barcode scan data. The target barcode scan data includes material information and material-related production plan transfer information. The target QR code data is filtered according to the preset difference threshold to obtain QR code data with an average difference greater than the difference threshold. Calculate the difference between two different scanned data, and average the difference between the different scanned data to obtain the average difference of all different scanned data. Based on the material information and transfer information of the target barcode data, a second linear relationship is set between the average difference of the differential barcode data and the transfer parameters, and the second transfer parameters of the target bin are determined based on the second linear relationship.

7. The batch material transfer method based on a bin robot according to claim 1, characterized in that, Based on the first and second transfer parameters, and using an accuracy prediction neural network, the accuracy parameters of the material transfer information in the target bin are determined, including the following steps: The first and second transfer parameters are regularized according to a preset regularization rule. The regularized first and second transfer parameters are then input into the accuracy prediction neural network to obtain the accuracy parameters of the material transfer information in the target bin. The accuracy prediction neural network is obtained through training multiple training barcode scanning data, training sensor data, training transfer parameters, and material transfer information training data with accuracy labels. The accuracy parameter is also used in the production plan to determine the material information corresponding to the target bins for batch material transfer, remove target bins that do not meet the accuracy threshold and recheck all materials in the transfer batch corresponding to the target bins, and determine whether there are any abnormalities in the material transfer information in the batch material transfer. If so, the transfer decision in the production plan is re-formulated, the abnormality is added to the material transfer object database, and the threshold parameter corresponding to the threshold judgment is updated.

8. A batch material transfer system based on a bin robot, characterized in that, The system is used to execute the batch material transfer method based on a bin robot as described in any one of claims 1-7, the system comprising: The data acquisition module is used to acquire the original barcode scanning data of multiple bin robots on the target bin, update the original barcode scanning data based on the verification and filtering algorithm, and obtain the target barcode scanning data. The data analysis module is used to determine the first transfer parameters of the target bin based on the transfer sensor data collected by the bin robot during the transfer period, and to determine the second transfer parameters of the target bin based on the target barcode data. The data confirmation module is used to determine the accuracy parameters of the material transfer information in the target bin based on the first transfer parameters and the second transfer parameters and an accuracy prediction neural network. The accuracy parameters are used to characterize the accuracy of the material transfer process in the production plan.

9. A batch material transfer device based on a bin robot, comprising: At least one processor, and, A memory communicatively connected to the at least one processor; characterized in that the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to perform the batch material transfer method based on a bin robot as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores, in the form of computer-readable instructions, a computer program implemented according to any one of claims 1 to 7, which, when invoked by a computer, executes the steps included in the corresponding method.

Images