ERP management system based on low-code platform fusion data real-time analysis

Abstract

The invention relates to an ERP (Enterprise Resource Planning) management system based on low-code platform fusion data real-time analysis, and relates to the technical field of ERP management systems.A time queue of a real-time inventory level is obtained, and market demand historical data is obtained based on timestamps at the beginning and the end of the time queue of the real-time inventory level; performing time sequence feature extraction and implicit association analysis on the real-time inventory level and the market demand historical data by adopting a data analysis and processing technology based on artificial intelligence, so as to perform adaptive query matching based on a time sequence with the real-time inventory level by taking the market demand historical data as query features; according to the technical scheme, the inventory management result indicating whether replenishment reminding is generated or not is intelligently generated in response to the matching result, the inventory level data can be acquired in real time, and the change mode of the inventory is described in combination with the time sequence dynamic change and trend of the inventory, so that the current inventory state can be timely and accurately mastered, and decision delay caused by data lag is reduced.

Description

Technical Field

[0001] The present application relates to the technical field of ERP management systems, and specifically, to an ERP management system based on a low-code platform for real-time analysis of fused data. Background Art

[0002] The ERP management system is an integrated information system used to help enterprises manage and coordinate various resources within the enterprise. It provides a unified platform by integrating various business processes and functions within the enterprise, enabling enterprises to plan, organize, execute and control their business activities more effectively.

[0003] As a core component of the ERP system, the efficiency and accuracy of inventory management directly affect the market competitiveness of enterprises. However, traditional inventory data is not updated in a timely manner, which will cause enterprises to react slowly to market dynamics and find it difficult to capture changes in market demand in a timely manner, so that production and replenishment plans cannot be adjusted quickly, resulting in inventory backlogs or out-of-stocks. In addition, many inventory management processes still rely on manual records, which have a low degree of automation, not only increasing the possibility of errors, but also consuming a lot of human resources.

[0004] Therefore, an optimized ERP management system is needed to solve the above technical problems. Summary of the invention

[0005] This summary is provided to introduce concepts in a brief form that will be described in detail in the detailed description below. This summary is not intended to identify key features or essential features of the claimed technical solution, nor is it intended to limit the scope of the claimed technical solution.

[0006] This application provides an ERP management system based on a low-code platform for real-time data analysis, the system comprising:

[0007] Real-time inventory level acquisition module, used to obtain the time queue of real-time inventory level;

[0008] A real-time inventory level timestamp extraction module, configured to extract a start timestamp and an end timestamp from a time queue of the real-time inventory level;

[0009] A market demand historical data interception module, used to intercept the market demand historical data based on the start timestamp and the end timestamp;

[0010] A market demand historical data time series feature extraction module, used for inputting the market demand historical data into a market demand time series feature extractor to obtain a market demand historical time series feature vector;

[0011] A real-time inventory level local time series association module, used for inputting the time queue of the real-time inventory level into the inventory level time series feature extractor to obtain a sequence of real-time inventory local time series implicit association feature vectors;

[0012] A real-time inventory-market demand adaptive query module is used to use the market demand historical time series feature vector as a query feature vector, and input the sequence of the market demand historical time series feature vector and the real-time inventory local time series implicit correlation feature vector into a fine-grained feature decoupling adaptive query encoding module to obtain a real-time inventory-market demand query response adaptive matching result representation matrix as a real-time inventory-market demand query response adaptive matching result representation feature;

[0013] The replenishment reminder judgment module is used to obtain an inventory management result based on the adaptive matching result representation characteristics of the real-time inventory-market demand query response, and the inventory management result is used to indicate whether a replenishment reminder is generated.

[0014] Optionally, the market demand historical data time series feature extraction module is used to: input the market demand historical data into a market demand time series feature extractor based on a 1D-CNN model to obtain the market demand historical time series feature vector.

[0015] Optionally, the real-time inventory level local time series association module is used to: input the time queue of the real-time inventory level into an inventory level time series feature extractor based on the RNN model to obtain a sequence of the real-time inventory local time series implicit association feature vectors.

[0016] Optionally, the real-time inventory-market demand adaptive query module includes: a fine-grained decoupling unit, which is used to perform fine-grained decoupling on the sequence of the real-time inventory local time series implicit association feature vectors to obtain a set of subsequences of the real-time inventory local time series implicit association feature vectors; a real-time inventory time series implicit association fine-grained local association encoding unit, which is used to input each subsequence of the real-time inventory local time series implicit association feature vector in the set of subsequences of the real-time inventory local time series implicit association feature vectors into a subsequence semantic enhancement encoder to obtain a set of real-time inventory time series implicit fine-grained decoupled local association feature vectors; and a covariance matrix calculation unit, which is used to calculate the covariance matrix between the query feature vector and each real-time inventory time series implicit fine-grained decoupled local association feature vector in the set of the real-time inventory time series implicit fine-grained decoupled local association feature vectors to obtain a query feature vector. a set of query real-time inventory time series response covariance matrices; a query real-time inventory time series response trace metric factor calculation unit, used to calculate the matrix trace of each query real-time inventory time series response covariance matrix in the set of query real-time inventory time series response covariance matrices to obtain a set of query real-time inventory time series response trace metric factors; a query real-time inventory time series response trace metric factor activation unit, used to activate the set of query real-time inventory time series response trace metric factors to obtain a set of query real-time inventory time series response trace attention weights; a weighted sum calculation unit, used to use each query real-time inventory time series response trace attention weight in the set of query real-time inventory time series response trace attention weights as a weight, and calculate the weighted sum of the set of query real-time inventory time series response covariance matrices to obtain the real-time inventory-market demand query response adaptive matching result representation matrix.

[0017] Optionally, the real-time inventory time series implicit association fine-grained local association encoding unit is used to: input each subsequence of the real-time inventory local time series implicit association feature vector in the set of subsequences of the real-time inventory local time series implicit association feature vectors into a subsequence semantic reinforcement encoder based on a Bi-LSTM model to obtain a set of the real-time inventory time series implicit fine-grained decoupled local association feature vectors.

[0018] Optionally, the query real-time inventory time series response trace metric factor calculation unit is used to: respectively calculate the sum of each element on the diagonal of each query real-time inventory time series response covariance matrix in the set of query real-time inventory time series response covariance matrices to obtain the set of query real-time inventory time series response trace metric factors.

[0019] Optionally, the query real-time inventory time series response trace measurement factor activation unit is used to: input the set of query real-time inventory time series response trace measurement factors into a Sigmoid activation function to obtain a set of query real-time inventory time series response trace attention weights.

[0020] Optionally, the replenishment reminder judgment module is used to: input the real-time inventory-market demand query response adaptive matching result representation matrix into a classifier-based inventory management result generator to obtain the inventory management result, and the inventory management result is used to indicate whether a replenishment reminder is generated.

[0021] Optionally, the replenishment reminder judgment module includes: a matrix expansion unit, used to expand the real-time inventory-market demand query response adaptive matching result representation matrix into a classification feature vector according to a row vector or a column vector; a fully connected encoding unit, used to use multiple fully connected layers of the classifier-based inventory management result generator to fully connect encode the classification feature vector to obtain an encoded classification feature vector; a classification unit, used to pass the encoded classification feature vector through the Softmax classification function of the classifier-based inventory management result generator to obtain the inventory management result.

[0022] The present application has at least the following technical effects: Compared with the prior art, the present application adopts the above technical solution, obtains the time queue of the real-time inventory level, and obtains the market demand history data based on the timestamps of the start and end of the time queue of the real-time inventory level, and adopts artificial intelligence-based data analysis and processing technology to extract the time series features and implicit association analysis of the real-time inventory level and the market demand history data, so as to perform adaptive query matching based on time series with the real-time inventory level according to the market demand history data as the query feature, so as to obtain the response matching result representation, thereby intelligently generating the inventory management result of whether to generate a replenishment reminder. The present application can obtain inventory level data in real time, and characterize the inventory change pattern in combination with the inventory's time series dynamic changes and trends, so as to timely and accurately grasp the current inventory status and reduce decision delays caused by data lags.

[0023] Other features and advantages of the present application will be described in detail in the subsequent detailed description. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] The above and other features, advantages and aspects of the embodiments of the present application will become more apparent with reference to the following detailed description in conjunction with the accompanying drawings. Throughout the drawings, the same or similar reference numerals represent the same or similar elements. It should be understood that the drawings are schematic and the originals and elements are not necessarily drawn to scale. In the drawings:

[0025] Figure 1 It is a block diagram of an ERP management system for real-time analysis of fused data based on a low-code platform according to an exemplary embodiment.

[0026] Figure 2It is a flowchart of an ERP management method based on a low-code platform for real-time analysis of fused data according to an exemplary embodiment. DETAILED DESCRIPTION

[0027] The embodiments of the present application will be described in more detail below with reference to the accompanying drawings. Although certain embodiments of the present application are shown in the accompanying drawings, it should be understood that the present application can be implemented in various forms and should not be construed as being limited to the embodiments described herein. Instead, these embodiments are provided to provide a more thorough and complete understanding of the present application. It should be understood that the drawings and embodiments of the present application are only for exemplary purposes and are not intended to limit the scope of protection of the present application.

[0028] It should be understood that the various steps described in the method implementation of the present application can be performed in different orders and / or performed in parallel. In addition, the method implementation may include additional steps and / or omit the steps shown. The scope of the present application is not limited in this respect.

[0029] The term "including" and its variations used herein are open inclusions, i.e., "including but not limited to". The term "based on" means "based at least in part on". The term "one embodiment" means "at least one embodiment"; the term "another embodiment" means "at least one additional embodiment"; the term "some embodiments" means "at least some embodiments". The relevant definitions of other terms will be given in the following description.

[0030] It should be noted that the concepts such as "first" and "second" mentioned in this application are only used to distinguish different devices, modules or units, and are not used to limit the order or interdependence of the functions performed by these devices, modules or units.

[0031] It should be noted that the modifications of "one" and "plurality" mentioned in the present application are illustrative rather than restrictive, and those skilled in the art should understand that unless otherwise clearly indicated in the context, it should be understood as "one or more".

[0032] The names of the messages or information exchanged between multiple devices in the embodiments of the present application are only used for illustrative purposes and are not used to limit the scope of these messages or information.

[0033] The specific implementation of the present application is described in detail below with reference to the accompanying drawings.

[0034] Figure 1 is a block diagram of an ERP management system based on low-code platform fusion data real-time analysis according to an exemplary embodiment. Figure 1 As shown, the ERP management system 100 includes:

[0035] A real-time inventory level acquisition module 101 is used to acquire a time queue of real-time inventory levels;

[0036] A real-time inventory level timestamp extraction module 102 is used to extract a start timestamp and an end timestamp from a time queue of the real-time inventory level;

[0037] A market demand historical data interception module 103, used to intercept the market demand historical data based on the start timestamp and the end timestamp;

[0038] A market demand historical data time series feature extraction module 104, used for inputting the market demand historical data into a market demand time series feature extractor to obtain a market demand historical time series feature vector;

[0039] A real-time inventory level local time series association module 105 is used to input the time queue of the real-time inventory level into the inventory level time series feature extractor to obtain a sequence of real-time inventory local time series implicit association feature vectors;

[0040] The real-time inventory-market demand adaptive query module 106 is used to use the market demand historical time series feature vector as the query feature vector, and input the sequence of the market demand historical time series feature vector and the real-time inventory local time series implicit correlation feature vector into the fine-grained feature decoupling adaptive query encoding module to obtain the real-time inventory-market demand query response adaptive matching result representation matrix as the real-time inventory-market demand query response adaptive matching result representation feature;

[0041] The replenishment reminder judgment module 107 is used to obtain an inventory management result based on the representation characteristics of the real-time inventory-market demand query response adaptive matching result, and the inventory management result is used to indicate whether to generate a replenishment reminder.

[0042] Specifically, the Low-Code Development Platform (LCDP) is a digital technology tool platform for two-way drive management for business personnel and IT personnel. It uses efficient methods such as graphical drag and drop and parameterized configuration to achieve functions such as rapid construction, data orchestration, ecological connection, and middle-office services. The low-code platform has the following core functions:

[0043] Scenario building capability: quickly build application scenarios required for operation management through graphical visualization and drag-and-drop, and agilely respond to rapid iterations of demand changes, so that research is development and development is deployment.

[0044] ‌Data orchestration capability‌: Through visual business rule orchestration, the scattered data within the enterprise can be revitalized to provide data services and business services for various front-end business scenarios;

[0045] ‌Connecting Ecosystem Capabilities‌: The platform’s integration capabilities can easily connect organizations and systems upstream and downstream of an enterprise, expand the enterprise’s business chain service boundaries, accumulate more data assets, and feed back data to the business, thus achieving more refined business scenario operations;

[0046] ‌Business middle platform capabilities‌: Rapidly build various innovative micro-applications required by the service center in a digital form, and realize online business in the core operation scenarios of the enterprise;

[0047] ‌Openness and scalability‌: The low-code platform not only supports personalized extensions through a small amount of code, but also supports plug-in extensions and social plug-in references, etc., to meet the various personalized scenario requirements of enterprise planning and design;

[0048] ‌Efficient Development‌: Generate applications quickly without coding or with a small amount of code, build business processes, logic and data models, and deliver applications for updates with one click;

[0049] As low-code platforms continue to develop, their applications in inventory management are becoming more extensive and in-depth.

[0050] Specifically, the EPR management system is a management platform that is based on information technology and provides decision-making and operation means for enterprise decision-makers and employees with a systematic management philosophy. It is characterized by integration, advancement, unity, integrity, and openness. It can integrate all resources together to plan procurement, production, cost, inventory, distribution, transportation, finance, and human resources, thereby achieving the best resource combination and obtaining the best benefits.

[0051] Correspondingly, the functions of the EPR management system in the finance / accounting department are: real-time and complete account information, timely collection of accounts and customer credit limit control, in compliance with the business tax law and taking into account the internal audit and internal control standards;

[0052] The functions of the EPR management system in procurement / business: meet the basic procurement work of the enterprise, effectively control the order processing status, record the change data in real time, accurately grasp the progress, and meet the delivery date required by customers;

[0053] The functions of the EPR management system in production and manufacturing: to accurately grasp the production progress, material preparation plan, process and outsourcing management, calculate the production cost in real time, shorten the manufacturing cycle, improve the capacity utilization rate and reduce the manufacturing cost;

[0054] The functions of the EPR management system for mid- and senior-level managers are: real-time information integration to effectively grasp the company's operating status, extend the ERP vision, and enhance corporate competitiveness in combination with business philosophy.

[0055] Based on this, the present application proposes an ERP management system based on real-time analysis of data fusion based on a low-code platform, which obtains the time queue of the real-time inventory level, and obtains the market demand historical data based on the timestamps of the start and end of the time queue of the real-time inventory level, and uses artificial intelligence-based data analysis and processing technology to extract the time series features and implicit association analysis of the real-time inventory level and the market demand historical data, so as to use the market demand historical data as a query feature to perform adaptive query matching based on the time series with the real-time inventory level, so as to obtain a response matching result representation, thereby intelligently generating an inventory management result of whether to generate a replenishment reminder. The present application can obtain inventory level data in real time, and characterize the inventory change pattern in combination with the time series dynamic changes and trends of the inventory, so as to be able to grasp the current inventory status in a timely and accurate manner, and reduce decision delays caused by data lag. At the same time, combined with the time series association analysis of the historical data of market demand, it can quickly identify market trends and demand changes, reduce dependence on manual analysis, and reduce errors and delays caused by human factors, thereby improving the intelligence level of the inventory management process, so as to provide enterprises with accurate inventory strategy adjustment suggestions.

[0056] Specifically, in the technical solution of the present application, first, a time queue of real-time inventory levels is obtained. Then, in order to better understand the dynamic trend of inventory changes and identify the patterns and laws of inventory fluctuations, in the technical solution of the present application, a start timestamp and an end timestamp are extracted from the time queue of the real-time inventory levels. That is, the timestamps provide time series data of changes in inventory levels, which helps to understand the dynamic changes in inventory levels over time, so that subsequent analysis can focus on the trends of inventory changes and market demand in a specific time period, thereby improving the accuracy of the analysis.

[0057] Specifically, in order to match the historical market demand data with the real-time inventory level, so as to help predict future inventory demand more accurately. Therefore, in the technical solution of the present application, the historical market demand data is intercepted based on the start timestamp and the end timestamp. That is, the start and end timestamps set a clear time range for data analysis. By selecting the inventory level data of a specific time period, the intercepted market demand data can be directly related to the real-time inventory level, so as to pay attention to the latest market demand changes, thereby improving the response speed of inventory management.

[0058] Specifically, considering that the market demand historical data is constantly changing over time, that is, the market data has certain time series feature information on a local time scale. The 1D-CNN model can effectively extract important local time series features from the time series data. Therefore, in the technical solution of the present application, the market demand historical data is input into a market demand time series feature extractor based on the 1D-CNN model to obtain a market demand historical time series feature vector, which can help identify the time series dynamic change trend of market demand. In particular, the essence of the market demand time series feature extractor based on the 1D-CNN model is a one-dimensional convolutional neural network model.

[0059] In one embodiment of the present application, the market demand historical data time series feature extraction module is used to: input the market demand historical data into a market demand time series feature extractor based on a 1D-CNN model to obtain the market demand historical time series feature vector.

[0060] Specifically, considering that the real-time inventory level in the time queue of the real-time inventory level is constantly changing with the change of time, that is, the real-time inventory level has dynamic changes in the time series, and the RNN model is used to process sequence data, which can capture the dynamic changes and time series dependencies in the time series. Based on this, in the technical solution of the present application, the time queue of the real-time inventory level is input into the inventory level time series feature extractor based on the RNN model to capture and mine the implicit correlation features of the real-time inventory level in the local time series, thereby obtaining a sequence of implicit correlation feature vectors of the local time series of the real-time inventory. In particular, the inventory level time series feature extractor based on the RNN model in the present application is essentially a recurrent neural network model.

[0061] In one embodiment of the present application, the real-time inventory level local time series association module is used to: input the time queue of the real-time inventory level into an inventory level time series feature extractor based on the RNN model to obtain a sequence of implicit association feature vectors of the real-time inventory local time series.

[0062] Furthermore, it is considered that there is a complex correlation response relationship between the market demand historical time series feature and the real-time inventory local time series implicit correlation feature. In order to more finely match the matching response relationship between the two in terms of time series features. In the technical solution of the present application, the market demand historical time series feature vector is used as a query feature vector, and the input fine-grained feature decoupled adaptive query encoding module is used to obtain a real-time inventory-market demand query response adaptive matching result representation matrix. Among them, the fine-grained feature decoupling adaptive query encoding module improves the matching degree between the market demand historical time series feature vector and the real-time inventory local time series implicit correlation feature vector sequence by fine-grained decoupling and adaptive correlation encoding of the real-time inventory local time series implicit correlation feature vector sequence. Specifically, firstly, the complex input features are decomposed by fine-grained decoupling of the sequence of the real-time inventory local time series implicit correlation feature vector, and independent and meaningful local detail sub-features are extracted. Then, the decoupled local sub-features are input into the sub-sequence semantic enhancement encoder of the Bi-LSTM model to process and enhance the contextual semantic information between the local sub-features. Then, the covariance matrix between the query features and the local correlation features containing contextual semantic information is calculated to quantify the semantic correlation between market demand and inventory level. After that, the matrix trace of the covariance matrix is ​​calculated to reflect the overall importance of the features, and the trace of each matrix is ​​used as a metric factor to input the sigmoid function to adaptively balance the relevance weights of each factor to obtain the attention weight. Finally, the attention weight is used as the weight value to perform weighted fusion on each covariance matrix to adjust the importance weight of each feature, ensuring that the model focuses on the features that are most critical to the prediction, and obtaining the real-time inventory-market demand query response adaptive matching result representation matrix.

[0063] In one embodiment of the present application, the real-time inventory-market demand adaptive query module includes: a fine-grained decoupling unit, which is used to perform fine-grained decoupling on the sequence of the real-time inventory local temporal implicit association feature vectors to obtain a set of subsequences of the real-time inventory local temporal implicit association feature vectors; a real-time inventory temporal implicit association fine-grained local association encoding unit, which is used to input each subsequence of the real-time inventory local temporal implicit association feature vector in the set of subsequences of the real-time inventory local temporal implicit association feature vectors into a subsequence semantic enhancement encoder to obtain a set of real-time inventory temporal implicit fine-grained decoupled local association feature vectors; a covariance matrix calculation unit, which is used to calculate the covariance matrix between the query feature vector and each real-time inventory temporal implicit fine-grained decoupled local association feature vector in the set of the real-time inventory temporal implicit fine-grained decoupled local association feature vectors. to obtain a set of query real-time inventory time series response covariance matrices; a query real-time inventory time series response trace metric factor calculation unit, used to calculate the matrix trace of each query real-time inventory time series response covariance matrix in the set of query real-time inventory time series response covariance matrices to obtain a set of query real-time inventory time series response trace metric factors; a query real-time inventory time series response trace metric factor activation unit, used to activate the set of query real-time inventory time series response trace metric factors to obtain a set of query real-time inventory time series response trace attention weights; a weighted sum calculation unit, used to use each query real-time inventory time series response trace attention weight in the set of query real-time inventory time series response trace attention weights as a weight, and calculate the weighted sum of the set of query real-time inventory time series response covariance matrices to obtain the real-time inventory-market demand query response adaptive matching result representation matrix.

[0064] Furthermore, in one embodiment of the present application, the real-time inventory time series implicit association fine-grained local association encoding unit is used to: input each subsequence of the real-time inventory local time series implicit association feature vector in the set of subsequences of the real-time inventory local time series implicit association feature vectors into a subsequence semantic enhancement encoder based on the Bi-LSTM model to obtain a set of the real-time inventory time series implicit fine-grained decoupled local association feature vectors.

[0065] Furthermore, in one embodiment of the present application, the query real-time inventory timing response trace metric factor calculation unit is used to: respectively calculate the sum of each element on the diagonal of each query real-time inventory timing response covariance matrix in the set of query real-time inventory timing response covariance matrices to obtain the set of query real-time inventory timing response trace metric factors.

[0066] Furthermore, in one embodiment of the present application, the query real-time inventory time series response trace measurement factor activation unit is used to: input the set of query real-time inventory time series response trace measurement factors into a Sigmoid activation function to obtain a set of query real-time inventory time series response trace attention weights.

[0067] Specifically, the market demand historical time series feature vector is used as the query feature vector, and the sequence of the market demand historical time series feature vector and the real-time inventory local time series implicit correlation feature vector is input into the fine-grained feature decoupling adaptive query encoding module, and processed with the following response adaptive matching formula to obtain the real-time inventory-market demand query response adaptive matching result representation matrix; wherein the response adaptive matching formula is:

[0068]

[0069]

[0070]

[0071]

[0072]

[0073]

[0074]

[0075] in, Each real-time inventory local time series implicit correlation feature vector in the sequence of the real-time inventory local time series implicit correlation feature vectors is represented, To decouple operations at a fine-grained level, represents a set of subsequences of implicitly associated feature vectors of the local time series of the real-time inventory, A subsequence of each real-time inventory local time series implicit correlation feature vector in a set of subsequences of the real-time inventory local time series implicit correlation feature vectors is represented, represents the Bi-LSTM encoder, is each real-time inventory time series implicit fine-grained decoupling local correlation feature vector in the set of the real-time inventory time series implicit fine-grained decoupling local correlation feature vectors, is the query feature vector, represents the operation of calculating the covariance matrix, is each query real-time inventory time series response covariance matrix in the set of query real-time inventory time series response covariance matrices, represents the calculation matrix trace, is each query real-time inventory time series response trace measurement factor in the set of query real-time inventory time series response trace measurement factors, yes function, is the attention weight of each query real-time inventory time series response trace in the set of attention weights of the query real-time inventory time series response trace, It is the real-time inventory-market demand query response adaptive matching result representation matrix.

[0076] Furthermore, the real-time inventory-market demand query response adaptive matching result representation matrix is ​​input into the classifier-based inventory management result generator to obtain an inventory management result, and the inventory management result is used to indicate whether a replenishment reminder is generated. That is, the real-time inventory-market demand query response adaptive matching result representation obtained by adaptive query encoding using the sequence of the market demand historical time series feature vector and the real-time inventory local time series implicit association feature vector is classified and processed to intelligently generate an inventory management result of whether a replenishment reminder is generated. In this way, inventory level data can be obtained in real time so that the current inventory status can be grasped in time and decision delays caused by data lags can be reduced. At the same time, combined with the time series association analysis of market demand historical data, market trends and demand changes can be quickly identified, reducing dependence on manual analysis, and reducing errors and delays caused by human factors, thereby improving the intelligence level of the inventory management process, so as to provide enterprises with accurate inventory strategy adjustment suggestions.

[0077] In one embodiment of the present application, the replenishment reminder judgment module is used to: input the real-time inventory-market demand query response adaptive matching result representation matrix into a classifier-based inventory management result generator to obtain the inventory management result, and the inventory management result is used to indicate whether a replenishment reminder is generated.

[0078] Furthermore, in one embodiment of the present application, the replenishment reminder judgment module includes: a matrix expansion unit, used to expand the real-time inventory-market demand query response adaptive matching result representation matrix into a classification feature vector according to a row vector or a column vector; a fully connected encoding unit, used to use multiple fully connected layers of the classifier-based inventory management result generator to fully connect encode the classification feature vector to obtain an encoded classification feature vector; a classification unit, used to pass the encoded classification feature vector through the Softmax classification function of the classifier-based inventory management result generator to obtain the inventory management result.

[0079] In a preferred example, considering that the sequence of the market demand historical time series feature vector and the real-time inventory local time series implicit association feature vector respectively represent the market demand local time series association feature and the real-time inventory local time series association, in this way, when the market demand historical time series feature vector is used as the query feature vector and the sequence of the market demand historical time series feature vector and the real-time inventory local time series implicit association feature vector are input into the fine-grained feature decoupling adaptive query encoding module, the real-time inventory-market demand query response adaptive matching result representation matrix obtained will also have interactive fusion distribution space structure differences due to the adaptive structure differences of the feature modal differences under the fine-grained time series, which will affect the accuracy of the inventory management results obtained by the real-time inventory-market demand query response adaptive matching result representation matrix input into the classifier-based inventory management result generator.

[0080] Based on this, when the real-time inventory-market demand query response adaptive matching result representation matrix is ​​input into the classifier-based inventory management result generator to obtain the inventory management result, the real-time inventory-market demand query response adaptive matching result representation matrix is ​​optimized, and the optimization process specifically includes:

[0081] The absolute value sum of all eigenvalues ​​of the real-time inventory-market demand query response adaptive matching result representation matrix is ​​calculated to obtain a first real-time inventory-market demand query response adaptive matching result space structure value, and the square root of the square sum is calculated to obtain a second real-time inventory-market demand query response adaptive matching result space structure value, that is:

[0082]

[0083]

[0084] in, The first one represents the real-time inventory-market demand query response adaptive matching result representation matrix eigenvalues, represents the number of eigenvalues ​​of all eigenvalues ​​of the real-time inventory-market demand query response adaptive matching result representation matrix, Represents the spatial structure value of the first real-time inventory-market demand query response adaptive matching result, represents the second real-time inventory-market demand query response adaptive matching result spatial structure value;

[0085] For each eigenvalue of the real-time inventory-market demand query response adaptive matching result representation matrix, the first real-time inventory-market demand query response adaptive matching result spatial structure value minus the product of the eigenvalue and the total number of eigenvalues ​​of the real-time inventory-market demand query response adaptive matching result representation matrix is ​​calculated to obtain the first real-time inventory-market demand query response adaptive matching result long-range dependency value ,in, represents the spatial structure value of the first real-time inventory-market demand query response adaptive matching result, The first one represents the real-time inventory-market demand query response adaptive matching result representation matrix eigenvalues, represents the number of eigenvalues ​​of all eigenvalues ​​of the real-time inventory-market demand query response adaptive matching result representation matrix, Indicates the long-range dependency value of the first real-time inventory-market demand query response adaptive matching result;

[0086] Calculate the square root of the total number of eigenvalues ​​multiplied by the product of the eigenvalues ​​minus the second real-time inventory-market demand query response adaptive matching result spatial structure value to obtain the second real-time inventory-market demand query response adaptive matching result long-range dependency value ,in, Represents the second real-time inventory-market demand query response adaptive matching result space structure value, The first one represents the real-time inventory-market demand query response adaptive matching result representation matrix eigenvalues, represents the number of eigenvalues ​​of all eigenvalues ​​of the real-time inventory-market demand query response adaptive matching result representation matrix, Indicates the long-range dependency value of the second real-time inventory-market demand query response adaptive matching result;

[0087] The index value obtained by taking the first real-time inventory-market demand query response adaptive matching result long-range dependency value as the exponent of the natural constant and the inverse of the second real-time inventory-market demand query response adaptive matching result long-range dependency value are weighted summed to obtain the optimized eigenvalue corresponding to each eigenvalue ,in, Indicates the long-range dependency value of the first real-time inventory-market demand query response adaptive matching result, Indicates the long-range dependency value of the second real-time inventory-market demand query response adaptive matching result, represents a natural constant, and represents the weighted hyperparameter, Indicates the optimized eigenvalue corresponding to each eigenvalue;

[0088] The optimized eigenvalues ​​are reshaped to obtain an optimized real-time inventory-market demand query response adaptive matching result representation matrix.

[0089] Based on this, in view of the possible spatial structure loss in the high-dimensional space of the feature set of the real-time inventory-market demand query response adaptive matching result representation matrix, which leads to the convergence inconsistency of the image semantic segmentation based on the implicit inference of the spatial structure information of the feature, a long-distance feature dependency relationship is established based on the overall feature scale of the real-time inventory-market demand query response adaptive matching result representation matrix relative to the spatial structure representation of the real-time inventory-market demand query response adaptive matching result representation matrix, so as to establish the feature local connectivity of the real-time inventory-market demand query response adaptive matching result representation matrix, and to capture the spatial ambiguity information of the object feature value through the unstructured feature value point prediction of the real-time inventory-market demand query response adaptive matching result representation matrix, thereby improving the spatial inductive deviation perception ability of the feature set of the real-time inventory-market demand query response adaptive matching result representation matrix, improving the semantic segmentation convergence consistency, and improving the accuracy of the inventory management result obtained by the inventory management result generator based on the classifier input by the optimized real-time inventory-market demand query response adaptive matching result representation matrix. In this way, the inventory level data can be obtained in real time, so that the current inventory status can be grasped in time, and the decision delay caused by data lag can be reduced. At the same time, combined with the time series correlation analysis of historical market demand data, it can quickly identify market trends and demand changes, reduce reliance on manual analysis, and reduce errors and delays caused by human factors, thereby improving the intelligence level of the inventory management process and providing companies with accurate inventory strategy adjustment recommendations.

[0090] In summary, the above scheme is adopted to obtain the time queue of the real-time inventory level, and obtain the market demand historical data based on the timestamps of the start and end of the time queue of the real-time inventory level, and use artificial intelligence-based data analysis and processing technology to extract the time series features and implicit association analysis of the real-time inventory level and the market demand historical data, so as to use the market demand historical data as the query feature to perform adaptive query matching based on the time series with the real-time inventory level to obtain the response matching result representation, thereby intelligently generating the inventory management result of whether to generate a replenishment reminder. This application can obtain inventory level data in real time, and characterize the inventory change pattern in combination with the time series dynamic changes and trends of the inventory, so as to be able to grasp the current inventory status in a timely and accurate manner, and reduce the decision delays caused by data lag. At the same time, combined with the time series association analysis of the market demand historical data, it can quickly identify market trends and demand changes, reduce the reliance on manual analysis, and reduce the errors and delays caused by human factors, thereby improving the intelligence level of the inventory management process, so as to provide enterprises with accurate inventory strategy adjustment suggestions.

[0091] Figure 2 is a flowchart of an ERP management method based on real-time analysis of low-code platform fusion data according to an exemplary embodiment. Figure 2 As shown, the method includes:

[0092] Step S201, obtaining a time queue of real-time inventory level;

[0093] Step S202, extracting a start timestamp and an end timestamp from the time queue of the real-time inventory level;

[0094] Step S203: intercepting market demand historical data based on the start timestamp and the end timestamp;

[0095] Step S204: inputting the market demand history data into a market demand time series feature extractor to obtain a market demand history time series feature vector;

[0096] Step S205, inputting the time queue of the real-time inventory level into the inventory level time series feature extractor to obtain a sequence of real-time inventory local time series implicit correlation feature vectors;

[0097] Step S206: using the market demand historical time series feature vector as a query feature vector, inputting the sequence of the market demand historical time series feature vector and the real-time inventory local time series implicit correlation feature vector into a fine-grained feature decoupling adaptive query encoding module to obtain a real-time inventory-market demand query response adaptive matching result representation matrix as a real-time inventory-market demand query response adaptive matching result representation feature;

[0098] Step S207: based on the representation characteristics of the adaptive matching result of the real-time inventory-market demand query response, an inventory management result is obtained, wherein the inventory management result is used to indicate whether a replenishment reminder is generated.

[0099] The above description is only a preferred embodiment of the present application and an explanation of the technical principles used. Those skilled in the art should understand that the scope of disclosure involved in the present application is not limited to the technical solution formed by a specific combination of the above technical features, but should also cover other technical solutions formed by any combination of the above technical features or their equivalent features without departing from the above disclosed concept. For example, the above features are replaced with the technical features with similar functions disclosed in this application (but not limited to) by each other to form a technical solution.

[0100] In addition, although each operation is described in a specific order, this should not be construed as requiring these operations to be performed in the specific order shown or to be performed in a sequential order. Under certain circumstances, multitasking and parallel processing may be advantageous. Similarly, although some specific implementation details are included in the above discussion, these should not be construed as limiting the scope of the application. Some features described in the context of a separate embodiment can also be implemented in a single embodiment in combination. On the contrary, the various features described in the context of a single embodiment can also be implemented in multiple embodiments individually or in any suitable sub-combination mode.

[0101] Although the subject matter has been described in language specific to structural features and / or method logic actions, it should be understood that the subject matter defined is not necessarily limited to the specific features or actions described above. On the contrary, the specific features and actions described above are merely exemplary forms of implementation. With respect to the apparatus in the above-mentioned embodiments, the specific manner in which each module performs the operation has been described in detail in the embodiments related to the method, and will not be described in detail here.

Claims

1. An ERP management system based on a low-code platform that integrates data and real-time analysis, characterized by: include: Real-time inventory level acquisition module, used to obtain the time queue of real-time inventory level; A real-time inventory level timestamp extraction module, configured to extract a start timestamp and an end timestamp from a time queue of the real-time inventory level; A market demand historical data interception module, used to intercept the market demand historical data based on the start timestamp and the end timestamp; A market demand historical data time series feature extraction module, used for inputting the market demand historical data into a market demand time series feature extractor to obtain a market demand historical time series feature vector; A real-time inventory level local time series association module, used for inputting the time queue of the real-time inventory level into the inventory level time series feature extractor to obtain a sequence of real-time inventory local time series implicit association feature vectors; A real-time inventory-market demand adaptive query module is used to use the market demand historical time series feature vector as a query feature vector, and input the sequence of the market demand historical time series feature vector and the real-time inventory local time series implicit correlation feature vector into a fine-grained feature decoupling adaptive query encoding module to obtain a real-time inventory-market demand query response adaptive matching result representation matrix as a real-time inventory-market demand query response adaptive matching result representation feature; The replenishment reminder judgment module is used to obtain an inventory management result based on the adaptive matching result representation characteristics of the real-time inventory-market demand query response, and the inventory management result is used to indicate whether a replenishment reminder is generated.

2. The ERP management system based on low-code platform fusion data real-time analysis according to claim 1 is characterized in that: The market demand historical data temporal feature extraction module is used to: input the market demand historical data into a market demand temporal feature extractor based on a 1D-CNN model to obtain the market demand historical temporal feature vector.

3. The ERP management system based on low-code platform fusion data real-time analysis according to claim 2 is characterized in that: The real-time inventory level local time series association module is used to: input the time queue of the real-time inventory level into the inventory level time series feature extractor based on the RNN model to obtain a sequence of the real-time inventory local time series implicit association feature vectors.

4. The ERP management system based on low-code platform fusion data real-time analysis according to claim 3 is characterized in that: The real-time inventory-market demand adaptive query module includes: A fine-grained decoupling unit, configured to perform fine-grained decoupling on the sequence of the real-time inventory local time series implicit correlation feature vectors to obtain a set of subsequences of the real-time inventory local time series implicit correlation feature vectors; A real-time inventory time series implicit association fine-grained local association encoding unit is used to input each subsequence of the real-time inventory local time series implicit association feature vector in the set of subsequences of the real-time inventory local time series implicit association feature vector into a subsequence semantic enhancement encoder to obtain a set of real-time inventory time series implicit fine-grained decoupled local association feature vectors; A covariance matrix calculation unit, used for calculating the covariance matrix between the query feature vector and each real-time inventory time series implicit fine-grained decoupled local correlation feature vector in the set of the real-time inventory time series implicit fine-grained decoupled local correlation feature vectors to obtain a set of query real-time inventory time series response covariance matrices; A real-time inventory time series response trace metric factor calculation unit is used to calculate the matrix trace of each real-time inventory time series response covariance matrix in the set of real-time inventory time series response covariance matrices to obtain a set of real-time inventory time series response trace metric factors; A query real-time inventory time series response trace metric factor activation unit, used to activate the set of query real-time inventory time series response trace metric factors to obtain a set of query real-time inventory time series response trace attention weights; A weighted sum calculation unit is used to calculate the weighted sum of the set of covariance matrices of the query real-time inventory time series response using each query real-time inventory time series response trace attention weight in the set of attention weights of the query real-time inventory time series response as a weight to obtain the real-time inventory-market demand query response adaptive matching result representation matrix.

5. The ERP management system based on low-code platform fusion data real-time analysis according to claim 4 is characterized in that: The real-time inventory time series implicit association fine-grained local association encoding unit is used to: input each subsequence of the real-time inventory local time series implicit association feature vector in the set of subsequences of the real-time inventory local time series implicit association feature vectors into a subsequence semantic reinforcement encoder based on the Bi-LSTM model to obtain a set of the real-time inventory time series implicit fine-grained decoupled local association feature vectors.

6. The ERP management system based on low-code platform fusion data real-time analysis according to claim 5 is characterized in that: The real-time inventory time series response trace metric factor calculation unit is used to respectively calculate the sum of the elements on the diagonal of each query real-time inventory time series response covariance matrix in the set of the query real-time inventory time series response covariance matrix to obtain the set of the query real-time inventory time series response trace metric factors.

7. The ERP management system based on low-code platform fusion data real-time analysis according to claim 6 is characterized in that: The query real-time inventory time series response trace measurement factor activation unit is used to: input the set of query real-time inventory time series response trace measurement factors into a Sigmoid activation function to obtain a set of query real-time inventory time series response trace attention weights.

8. The ERP management system based on low-code platform fusion data real-time analysis according to claim 7 is characterized in that: The replenishment reminder judgment module is used to: input the real-time inventory-market demand query response adaptive matching result representation matrix into a classifier-based inventory management result generator to obtain the inventory management result, and the inventory management result is used to indicate whether a replenishment reminder is generated.

9. The ERP management system based on low-code platform fusion data real-time analysis according to claim 8 is characterized in that: The replenishment reminder judgment module includes: A matrix expansion unit, used to expand the real-time inventory-market demand query response adaptive matching result representation matrix into a classification feature vector according to row vectors or column vectors; A fully connected encoding unit is used to perform fully connected encoding on the classification feature vector using multiple fully connected layers of the classifier-based inventory management result generator to obtain an encoded classification feature vector.

10. The ERP management system based on low-code platform fusion data real-time analysis according to claim 9 is characterized in that: The replenishment reminder judgment module also includes: A classification unit is used to pass the encoded classification feature vector through the Softmax classification function of the classifier-based inventory management result generator to obtain the inventory management result.

Images