Material management system and method for medical rescue scene

Abstract

The invention relates to a data analysis technology, and discloses a material management system and method for a medical rescue scene, and the method comprises the steps: obtaining multi-stage warehouse data of materials, obtaining material path data according to the multi-stage warehouse data, generating a spatial feature matrix according to the multi-stage warehouse data and the material path data, and obtaining historical data of material consumption, predicting material demand data in a first preset duration according to historical data and the spatial feature matrix, calculating stockout coefficients of different types of materials according to the multi-level warehouse data, the material demand data and the spatial feature matrix, and obtaining real consumption data of the materials in a second preset duration, and obtaining predicted demand data included in the material demand data in a second preset duration, calculating a demand error value of the real consumption data and the predicted demand data, and generating replenishment data based on the demand error value and the stockout coefficient. According to the invention, the efficiency of material replenishment in a medical rescue scene can be improved.

Description

Technical Field

[0001] The present invention relates to the technical field of data analysis, and in particular to a material management system and method thereof for medical rescue scenarios. Background Art

[0002] In medical rescue scenarios, the timely supply of materials is crucial. However, there are many problems with the current inventory management of medical rescue materials. Traditional inventory management methods often rely on experience and lack scientific data analysis and prediction, resulting in insufficient or excessive material reserves. Insufficient reserves may affect the smooth progress of rescue work, while excessive reserves will cause waste of resources and increased costs. In addition, the existing inventory management system is difficult to track the dynamic changes of materials in real time, and cannot accurately allocate inventory according to the needs of different rescue scenarios. It is impossible to analyze the appropriate replenishment data in real time based on the current material consumption data and inventory data, resulting in reduced replenishment efficiency, which in turn reduces the efficiency of material supply in rescue scenarios. Summary of the invention

[0003] The present invention provides a material management system and method for medical rescue scenarios, the main purpose of which is to solve the problem of low replenishment efficiency of existing material management methods for medical rescue scenarios.

[0004] To achieve the above-mentioned purpose, the present invention provides a material management system for medical rescue scenarios, characterized in that the system includes a data acquisition module, a feature generation module, a data prediction module, a data calculation module and a data generation module, wherein: The data acquisition module is used to acquire multi-level warehouse data of materials, and acquire material path data according to the multi-level warehouse data; The feature generation module is used to generate a spatial feature matrix according to the multi-level warehouse data and the material path data; The data prediction module is used to predict material demand data within a first preset time period based on historical data of material consumption and the spatial feature matrix; The data calculation module is used to calculate the out-of-stock coefficient according to the multi-level warehouse data, the material demand data and the spatial feature matrix, and calculate the demand error value according to the actual consumption data of the materials within the second preset time period and the predicted demand data contained in the material demand data within the second preset time period; The data generation module is used to generate replenishment data based on the demand error value and the out-of-stock coefficient.

[0005] Optionally, when the feature generation module generates a spatial feature matrix according to the multi-level warehouse data and the material path data, it is specifically used to: Constructing a warehouse location map based on the multi-level warehouse data; Constructing a warehouse connection map according to the material path data and the warehouse location map; Feature extraction is performed on the warehouse connection graph to obtain a spatial feature matrix.

[0006] Optionally, when the feature generation module constructs the warehouse connection map according to the material path data and the warehouse location map, it is specifically used to: Generate an adaptive path curve of corresponding proportion according to the material path data; Obtaining the path connection relationship data contained in the material path duration data; The adaptive path curve is drawn into the warehouse location map based on the path connection relationship data to obtain a warehouse connection map.

[0007] Optionally, when predicting the material demand data within the first preset time period according to the historical data of material consumption and the spatial feature matrix, the data prediction module is specifically used to: Generate a splicing feature according to the historical data and the spatial feature matrix; Using a pre-trained material consumption prediction model to predict demand data within a first preset time period according to the splicing features, to obtain initial demand data; The initial demand data is adjusted based on a preset loss weight to obtain material demand data.

[0008] Optionally, when the data calculation module calculates the stock-out coefficient of a material according to the material demand data and the spatial feature matrix, it is specifically used to: Obtaining the inventory of each level of warehouse in the multi-level warehouse data, and obtaining the warehouse demand of each level of warehouse in the material demand data; Calculate the total inventory and total material demand based on the multi-level warehouse data and material demand data; The logarithmic term is obtained by performing logarithmic operation on the warehouse demand and the total inventory; Perform multiplication and division operations on the element mean of the spatial feature matrix, the warehouse demand, and the inventory to obtain a ratio term; The logarithmic term and the ratio term are summed and then multiplied by a preset weight parameter to obtain a comprehensive weight term; Calculate the ratio of the total material demand to the total inventory and then multiply it by a preset supply-demand ratio weight parameter to obtain a supply-demand ratio item; The comprehensive weight item is added to the supply-demand ratio item to obtain the warehouse out-of-stock coefficient of each level of warehouse, and the warehouse out-of-stock coefficients of warehouses of all levels are added to obtain the out-of-stock coefficient.

[0009] Optionally, in the data calculation module, the calculation formula of the out-of-stock coefficient is as follows: in, is the out-of-stock coefficient, is the warehouse level, For the The weight coefficient of the level warehouse, The material demand data The warehouse demand of the first-level warehouse, For multi-level warehouse data The inventory level of the warehouse, is the total number of elements in the spatial feature matrix, is the spatial feature matrix The value of the element, To preset a very small number, is the supply-demand ratio weight parameter, is the total material demand in the material demand data, It is the total inventory in the multi-level warehouse data.

[0010] Optionally, when the data calculation module calculates the demand error value according to the actual consumption data of the material within the second preset time period and the predicted demand data included in the material demand data within the second preset time period, it is specifically used to: Dividing the actual consumption data and the predicted demand data into time windows to obtain actual window data and predicted window data; Calculate the mean of each time window in the real window data and the predicted window data to obtain real window mean data and predicted window mean data; A demand error value is calculated according to the real window data, the predicted window data, the real window mean data and the predicted window mean data.

[0011] Optionally, in the data calculation module, the calculation formula of the required error value is as follows: The required error value is calculated using the following formula: in, is the required error value, is the number of windows, represents the hyperbolic tangent function, Indicates the real window mean data The mean data of the windows, Indicates the first The mean data of the windows, represents the number of samples in a single window, Indicates the real window data In the window data, Indicates the prediction window data In the window data, is the preset confidence deviation penalty coefficient, is the base of natural logarithms.

[0012] Optionally, when generating replenishment data based on the demand error value and the out-of-stock coefficient, the data generating module is specifically used to: Calculate initial replenishment data according to the preset ratio parameter and the out-of-stock coefficient; The replenishment data is confirmed according to the magnitude relationship between the demand error value and a preset error threshold value and the initial replenishment data.

[0013] In order to solve the above problems, the present invention also provides a material management method for medical rescue scenarios, the method comprising: Acquire multi-level warehouse data of materials, and acquire material path data according to the multi-level warehouse data; Generate a spatial feature matrix according to the multi-level warehouse data and the material path data; Predicting material demand data within a first preset time period based on historical data of material consumption and the spatial feature matrix; Calculate the out-of-stock coefficient according to the multi-level warehouse data, the material demand data and the spatial feature matrix; Calculating a demand error value according to actual consumption data of materials within a second preset time period and predicted demand data included in the material demand data within the second preset time period; Replenishment data is generated based on the demand error value and the out-of-stock factor.

[0014] The embodiment of the present invention obtains multi-level warehouse data of materials, obtains material path data based on the multi-level warehouse data, generates a spatial feature matrix based on the multi-level warehouse data and the material path data, predicts material demand data within a first preset time period based on historical data of material consumption and the spatial feature matrix, calculates a stock-out coefficient based on the multi-level warehouse data, the material demand data and the spatial feature matrix, calculates a demand error value based on actual material consumption data within a second preset time period and the predicted demand data contained in the material demand data within the second preset time period, and generates replenishment data based on the demand error value and the stock-out coefficient. Therefore, the material management system and method for medical rescue scenarios proposed by the present invention can solve the problem of low replenishment efficiency in existing material management methods for medical rescue scenarios. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 A functional module diagram of a material management system for medical rescue scenarios provided by an embodiment of the present invention; Figure 2 A schematic diagram of a flow chart of a material management method for a medical rescue scenario provided by an embodiment of the present invention; The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0016] It should be understood that the specific embodiments described herein are only used to explain the present invention, and are not used to limit the present invention.

[0017] The embodiment of the present application provides a material management method for a medical rescue scenario. The execution subject of the material management method for a medical rescue scenario includes but is not limited to at least one of the electronic devices such as a server, a terminal, etc. that can be configured to execute the method provided by the embodiment of the present application. In other words, the material management method for a medical rescue scenario can be executed by software or hardware installed on a terminal device or a server device, and the software can be a blockchain platform. The server includes but is not limited to: a single server, a server cluster, a cloud server or a cloud server cluster, etc. The server can be an independent server, or it can be a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, content delivery networks (Content Delivery Network, CDN), and big data and artificial intelligence platforms.

[0018] like Figure 1 1 is a functional module diagram of a material management system for medical rescue scenarios provided by one embodiment of the present invention.

[0019] The material management system 100 for medical rescue scenarios of the present invention can be installed in an electronic device. According to the functions to be implemented, the material management system 100 for medical rescue scenarios can include a data acquisition module 101, a feature generation module 102, a data prediction module 103, a data calculation module 104 and a data generation module 105. The module of the present invention can also be called a unit, which refers to a series of computer program segments that can be executed by an electronic device processor and can complete fixed functions, which are stored in the memory of the electronic device.

[0020] In this embodiment, the functions of each module / unit are as follows: The data acquisition module 101 is used to acquire multi-level warehouse data of materials, and acquire material path data according to the multi-level warehouse data; The feature generation module 102 is used to generate a spatial feature matrix according to the multi-level warehouse data and the material path data; The data prediction module 103 is used to predict material demand data within a first preset time period according to the historical data of material consumption and the spatial feature matrix; The data calculation module 104 is used to calculate the out-of-stock coefficient according to the multi-level warehouse data, the material demand data and the spatial feature matrix, and calculate the demand error value according to the actual consumption data of the material within the second preset time period and the predicted demand data contained in the material demand data within the second preset time period; The data generating module 105 is used to generate replenishment data based on the demand error value and the out-of-stock coefficient.

[0021] The data acquisition module 101 is used to acquire multi-level warehouse data of materials, and acquire material path data according to the multi-level warehouse data.

[0022] In the embodiment of the present invention, the multi-level warehouse data may refer to material-related data of warehouses of different levels, and the warehouses of different levels may include a central warehouse, a regional warehouse, and an emergency vehicle warehouse.

[0023] In the embodiment of the present invention, the multi-level warehouse data of materials is obtained by real-time data connection with the management system of each level of warehouses to collect relevant information of various materials, including the type, quantity, inventory location, entry and exit time of the materials, etc. These data reflect the real-time status of the materials in each warehouse and provide a basis for subsequent analysis.

[0024] In the embodiment of the present invention, the material path data is obtained according to the multi-level warehouse data, and the flow trajectory of the materials between different warehouses is analyzed based on the obtained multi-level warehouse data. By tracking the in-and-out records of the materials, it can be determined which intermediate warehouses the materials pass through from the starting warehouse to the final destination of the materials. This information constitutes the material path data, which helps to understand the material circulation process, clarify the allocation relationship between materials in different warehouses, and provide a basis for subsequent analysis.

[0025] In detail, the material path data includes data on the time it takes for materials to be shipped from one warehouse to another warehouse.

[0026] In the embodiment of the present invention, by acquiring multi-level warehouse data of materials and acquiring material path data according to the multi-level warehouse data, the efficiency of subsequent generation of a spatial feature matrix can be improved.

[0027] The feature generation module 102 is used to generate a spatial feature matrix according to the multi-level warehouse data and the material path data.

[0028] In the embodiment of the present invention, the spatial characteristic matrix is ​​generated according to the multi-level warehouse data and the material path data, which presents the complex material spatial distribution and flow information in the form of a mathematical matrix for subsequent data analysis and processing. This matrix integrates the multi-level warehouse data and the material path data, and can fully reflect the spatial characteristics of the material.

[0029] In detail, the rows of the matrix can represent different warehouses or material storage nodes, and the columns can represent various material-related features. For example, a row corresponds to a specific warehouse, and the elements in the column may include the inventory quantity of different types of materials in the warehouse, the connection relationship between the warehouse and other warehouses (reflecting the material flow path), the geographical location information of the warehouse, etc. In this way, the spatial distribution, quantity information, and flow path of materials are integrated into a matrix, providing a unified data structure for subsequent material demand forecasting and analysis.

[0030] In the embodiment of the present invention, when the feature generation module 102 generates a spatial feature matrix according to the multi-level warehouse data and the material path data, it is specifically used to: Constructing a warehouse location map based on the multi-level warehouse data; Constructing a warehouse connection map according to the material path data and the warehouse location map; Feature extraction is performed on the warehouse connection graph to obtain a spatial feature matrix.

[0031] In detail, constructing a warehouse location map based on the multi-level warehouse data is accomplished by acquiring the warehouse location data contained in the multi-level warehouse data and constructing a warehouse location map based on the warehouse location data.

[0032] In detail, the feature extraction of the warehouse connection map to obtain the spatial feature matrix is ​​carried out by graying the warehouse connection map to obtain a grayed image, downsampling the grayed image to obtain a downsampled image, and digitizing the downsampled image to obtain the spatial feature matrix.

[0033] In detail, the warehouse location data includes the latitude and longitude data of the central warehouse, the regional warehouse, and the emergency vehicle warehouse.

[0034] In detail, constructing a warehouse location map based on the warehouse location data is to draw an image containing each warehouse location according to a preset scale based on the relative positions between warehouses of different levels.

[0035] In detail, the grayscale processing of the warehouse connection map to obtain the grayscale image is to convert each pixel of the image from a plurality of color channels (such as RGB) into a single grayscale value. The grayscale value is usually an integer between 0 and 255, indicating the brightness of the pixel.

[0036] In detail, downsampling the grayscale image to obtain a downsampled image may refer to dividing the grayscale image into multiple pixel blocks of the same size, taking the pixel mean of each pixel block as the feature value of the pixel block, and summarizing all the feature values ​​to obtain a new image, which has a smaller size but higher feature expression capability.

[0037] In detail, the numerical processing of the downsampled image refers to converting each pixel point according to the downsampled image into a numerical value according to its gray value.

[0038] In detail, when the feature generation module 102 constructs the warehouse connection map according to the material path data and the warehouse location map, it is specifically used to: Generate an adaptive path curve of corresponding proportion according to the material path data; Obtaining the path connection relationship data contained in the material path duration data; The adaptive path curve is drawn into the warehouse location map based on the path connection relationship data to obtain a warehouse connection map.

[0039] In detail, the method of generating an adaptive path curve of corresponding proportion according to the material path data is performed by acquiring the material path duration data contained in the material path data, and generating an adaptive path curve of corresponding proportion based on each path duration in the material path duration data.

[0040] In detail, the material path duration data includes the duration data of each path.

[0041] In an embodiment of the present invention, since the road conditions in the medical rescue scenario may be more complicated, the corresponding path data is also more complicated. For example, there are two regional warehouses with the same straight-line distance from the central warehouse, but the actual paths for transporting materials may be different, which can be reflected in the different path passing time in this solution.

[0042] In detail, the path connection relationship data included in the material path duration data is obtained by obtaining the connection relationship between the central warehouse and multiple regional warehouses, and the connection relationship between each regional warehouse and multiple emergency vehicle warehouses.

[0043] In detail, the step of drawing the adaptive path curve into the warehouse location map based on the path connection relationship data to obtain the warehouse connection map means that the warehouse location map is used as the base map, and the adaptive path curve is drawn between the corresponding warehouse locations according to the connection relationship between warehouses determined by the path connection relationship data. In this way, by connecting each warehouse with a curve, a warehouse connection map is formed. In this map, not only can the location distribution of the warehouses be seen, but also the flow path of materials between different warehouses and the time cost difference of the path can be intuitively understood through the shape and connection method of the curve.

[0044] In detail, the adaptive path curve refers to a curve that can adaptively adjust the curvature according to the ratio of the straight-line distance between two warehouses and the path passing time, so as to realize the connection between the two warehouses. For example, when the straight-line distance between the two warehouses is fixed, the longer the actual passing time for materials to complete the entry and exit of the two warehouses, the longer the corresponding curve length. The curve can adaptively adjust the curvature of the curve to adapt to the difference between the straight-line distance and the passing time.

[0045] In the embodiment of the present invention, by generating a spatial feature matrix according to the multi-level warehouse data and the material path data, the accuracy of subsequent prediction of material demand data is improved.

[0046] The data prediction module 103 is used to predict material demand data within a first preset time period according to historical data of material consumption and the spatial feature matrix.

[0047] In the embodiment of the present invention, when the data prediction module 103 predicts the material demand data within the first preset time period according to the historical data of material consumption and the spatial feature matrix, it is specifically used to: Generate a splicing feature according to the historical data and the spatial feature matrix; Using a pre-trained material consumption prediction model to predict demand data within a first preset time period according to the splicing features, to obtain initial demand data; The initial demand data is adjusted based on a preset loss weight to obtain material demand data.

[0048] In detail, the splicing features are generated based on the historical data and the spatial feature matrix by extracting the periodic features and trend features of the historical data, and performing tensor splicing of the periodic features and the trend features with the spatial feature matrix to obtain the splicing features.

[0049] In detail, the material consumption prediction model is a spatiotemporal graph convolutional network model, which is obtained by obtaining a set of material consumption history data and dividing it into training data and test data; The periodic characteristics and trend characteristics of the training data are extracted, and the spatial characteristics of the training data are obtained. After splicing the extracted features, the spatiotemporal graph convolutional network model is used to predict the future consumption data of materials based on the spliced ​​features. The mean square error function is used as the loss function to calculate the loss value of the future consumption data and the test data, and the parameters of the optimization model are adjusted according to the loss value. When the loss value converges or reaches the preset number of iterations, the model training is confirmed to be completed.

[0050] In detail, the extraction of the periodic characteristics and trend characteristics of the historical data can be based on the periodic decomposition of the time series of the historical data to obtain the periodic characteristics, and the calculation of the average value of the historical data within a certain time window can smooth the data and extract trend characteristics.

[0051] In detail, in the field of mathematics and computer science, a tensor is an array structure that can represent data of different dimensions. The tensor splicing is to merge multiple tensors according to a specific dimension to form a new tensor. The tensor splicing of the periodic feature and the trend feature with the spatial feature matrix to obtain the spliced ​​feature is to merge the tensors representing the periodic feature and the trend feature with the spatial feature matrix according to a preset rule.

[0052] In detail, the initial demand data is adjusted based on the preset loss weight to obtain the material demand data, which means assigning higher weights to material demand nodes with higher importance. For example, the importance of the emergency vehicle warehouse requires adjusting the material demand data of the emergency vehicle warehouse based on a higher weight to ensure sufficient supply of materials at the emergency site.

[0053] In the embodiment of the present invention, the time span of the historical data needs to be greater than the first preset time length.

[0054] In the embodiment of the present invention, by predicting the material demand data within the first preset time period based on the historical data of material consumption and the spatial feature matrix, the gap in material demand can be calculated first, so as to facilitate immediate replenishment.

[0055] The data calculation module 104 is used to calculate the out-of-stock coefficient based on the multi-level warehouse data, the material demand data and the spatial feature matrix, and calculate the demand error value based on the actual consumption data of the materials within the second preset time period and the predicted demand data contained in the material demand data within the second preset time period.

[0056] In the embodiment of the present invention, the out-of-stock coefficient is a coefficient used to represent the out-of-stock quantity of a certain material, and is a key parameter used for subsequent replenishment data.

[0057] In the embodiment of the present invention, when the data calculation module 104 calculates the stock-out coefficient of a material according to the material demand data and the spatial characteristic matrix, it is specifically used to: Obtaining the inventory of each level of warehouse in the multi-level warehouse data, and obtaining the warehouse demand of each level of warehouse in the material demand data; Calculate the total inventory and total material demand based on the multi-level warehouse data and material demand data; The logarithmic term is obtained by performing logarithmic operation on the warehouse demand and the total inventory; Perform multiplication and division operations on the element mean of the spatial feature matrix, the warehouse demand, and the inventory to obtain a ratio term; The logarithmic term and the ratio term are summed and then multiplied by a preset weight parameter to obtain a comprehensive weight term; Calculate the ratio of the total material demand to the total inventory and then multiply it by a preset supply-demand ratio weight parameter to obtain a supply-demand ratio item; The comprehensive weight item is added to the supply-demand ratio item to obtain the warehouse out-of-stock coefficient of each level of warehouse, and the warehouse out-of-stock coefficients of warehouses of all levels are added to obtain the out-of-stock coefficient.

[0058] In detail, in the data calculation module 104, the calculation formula of the out-of-stock coefficient is as follows: in, is the out-of-stock coefficient, is the warehouse level, For the The weight coefficient of the level warehouse, The material demand data The warehouse demand of the first-level warehouse, For multi-level warehouse data The inventory level of the warehouse, is the total number of elements in the spatial feature matrix, is the spatial feature matrix The value of the element, To preset a very small number, is the supply-demand ratio weight parameter, is the total material demand in the material demand data, It is the total inventory in the multi-level warehouse data.

[0059] In the embodiment of the present invention, the number of warehouse levels may be 3. A level warehouse can refer to a level 1 warehouse, a level 2 warehouse, or a level 3 warehouse.

[0060] In detail, the first-level warehouse may refer to a central warehouse, the second-level warehouse may refer to a regional warehouse, and the third-level warehouse may refer to an emergency vehicle warehouse.

[0061] In detail, the total material demand is obtained by summing up the material demand of each warehouse in the material demand data.

[0062] In detail, the total inventory amount is obtained by summing up the inventory amount of each warehouse in the multi-level warehouse data.

[0063] In an embodiment of the present invention, the second preset time length is less than the preset first time length, and the second preset time length has the same starting point as the first preset time length. For example, the preset first time length is from 8 a.m. to 8 p.m. on the same day, and the preset second time length may be from 8 a.m. to 12 noon on the same day.

[0064] In an embodiment of the present invention, the demand error value is calculated based on the actual consumption data of materials within the second preset time period and the predicted demand data contained in the material demand data within the second preset time period. This is to judge the prediction accuracy of the predicted data by comparing the predicted data with the actual data in a short period of time, which can improve the efficiency of judgment and detect prediction errors as early as possible.

[0065] In the embodiment of the present invention, when the data calculation module 104 calculates the demand error value according to the actual consumption data of the material within the second preset time period and the predicted demand data contained in the material demand data within the second preset time period, it is specifically used to: Dividing the actual consumption data and the predicted demand data into time windows to obtain actual window data and predicted window data; Calculate the mean of each time window in the real window data and the predicted window data to obtain real window mean data and predicted window mean data; A demand error value is calculated according to the real window data, the predicted window data, the real window mean data and the predicted window mean data.

[0066] In detail, in the data calculation module 104, the calculation formula of the required error value is as follows: in, is the required error value, is the number of windows, represents the hyperbolic tangent function, Indicates the real window mean data The mean data of the windows, Indicates the first The mean data of the windows, represents the number of samples in a single window, Indicates the real window data In the window data, Indicates the prediction window data In the window data, is the preset confidence deviation penalty coefficient, is the base of natural logarithms.

[0067] In detail, the hyperbolic tangent function can realize normalization of the value of the input function. In detail, the confidence deviation penalty coefficient may be 1.2.

[0068] In an embodiment of the present invention, by calculating the demand error value based on the actual consumption data of materials within the second preset time period and the predicted demand data contained in the material demand data within the second preset time period, it is possible to promptly determine whether the predicted demand data is accurate, thereby ensuring the accuracy of subsequent calculations of replenishment data.

[0069] The data generation module 105 is used to generate replenishment data based on the demand error value and the out-of-stock coefficient.

[0070] In the embodiment of the present invention, when the data generation module 105 generates replenishment data based on the demand error value and the out-of-stock coefficient, it is specifically used to: Calculate initial replenishment data according to the preset ratio parameter and the out-of-stock coefficient; The replenishment data is confirmed according to the magnitude relationship between the demand error value and a preset error threshold value and the initial replenishment data.

[0071] In detail, the step of confirming the replenishment data according to the magnitude relationship between the demand error value and the preset error threshold and the initial replenishment data is performed by judging whether the demand error value is less than the preset error threshold, and if the demand error value is greater than or equal to the error threshold, correcting the material demand data according to the error threshold, recalculating the stock-out coefficient according to the corrected material demand data, and returning to the step of calculating the initial replenishment data according to the preset ratio parameter and the stock-out coefficient, and if the demand error value is less than the error threshold, confirming the initial replenishment data as the final replenishment data.

[0072] In the embodiment of the present invention, by real-time connection with warehouse management systems at all levels, data such as the type, quantity, inventory location, and entry and exit time of different types of materials in multi-level warehouses such as central warehouses, regional warehouses, and emergency vehicle warehouses are obtained. Based on these data, the material flow trajectory is tracked, and material path data including the material flow time is obtained. The multi-level warehouse data and material path data are integrated to present the spatial distribution and flow characteristics of the materials in the form of a mathematical matrix. The rows of the matrix represent warehouses or storage nodes, and the columns represent material-related characteristics. The generation process includes obtaining warehouse location data to construct a map, combining material path duration data to construct a warehouse connection map, and then graying, downsampling and numerically processing the map in turn to obtain a spatial feature matrix, extracting the periodicity and trend characteristics of historical material consumption data, and performing tensor splicing with the spatial feature matrix. The pre-trained spatiotemporal graph convolutional network model is used to predict the initial demand data within the first preset duration, and then the material demand data is obtained based on the preset loss weight adjustment. The inventory and demand of warehouses at all levels are obtained, and the out-of-stock coefficient is calculated using a specific formula in combination with the total inventory, the total material demand and the spatial feature matrix. In the second preset duration that is less than the first preset duration and has the same time starting point, the real consumption data and predicted demand data are obtained. The mean is calculated by dividing the time window, and the demand error value is calculated using a specific formula to determine the accuracy of the predicted data, ensure the accuracy of the subsequent replenishment data, and calculate the initial replenishment data according to the preset ratio parameter and the out-of-stock coefficient. If the demand error value is greater than or equal to the preset threshold, the material demand data is corrected and the out-of-stock coefficient is recalculated, and the initial replenishment data is calculated again; if it is less than the threshold, the initial replenishment data is determined to be the final replenishment data. It ensures the efficiency of material replenishment in medical rescue scenarios.

[0073] Reference Figure 2 FIG. 1 is a flow chart of a material management method for a medical rescue scenario provided by an embodiment of the present invention. In this embodiment, the material management method for a medical rescue scenario includes: S1. Acquire multi-level warehouse data of materials, and acquire material path data according to the multi-level warehouse data.

[0074] In the embodiment of the present invention, the multi-level warehouse data may refer to material-related data of warehouses of different levels, and the warehouses of different levels may include a central warehouse, a regional warehouse, and an emergency vehicle warehouse.

[0075] In the embodiment of the present invention, the multi-level warehouse data of materials is obtained by real-time data connection with the management system of each level of warehouses to collect relevant information of various materials, including the type, quantity, inventory location, entry and exit time of the materials, etc. These data reflect the real-time status of the materials in each warehouse and provide a basis for subsequent analysis.

[0076] In the embodiment of the present invention, the material path data is obtained according to the multi-level warehouse data, and the flow trajectory of the materials between different warehouses is analyzed based on the obtained multi-level warehouse data. By tracking the in-and-out records of the materials, it can be determined which intermediate warehouses the materials pass through from the starting warehouse to the final destination of the materials. This information constitutes the material path data, which helps to understand the material circulation process, clarify the allocation relationship between materials in different warehouses, and provide a basis for subsequent analysis.

[0077] In detail, the material path data includes data on the time it takes for materials to be shipped from one warehouse to another warehouse.

[0078] In the embodiment of the present invention, by acquiring multi-level warehouse data of materials and acquiring material path data according to the multi-level warehouse data, the efficiency of subsequent generation of a spatial feature matrix can be improved.

[0079] S2. Generate a spatial feature matrix based on the multi-level warehouse data and the material path data.

[0080] In the embodiment of the present invention, the spatial characteristic matrix is ​​generated according to the multi-level warehouse data and the material path data, which presents the complex material spatial distribution and flow information in the form of a mathematical matrix for subsequent data analysis and processing. This matrix integrates the multi-level warehouse data and the material path data, and can fully reflect the spatial characteristics of the material.

[0081] In detail, the rows of the matrix can represent different warehouses or material storage nodes, and the columns can represent various material-related features. For example, a row corresponds to a specific warehouse, and the elements in the column may include the inventory quantity of different types of materials in the warehouse, the connection relationship between the warehouse and other warehouses (reflecting the material flow path), the geographical location information of the warehouse, etc. In this way, the spatial distribution, quantity information, and flow path of materials are integrated into a matrix, providing a unified data structure for subsequent material demand forecasting and analysis.

[0082] In the embodiment of the present invention, generating a spatial feature matrix according to the multi-level warehouse data and the material path data includes: Constructing a warehouse location map based on the multi-level warehouse data; Constructing a warehouse connection map according to the material path data and the warehouse location map; Feature extraction is performed on the warehouse connection graph to obtain a spatial feature matrix.

[0083] In detail, constructing a warehouse location map based on the multi-level warehouse data is accomplished by acquiring the warehouse location data contained in the multi-level warehouse data and constructing a warehouse location map based on the warehouse location data.

[0084] In detail, the feature extraction of the warehouse connection map to obtain the spatial feature matrix is ​​carried out by graying the warehouse connection map to obtain a grayed image, downsampling the grayed image to obtain a downsampled image, and digitizing the downsampled image to obtain the spatial feature matrix.

[0085] In detail, the warehouse location data includes the latitude and longitude data of the central warehouse, the regional warehouse, and the emergency vehicle warehouse.

[0086] In detail, constructing a warehouse location map based on the warehouse location data is to draw an image containing each warehouse location according to a preset scale based on the relative positions between warehouses of different levels.

[0087] In detail, the grayscale processing of the warehouse connection map to obtain the grayscale image is to convert each pixel of the image from a plurality of color channels (such as RGB) into a single grayscale value. The grayscale value is usually an integer between 0 and 255, indicating the brightness of the pixel.

[0088] In detail, downsampling the grayscale image to obtain a downsampled image may refer to dividing the grayscale image into multiple pixel blocks of the same size, taking the pixel mean of each pixel block as the feature value of the pixel block, and summarizing all the feature values ​​to obtain a new image, which has a smaller size but higher feature expression capability.

[0089] In detail, the numerical processing of the downsampled image refers to converting each pixel point according to the downsampled image into a numerical value according to its gray value.

[0090] In detail, constructing a warehouse connection map according to the material path data and the warehouse location map includes: Generate an adaptive path curve of corresponding proportion according to the material path data; Obtaining the path connection relationship data contained in the material path duration data; The adaptive path curve is drawn into the warehouse location map based on the path connection relationship data to obtain a warehouse connection map.

[0091] In detail, the method of generating an adaptive path curve of corresponding proportion according to the material path data is performed by acquiring the material path duration data contained in the material path data, and generating an adaptive path curve of corresponding proportion based on each path duration in the material path duration data.

[0092] In detail, the material path duration data includes the duration data of each path.

[0093] In an embodiment of the present invention, since the road conditions in the medical rescue scenario may be more complicated, the corresponding path data is also more complicated. For example, there are two regional warehouses with the same straight-line distance from the central warehouse, but the actual paths for transporting materials may be different, which can be reflected in the different path passing time in this solution.

[0094] In detail, the path connection relationship data included in the material path duration data is obtained by obtaining the connection relationship between the central warehouse and multiple regional warehouses, and the connection relationship between each regional warehouse and multiple emergency vehicle warehouses.

[0095] In detail, the step of drawing the adaptive path curve into the warehouse location map based on the path connection relationship data to obtain the warehouse connection map means that the warehouse location map is used as the base map, and the adaptive path curve is drawn between the corresponding warehouse locations according to the connection relationship between warehouses determined by the path connection relationship data. In this way, by connecting each warehouse with a curve, a warehouse connection map is formed. In this map, not only can the location distribution of the warehouses be seen, but also the flow path of materials between different warehouses and the time cost difference of the path can be intuitively understood through the shape and connection method of the curve.

[0096] In detail, the adaptive path curve refers to a curve that can adaptively adjust the curvature according to the ratio of the straight-line distance between two warehouses and the path passing time, so as to realize the connection between the two warehouses. For example, when the straight-line distance between the two warehouses is fixed, the longer the actual passing time for materials to complete the entry and exit of the two warehouses, the longer the corresponding curve length. The curve can adaptively adjust the curvature of the curve to adapt to the difference between the straight-line distance and the passing time.

[0097] In the embodiment of the present invention, by generating a spatial feature matrix according to the multi-level warehouse data and the material path data, the accuracy of subsequent prediction of material demand data is improved.

[0098] S3. Predicting material demand data within a first preset time period based on historical data of material consumption and the spatial feature matrix.

[0099] In the embodiment of the present invention, the predicting of material demand data within a first preset time period based on the historical data of material consumption and the spatial feature matrix includes: Generate a splicing feature according to the historical data and the spatial feature matrix; Using a pre-trained material consumption prediction model to predict demand data within a first preset time period according to the splicing features, to obtain initial demand data; The initial demand data is adjusted based on a preset loss weight to obtain material demand data.

[0100] In detail, the splicing features are generated based on the historical data and the spatial feature matrix by extracting the periodic features and trend features of the historical data, and performing tensor splicing of the periodic features and the trend features with the spatial feature matrix to obtain the splicing features.

[0101] In detail, the material consumption prediction model is a spatiotemporal graph convolutional network model, which is obtained by obtaining a set of material consumption history data and dividing it into training data and test data; The periodic characteristics and trend characteristics of the training data are extracted, and the spatial characteristics of the training data are obtained. After splicing the extracted features, the spatiotemporal graph convolutional network model is used to predict the future consumption data of materials based on the spliced ​​features. The mean square error function is used as the loss function to calculate the loss value of the future consumption data and the test data, and the parameters of the optimization model are adjusted according to the loss value. When the loss value converges or reaches the preset number of iterations, the model training is confirmed to be completed.

[0102] In detail, the extraction of the periodic characteristics and trend characteristics of the historical data can be based on the periodic decomposition of the time series of the historical data to obtain the periodic characteristics, and the calculation of the average value of the historical data within a certain time window can smooth the data and extract trend characteristics.

[0103] In detail, in the field of mathematics and computer science, a tensor is an array structure that can represent data of different dimensions. The tensor splicing is to merge multiple tensors according to a specific dimension to form a new tensor. The tensor splicing of the periodic feature and the trend feature with the spatial feature matrix to obtain the spliced ​​feature is to merge the tensors representing the periodic feature and the trend feature with the spatial feature matrix according to a preset rule.

[0104] In detail, the initial demand data is adjusted based on the preset loss weight to obtain the material demand data, which means assigning higher weights to material demand nodes with higher importance. For example, the importance of the emergency vehicle warehouse requires adjusting the material demand data of the emergency vehicle warehouse based on a higher weight to ensure sufficient supply of materials at the emergency site.

[0105] In the embodiment of the present invention, the time span of the historical data needs to be greater than the first preset time length.

[0106] In the embodiment of the present invention, by predicting the material demand data within the first preset time period based on the historical data of material consumption and the spatial feature matrix, the gap in material demand can be calculated first, so as to facilitate immediate replenishment.

[0107] S4. Calculate the out-of-stock coefficient according to the multi-level warehouse data, the material demand data and the spatial feature matrix.

[0108] In the embodiment of the present invention, the out-of-stock coefficient is a coefficient used to represent the out-of-stock quantity of a certain material, and is a key parameter used for subsequent replenishment data.

[0109] In the embodiment of the present invention, the calculating of the stock-out coefficient of the material according to the material demand data and the spatial characteristic matrix includes: Obtaining the inventory of each level of warehouse in the multi-level warehouse data, and obtaining the warehouse demand of each level of warehouse in the material demand data; Calculate the total inventory and total material demand based on the multi-level warehouse data and material demand data; The logarithmic term is obtained by performing logarithmic operation on the warehouse demand and the total inventory; Perform multiplication and division operations on the element mean of the spatial feature matrix, the warehouse demand, and the inventory to obtain a ratio term; The logarithmic term and the ratio term are summed and then multiplied by a preset weight parameter to obtain a comprehensive weight term; Calculate the ratio of the total material demand to the total inventory and then multiply it by a preset supply-demand ratio weight parameter to obtain a supply-demand ratio item; The comprehensive weight item is added to the supply-demand ratio item to obtain the warehouse out-of-stock coefficient of each level of warehouse, and the warehouse out-of-stock coefficients of warehouses of all levels are added to obtain the out-of-stock coefficient.

[0110] In detail, the calculation formula of the out-of-stock coefficient is as follows: in, is the out-of-stock coefficient, is the warehouse level, For the The weight coefficient of the level warehouse, The material demand data The warehouse demand of the first-level warehouse, For multi-level warehouse data The inventory level of the warehouse, is the total number of elements in the spatial feature matrix, is the spatial feature matrix The value of the element, To preset a very small number, is the supply-demand ratio weight parameter, is the total material demand in the material demand data, It is the total inventory in the multi-level warehouse data.

[0111] In the embodiment of the present invention, the number of warehouse levels may be 3. A level warehouse can refer to a level 1 warehouse, a level 2 warehouse, or a level 3 warehouse.

[0112] In detail, the first-level warehouse may refer to a central warehouse, the second-level warehouse may refer to a regional warehouse, and the third-level warehouse may refer to an emergency vehicle warehouse.

[0113] S5. Calculate a demand error value according to actual consumption data of materials within a second preset time period and predicted demand data included in the material demand data within the second preset time period.

[0114] In an embodiment of the present invention, the second preset time length is less than the preset first time length, and the second preset time length has the same starting point as the first preset time length. For example, the preset first time length is from 8 a.m. to 8 p.m. on the same day, and the preset second time length may be from 8 a.m. to 12 noon on the same day.

[0115] In an embodiment of the present invention, the demand error value is calculated based on the actual consumption data of materials within the second preset time period and the predicted demand data contained in the material demand data within the second preset time period. This is to judge the prediction accuracy of the predicted data by comparing the predicted data with the actual data in a short period of time, which can improve the efficiency of judgment and detect prediction errors as early as possible.

[0116] In the embodiment of the present invention, the calculation of the demand error value according to the actual consumption data of the material within the second preset time period and the predicted demand data contained in the material demand data within the second preset time period includes: Dividing the actual consumption data and the predicted demand data into time windows to obtain actual window data and predicted window data; Calculate the mean of each time window in the real window data and the predicted window data to obtain real window mean data and predicted window mean data; A demand error value is calculated according to the real window data, the predicted window data, the real window mean data and the predicted window mean data.

[0117] In detail, the calculating of the demand error value according to the real window data, the predicted window data, the real window mean data and the predicted window mean data includes: The required error value is calculated using the following formula: in, is the required error value, is the number of windows, represents the hyperbolic tangent function, Indicates the real window mean data The mean data of the windows, Indicates the first The mean data of the windows, represents the number of samples in a single window, Indicates the real window data In the window data, Indicates the prediction window data In the window data, is the preset confidence deviation penalty coefficient, is the base of natural logarithms.

[0118] In detail, the hyperbolic tangent function can realize normalization of the value of the input function. In detail, the confidence deviation penalty coefficient may be 1.2.

[0119] In an embodiment of the present invention, by calculating the demand error value based on the actual consumption data of materials within the second preset time period and the predicted demand data contained in the material demand data within the second preset time period, it is possible to promptly determine whether the predicted demand data is accurate, thereby ensuring the accuracy of subsequent calculations of replenishment data.

[0120] S6. Generate replenishment data based on the demand error value and the out-of-stock coefficient.

[0121] In the embodiment of the present invention, the generating replenishment data based on the demand error value and the out-of-stock coefficient includes: Calculate initial replenishment data according to the preset ratio parameter and the out-of-stock coefficient; The replenishment data is confirmed according to the magnitude relationship between the demand error value and a preset error threshold value and the initial replenishment data.

[0122] In detail, the step of confirming the replenishment data according to the magnitude relationship between the demand error value and the preset error threshold and the initial replenishment data is performed by judging whether the demand error value is less than the preset error threshold, and if the demand error value is greater than or equal to the error threshold, correcting the material demand data according to the error threshold, recalculating the stock-out coefficient according to the corrected material demand data, and returning to the step of calculating the initial replenishment data according to the preset ratio parameter and the stock-out coefficient, and if the demand error value is less than the error threshold, confirming the initial replenishment data as the final replenishment data.

[0123] In the embodiment of the present invention, by real-time connection with warehouse management systems at all levels, data such as the type, quantity, inventory location, and entry and exit time of different types of materials in multi-level warehouses such as central warehouses, regional warehouses, and emergency vehicle warehouses are obtained. Based on these data, the material flow trajectory is tracked, and material path data including the material flow time is obtained. The multi-level warehouse data and material path data are integrated to present the spatial distribution and flow characteristics of the materials in the form of a mathematical matrix. The rows of the matrix represent warehouses or storage nodes, and the columns represent material-related characteristics. The generation process includes obtaining warehouse location data to construct a map, combining material path duration data to construct a warehouse connection map, and then graying, downsampling and numerically processing the map in turn to obtain a spatial feature matrix, extracting the periodicity and trend characteristics of historical material consumption data, and performing tensor splicing with the spatial feature matrix. The pre-trained spatiotemporal graph convolutional network model is used to predict the initial demand data within the first preset duration, and then the material demand data is obtained based on the preset loss weight adjustment. The inventory and demand of warehouses at all levels are obtained, and the out-of-stock coefficient is calculated using a specific formula in combination with the total inventory, the total material demand and the spatial feature matrix. In the second preset duration that is less than the first preset duration and has the same time starting point, the real consumption data and predicted demand data are obtained. The mean is calculated by dividing the time window, and the demand error value is calculated using a specific formula to determine the accuracy of the predicted data, ensure the accuracy of the subsequent replenishment data, and calculate the initial replenishment data according to the preset ratio parameter and the out-of-stock coefficient. If the demand error value is greater than or equal to the preset threshold, the material demand data is corrected and the out-of-stock coefficient is recalculated, and the initial replenishment data is calculated again; if it is less than the threshold, the initial replenishment data is determined to be the final replenishment data. It ensures the efficiency of material replenishment in medical rescue scenarios.

[0124] In the embodiments provided by the present invention, it should be understood that the disclosed devices, systems and methods can be implemented in other ways. For example, the system embodiments described above are only schematic, for example, the division of the modules is only a logical function division, and there may be other division methods in actual implementation.

[0125] The modules described as separate components may or may not be physically separated, and the components shown as modules may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0126] In addition, each functional module in each embodiment of the present invention may be integrated into one processing unit, each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of hardware plus software functional modules.

[0127] It is obvious to those skilled in the art that the present invention is not limited to the details of the above exemplary embodiments, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0128] Therefore, no matter from which point of view, the embodiments should be regarded as illustrative and non-restrictive, and the scope of the present invention is limited by the appended claims rather than the above description, so it is intended that all changes falling within the meaning and scope of the equivalent elements of the claims are included in the present invention. Any attached figure mark in the claims should not be regarded as limiting the claims involved.

[0129] The embodiments of the present application can acquire and process relevant data based on artificial intelligence technology. Artificial Intelligence (AI) is the theory, method, technology and application system that uses digital computers or machines controlled by digital computers to simulate, extend and expand human intelligence, perceive the environment, acquire knowledge and use knowledge to obtain the best results.

[0130] In addition, it is clear that the word "comprising" does not exclude other units or steps, and the singular does not exclude the plural. Multiple units or systems stated in a system claim can also be implemented by one unit or system through software or hardware. The words first, second, etc. are used to indicate names, and do not indicate any particular order.

[0131] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solution of the present invention can be modified or replaced by equivalents without departing from the spirit and scope of the technical solution of the present invention.

Claims

1. A material management system for medical rescue scenarios, characterized in that: It includes a data acquisition module, a feature generation module, a data prediction module, a data calculation module and a data generation module, among which: The data acquisition module is used to acquire multi-level warehouse data of materials, and acquire material path data according to the multi-level warehouse data; The feature generation module is used to generate a spatial feature matrix according to the multi-level warehouse data and the material path data; The data prediction module is used to predict material demand data within a first preset time period based on historical data of material consumption and the spatial feature matrix; The data calculation module is used to calculate the out-of-stock coefficient according to the multi-level warehouse data, the material demand data and the spatial feature matrix, and calculate the demand error value according to the actual consumption data of the materials within the second preset time period and the predicted demand data contained in the material demand data within the second preset time period; The data generation module is used to generate replenishment data based on the demand error value and the out-of-stock coefficient.

2. The material management system for medical rescue scenarios according to claim 1, characterized in that: When the feature generation module generates a spatial feature matrix according to the multi-level warehouse data and the material path data, it is specifically used to: Constructing a warehouse location map based on the multi-level warehouse data; Constructing a warehouse connection map according to the material path data and the warehouse location map; Feature extraction is performed on the warehouse connection graph to obtain a spatial feature matrix.

3. The material management system for medical rescue scenarios according to claim 2, characterized in that: When the feature generation module constructs the warehouse connection map according to the material path data and the warehouse location map, it is specifically used to: Generate an adaptive path curve of corresponding proportion according to the material path data; Obtaining the path connection relationship data contained in the material path duration data; The adaptive path curve is drawn into the warehouse location map based on the path connection relationship data to obtain a warehouse connection map.

4. The material management system for medical rescue scenarios according to claim 1, characterized in that: When the data prediction module predicts the material demand data within the first preset time period according to the historical data of material consumption and the spatial feature matrix, it is specifically used to: Generate a splicing feature according to the historical data and the spatial feature matrix; Using a pre-trained material consumption prediction model to predict demand data within a first preset time period according to the splicing features, to obtain initial demand data; The initial demand data is adjusted based on a preset loss weight to obtain material demand data.

5. The material management system for medical rescue scenarios according to claim 1, characterized in that: When the data calculation module calculates the stock-out coefficient of a material according to the material demand data and the spatial characteristic matrix, it is specifically used to: Obtaining the inventory of each level of warehouse in the multi-level warehouse data, and obtaining the warehouse demand of each level of warehouse in the material demand data; Calculate the total inventory and total material demand based on the multi-level warehouse data and material demand data; The logarithmic term is obtained by performing logarithmic operation on the warehouse demand and the total inventory; Perform multiplication and division operations on the element mean of the spatial feature matrix, the warehouse demand, and the inventory to obtain a ratio term; The logarithmic term and the ratio term are summed and then multiplied by a preset weight parameter to obtain a comprehensive weight term; Calculating the ratio of the total material demand to the total inventory and then multiplying it by a preset supply-demand ratio weight parameter to obtain a supply-demand ratio item; The comprehensive weight item is added to the supply-demand ratio item to obtain the warehouse out-of-stock coefficient of each level of warehouse, and the warehouse out-of-stock coefficients of warehouses of all levels are added to obtain the out-of-stock coefficient.

6. The material management system for medical rescue scenarios according to claim 1, characterized in that: In the data calculation module, the calculation formula of the out-of-stock coefficient is as follows: in, is the out-of-stock coefficient, is the warehouse level, For the The weight coefficient of the level warehouse, The material demand data The warehouse demand of the first-level warehouse, For multi-level warehouse data The inventory level of the warehouse, is the total number of elements in the spatial feature matrix, is the spatial feature matrix The value of the element, To preset a very small number, is the supply-demand ratio weight parameter, is the total material demand in the material demand data, It is the total inventory in the multi-level warehouse data.

7. The material management system for medical rescue scenarios according to claim 1, characterized in that: When the data calculation module calculates the demand error value according to the actual consumption data of the material within the second preset time period and the predicted demand data contained in the material demand data within the second preset time period, it is specifically used to: Dividing the actual consumption data and the predicted demand data into time windows to obtain actual window data and predicted window data; Calculate the mean of each time window in the real window data and the predicted window data to obtain real window mean data and predicted window mean data; A demand error value is calculated according to the real window data, the predicted window data, the real window mean data and the predicted window mean data.

8. The material management system for medical rescue scenarios according to claim 7, characterized in that: In the data calculation module, the calculation formula of the required error value is as follows: in, is the required error value, is the number of windows, represents the hyperbolic tangent function, Indicates the real window mean data The mean data of the windows, Indicates the first The mean data of the windows, represents the number of samples in a single window, Indicates the real window data In the window data, Indicates the prediction window data In the window data, is the preset confidence deviation penalty coefficient, is the base of natural logarithms.

9. The material management system for medical rescue scenarios according to claim 1, characterized in that: When the data generation module generates replenishment data based on the demand error value and the out-of-stock coefficient, it is specifically used to: Calculate initial replenishment data according to the preset ratio parameter and the out-of-stock coefficient; The replenishment data is confirmed according to the magnitude relationship between the demand error value and a preset error threshold value and the initial replenishment data.

10. A material management method for medical rescue scenarios, characterized in that: The method comprises: Acquire multi-level warehouse data of materials, and acquire material path data according to the multi-level warehouse data; Generate a spatial feature matrix according to the multi-level warehouse data and the material path data; Predicting material demand data within a first preset time period based on historical data of material consumption and the spatial feature matrix; Calculate the out-of-stock coefficient according to the multi-level warehouse data, the material demand data and the spatial feature matrix; Calculating a demand error value according to actual consumption data of materials within a second preset time period and predicted demand data included in the material demand data within the second preset time period; Replenishment data is generated based on the demand error value and the out-of-stock factor.

Images