A dynamic early warning method and system for medical supplies based on dynamic scheduling

Abstract

The present invention relates to the field of data processing, and specifically to a dynamic early warning method and system for medical supplies based on dynamic scheduling, which has the main purpose of reducing scheduling planning, and obtains medical supplies with usage associations by analyzing historical usage speed information. Then, when analyzing any kind of medical supplies, if it is found that the usage speed of the medical supplies has increased in a short period of time, it means that the usage speed of the medical supplies with usage associations may also increase. Therefore, a pre-constructed statistical model is used to predict the usage speed at future time points, and the associated medical supplies are also predicted and warned. This avoids the situation where the inventory of multiple medical supplies triggers the warning line multiple times in a row within a period of time, thereby helping relevant personnel to make reasonable scheduling plans.

Description

Technical Field

[0001] The present invention relates to the field of data processing, and in particular to a dynamic early warning method and system for medical supplies based on dynamic scheduling. Background Art

[0002] Inventory management of medical supplies is crucial to the effective operation of medical institutions, and its significance is far-reaching and multifaceted. Effective inventory management can ensure the continuity of medical services, improve emergency response capabilities, optimize resource utilization and cost control, etc. Therefore, a sound medical supplies inventory management system is very important for medical institutions.

[0003] Existing inventory management is mainly based on manual inventory of warehouses, and then the maintenance time of medical supplies is estimated based on the daily usage speed, so as to realize the early warning and scheduling planning of materials. However, different types of medical substances actually have usage associations, such as sutures, disinfectants, anesthetics and other surgical supplies. In the existing inventory warning, generally only a separate inventory warning is carried out for each medical material, resulting in scheduling planning can only be targeted at a single medical material. As a result, in some cases, the inventory of multiple medical supplies will trigger the warning line multiple times in a row within a period of time, requiring frequent scheduling planning, resulting in unreasonable scheduling planning and low efficiency. Summary of the invention

[0004] In view of this, an object of the present invention is to provide a dynamic early warning method and system for medical supplies based on dynamic scheduling to solve the problems in the background technology.

[0005] In order to achieve the above object, the present invention adopts the following technical solutions:

[0006] A dynamic early warning method for medical supplies based on dynamic scheduling of the present invention comprises the following steps:

[0007] Obtaining historical usage speed information of multiple medical supplies in multiple time periods, and obtaining current usage speed information and inventory of multiple medical supplies in a current time period, wherein the historical usage speed information includes usage speeds at multiple historical time points, and the current usage speed information includes usage speeds at multiple current time points;

[0008] Performing correlation analysis based on historical usage speed information of multiple medical supplies in multiple time periods to obtain a material usage speed correlation map, wherein the material usage speed correlation map includes a correlation relationship between the usage speeds of any two medical supplies;

[0009] Extracting the change characteristics in the current usage speed information, and taking the medical supplies whose change characteristics satisfy the unidirectional change as the target medical supplies, and predicting the usage speed of the target medical supplies and the related medical supplies as well as the predicted usage speed at a future time point based on the pre-built statistical model, the change characteristics of the target medical supplies, and the change characteristics of the related materials of the target medical supplies;

[0010] An early warning is issued based on the predicted usage speed of the target medical supplies at a future time point, the current inventory of the target medical supplies, the predicted usage speed of the related supplies at a future time point, and the current inventory of the related supplies.

[0011] In one embodiment of the present application, a correlation analysis is performed based on historical usage speed information of multiple medical supplies in multiple time periods to obtain a material usage speed correlation map, including:

[0012] Conduct regularity analysis on the historical usage speed information of each medical material in multiple time periods to obtain regular medical materials with regular characteristics;

[0013] Based on the regular characteristics, correlation analysis is performed on multiple regular medical supplies to obtain multiple groups of medical supplies with correlation characteristics, wherein the correlation characteristics include positive correlation and negative correlation;

[0014] A material usage speed correlation map is constructed based on all groups of medical materials with correlation characteristics.

[0015] In one embodiment of the present application, the historical usage speed information of each medical material in multiple time periods is analyzed for regularity, and regular medical materials with regular characteristics are obtained, including:

[0016] The usage speed of each medical material at multiple historical time points in each time period is curve fitted to obtain the usage speed fluctuation curve of each medical material in multiple time periods, wherein the usage speed fluctuation curve is located at In a two-dimensional coordinate system;

[0017] Smoothing the usage speed fluctuation curve based on a sliding average to obtain a usage speed smooth curve;

[0018] Calculate the average speed of the speed fluctuation curve and using speed variance , and construct a first usage speed mean reference line based on the usage speed average and the usage speed variance and the second use speed mean reference line ,in, To adjust the parameters;

[0019] Extract the first usage speed mean reference line in the usage speed fluctuation curve The part above the speed fluctuation curve is obtained to obtain the peak section; and the second speed average reference line is extracted from the speed fluctuation curve. The lower part gets the trough section;

[0020] Extract the upper inflection point of the peak segment and the lower inflection point in the trough section ;

[0021] For all upper inflection points of each medical material Density-based clustering is performed to obtain multiple upper inflection point clusters; and all lower inflection points of each medical material are clustered. Performing density-based clustering to obtain a plurality of lower inflection point clusters; wherein the maximum intra-cluster distances between the upper inflection point cluster and the lower inflection point cluster are preset;

[0022] The upper inflection point cluster or the lower inflection point cluster whose number of data points in the cluster is greater than the preset number threshold is taken as the target cluster, and the time range of the data points in the target cluster is calculated and speed range , obtain regular characteristics; and take the medical supplies with regular characteristics as regular medical supplies, where, is the time average of the data points in the target cluster, is the time standard deviation of the data points in the target cluster, is the average usage speed of the data points in the target cluster, is the standard deviation of the speed of the data points in the target cluster, is the range adjustment factor.

[0023] In one embodiment of the present application, a correlation analysis is performed on a plurality of regular medical supplies based on the regular characteristics to obtain a plurality of groups of medical supplies with correlation characteristics, including:

[0024] The time period is divided into a plurality of time periods, and the plurality of time periods are assigned values ​​based on the regular characteristics of each regular medical resource to obtain a regular code for each regular medical resource, wherein the value of each bit of the regular code is positively correlated with the usage speed range of the position corresponding to the regular characteristic;

[0025] The regular codes of any two regular medical supplies are subjected to bitwise difference and absolute value calculation to obtain the difference code;

[0026] Calculating the sum of all values ​​in the difference code, and comparing the sum of all values ​​in the difference code with a pre-constructed first screening threshold and a second screening threshold, wherein the first screening threshold is greater than the second screening threshold;

[0027] When the sum of all values ​​in the difference code is less than the second screening threshold, it is determined that the two regular medical supplies are positively associated; when the sum of all values ​​in the difference code is greater than the first screening threshold, it is determined that the two regular medical supplies are negatively associated; otherwise, it is determined that the two regular medical supplies are not associated;

[0028] Based on the association relationship between any two regular medical supplies, multiple groups of medical supplies with associated characteristics are obtained.

[0029] In an embodiment of the present application, extracting the change feature in the current usage speed information includes: mapping the usage speeds at multiple current time points to a two-dimensional coordinate system, wherein the two-dimensional coordinate system includes a time axis and a usage speed axis;

[0030] Slide along the time axis based on the pre-built sliding window, and calculate the average value of all data points in the sliding window at each slide ,in, is the number of sliding steps;

[0031] Determine whether multiple averages meet and ,or, and ; If satisfied, it is determined that there is a unidirectional change trend and the change slope is calculated , ,in, For the The average value of sliding calculations, is the average value of the first calculation, The average value calculated from the last slide, For the The average value of sliding calculations, is the time interval between any two slides.

[0032] In one embodiment of the present application, the process of constructing the statistical model includes:

[0033] The usage speed of each medical material at multiple historical time points in each time period is curve fitted to obtain the usage speed fluctuation curve of each medical material in multiple time periods, wherein the usage speed fluctuation curve is located at In a two-dimensional coordinate system;

[0034] Smoothing the usage speed fluctuation curve based on a sliding average to obtain a usage speed smooth curve;

[0035] Use a sliding window along the two-dimensional coordinate system The axis slides, and at each slide, the historical change characteristics of the data points in the sliding window are extracted;

[0036] When the historical change feature satisfies a unidirectional change, the slope in the change feature is extracted, and the first inflection point sample after the historical change feature is extracted , inflection point samples A plurality of sampling points between the current position center of the sliding window and the inflection point sample, and a time interval between the current position center of the sliding window and the inflection point sample , wherein the inflection point sample is an upper inflection point or a lower inflection point, is the time range parameter, and the time interval between adjacent sampling points is fixed;

[0037] Based on slope, multiple sampling points, inflection point samples and time interval Construct sample data to obtain sample data set; and based on the usage speed in the inflection point sample and time interval Cluster the sample data set to obtain multiple sample clusters;

[0038] Calculate the slope mean of the sample data in the target sample cluster whose sample data volume is greater than the preset number threshold , slope variance ; and in the slope variance When the variance is greater than the preset threshold, the sample cluster with the slope mean is removed. Sample data whose deviation rate is greater than the deviation rate threshold, and return to calculate the slope mean of the sample data in the target sample cluster , slope variance , until the slope variance When the variance is not greater than the preset threshold and the amount of sample data in the target sample cluster is greater than the preset threshold, the slope mean in the target sample cluster is used as the Construct a slope range that satisfies 3 times the standard deviation and calculate the average inflection point speed corresponding to the slope range and the average time interval ;

[0039] Aligning sampling points of different sample data in the target sample cluster, and calculating the average value of the aligned sampling points to obtain multiple reference sampling points;

[0040] Construct multiple reference sampling points and inflection point samples corresponding to multiple slope ranges Constructing a speed change reference curve; and constructing a statistical model based on the speed change reference curves in multiple slope ranges.

[0041] In one embodiment of the present application, based on a pre-built statistical model, the change characteristics of the target medical supplies, and the change characteristics of the associated materials of the target medical supplies, the target medical supplies and the associated medical supplies are predicted to have a predicted usage speed at a future time point, including:

[0042] Substitute the slope of the target medical material or the slope of the associated medical material into the statistical model to obtain a corresponding speed change reference curve; and read the predicted usage speed at a future time point from the speed change reference curve.

[0043] In one embodiment of the present application, an early warning is provided based on the predicted usage speed of the target medical supplies at a future time point, the current inventory of the target medical supplies, the predicted usage speed of the associated supplies at a future time point, and the current inventory of the associated supplies, including:

[0044] When the associated medical supplies do not show a unidirectional change feature, the associated medical supplies are marked, and the available time of the target medical supplies is calculated based on the predicted usage speed of the target medical supplies at a future time point and the current inventory of the target medical supplies;

[0045] When there is a one-way change feature in the associated medical supplies, the available time of the target medical supplies is calculated based on the predicted usage speed of the target medical supplies at a future time point and the current inventory of the target medical supplies, and the available time of the associated medical supplies is calculated based on the predicted usage speed of the associated medical supplies at a future time point and the current inventory of the associated medical supplies;

[0046] When the available time is less than the first warning value, a warning message is generated, and the warning message and the marked associated medical supplies are output to the target object.

[0047] In one embodiment of the present application, it also includes:

[0048] When the current inventory of the marked associated medical supplies is less than a second warning value, a warning message is generated, wherein the first warning value is less than the second warning value.

[0049] The present application also provides a dynamic early warning system for medical supplies based on dynamic scheduling, including:

[0050] An acquisition module, used to acquire historical usage speed information of multiple medical supplies in multiple time periods, and to acquire current usage speed information and inventory of multiple medical supplies in a current time period, wherein the historical usage speed information includes usage speeds at multiple historical time points, and the current usage speed information includes usage speeds at multiple current time points;

[0051] A correlation analysis module, used to perform correlation analysis based on historical usage speed information of multiple medical supplies in multiple time periods to obtain a material usage speed correlation map, wherein the material usage speed correlation map includes a correlation relationship between the usage speeds of any two medical supplies;

[0052] A feature extraction module is used to extract the change features in the current usage speed information, and take the medical supplies whose change features satisfy the unidirectional change as the target medical supplies, and predict the usage speed of the target medical supplies and the related medical supplies as well as the predicted usage speed at a future time point based on a pre-built statistical model, the change features of the target medical supplies, and the change features of the related materials of the target medical supplies;

[0053] The early warning module is used to issue an early warning based on the predicted usage speed of the target medical supplies at a future time point, the current inventory of the target medical supplies, the predicted usage speed of the related supplies at a future time point, and the current inventory of the related supplies.

[0054] The beneficial effects of the present invention are as follows: a dynamic early warning method and system for medical supplies based on dynamic scheduling of the present invention, with the main purpose of reducing scheduling planning, obtains medical supplies with usage associations by analyzing historical usage speed information. Then, when analyzing any kind of medical supplies, if it is found that the usage speed of the medical supplies has increased in a short period of time, it means that the usage speed of the medical supplies with associated usage may also increase. Therefore, a pre-built statistical model is used to predict the usage speed at future time points, and the associated medical supplies are also predicted and warned. This avoids the situation where the inventory of multiple medical supplies triggers the warning line multiple times in a row within a period of time, thereby helping relevant personnel to make reasonable scheduling plans. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] The present invention will be further described below in conjunction with the accompanying drawings and embodiments:

[0056] Figure 1 It is a flow chart of a dynamic early warning method for medical supplies based on dynamic scheduling shown in an embodiment of the present application;

[0057] Figure 2 This is a schematic diagram of the analysis process of regular medical supplies in one embodiment of the present application;

[0058] Figure 3 This is a schematic diagram of a material usage speed correlation map in one embodiment of the present application;

[0059] Figure 4 This is a schematic diagram of the target sample cluster screening process in one embodiment of the present application;

[0060] Figure 5 It is a structural diagram of a dynamic early warning system for medical supplies based on dynamic scheduling shown in one embodiment of the present application. DETAILED DESCRIPTION

[0061] The following describes the embodiments of the present invention by specific examples, and those skilled in the art can easily understand other advantages and effects of the present invention from the contents disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that the following embodiments and features in the embodiments can be combined with each other without conflict.

[0062] It should be noted that the illustrations provided in the following embodiments are only schematic illustrations of the basic concept of the present invention, and thus the drawings only show the layers related to the present invention rather than being drawn according to the number, shape and size of the layers in actual implementation. In actual implementation, the type, quantity and proportion of each layer may be changed arbitrarily, and the layer layout may also be more complicated.

[0063] In the following description, numerous details are discussed to provide a more thorough explanation of embodiments of the present invention; however, it is apparent to one skilled in the art that embodiments of the present invention may be practiced without these specific details.

[0064] Figure 1 is a flow chart of a dynamic early warning method for medical supplies based on dynamic scheduling shown in an embodiment of the present application, such as Figure 1 As shown, a dynamic early warning method for medical supplies based on dynamic scheduling in this embodiment may include the following steps:

[0065] S110, obtaining historical usage speed information of multiple medical supplies in multiple time periods, and obtaining current usage speed information and inventory of multiple medical supplies in a current time period, wherein the historical usage speed information includes usage speeds at multiple historical time points, and the current usage speed information includes usage speeds at multiple current time points;

[0066] In this application, historical usage speed information within multiple time periods is extracted from inventory count records. The time period in this application is one year. The actual material consumption is the outbound quantity minus the inbound quantity. Taking masks as an example, the following table shows the inventory record of masks:

[0067]

[0068] As can be seen from the table above, 350 masks were used from 2020-01-01 to 2020-01-04, with an average usage rate of 87.5 / day.

[0069] S120, performing correlation analysis based on historical usage speed information of multiple medical supplies in multiple time periods to obtain a material usage speed correlation map, wherein the material usage speed correlation map includes a correlation relationship between usage speeds of any two medical supplies;

[0070] Different medical supplies are related when used. For example, in addition to surgical instruments, sterile gloves, masks, surgical gowns, disinfectants, anesthetics, infusion bags, etc. are also needed during surgery. These supplies together constitute a complete surgical environment. For example, in long-term care or intensive care, some basic medical supplies such as gauze, bandages, catheters, etc. are often used, and specific drugs and nutritional supplements may be required. In addition, monitoring equipment such as blood pressure monitors and blood glucose meters are also indispensable.

[0071] The usage association between different dependent materials is relatively complex. This application uses big data statistics to construct an association map to find the usage speed relationship between different medical materials. The specific process is as follows:

[0072] S121, performing regularity analysis on historical usage speed information of each medical material in multiple time periods to obtain regular medical materials with regular characteristics;

[0073] Medical supplies that do not have regular usage patterns have complex correlation patterns and are difficult to analyze. Therefore, this application conducts correlation analysis on regular medical supplies that have regular characteristics.

[0074] Under normal circumstances, there will be a high incidence of influenza or other epidemics every year, and the medical supplies used during these periods are relatively fixed. This application defines them as regular medical supplies. Figure 2 This is a schematic diagram of the analysis process of regular medical supplies in one embodiment of the present application. Figure 2 As shown, specifically including the following:

[0075] S1211, performing curve fitting on the usage speed of each medical material at multiple historical time points in each time period to obtain a usage speed fluctuation curve of each medical material in multiple time periods, wherein the usage speed fluctuation curve is located at In a two-dimensional coordinate system;

[0076] In this application, the daily or weekly usage speed of medical supplies is extracted from the inventory change record. Generally, this usage speed is relatively stable. However, during the high-incidence season of epidemics, there will be a peak period. Therefore, this application uses curve fitting to perform curve fitting on the daily or weekly usage speed of medical supplies within a year to obtain a usage speed fluctuation curve.

[0077] The fitting method can be least square fitting, spline curve fitting and other fitting methods.

[0078] S1212, smoothing the usage speed fluctuation curve based on a sliding average to obtain a usage speed smooth curve;

[0079] If daily or weekly data is used, due to the large number of data points, small fluctuations may occur in multiple places on the curve. This application mainly analyzes its overall trend, so smoothing is also required.

[0080] This application adopts sliding smoothing, that is, using a sliding window to slide along the time axis, then calculating the average value of each step, refitting all the average values, and obtaining a smooth curve.

[0081] S1213, calculating the average usage speed of the usage speed fluctuation curve and using speed variance , and construct a first usage speed mean reference line based on the usage speed average and the usage speed variance and the second use speed mean reference line ,in, To adjust the parameters;

[0082] Since the fluctuations of the usage speed fluctuation curves generated by different medical supplies are different, some supplies, such as reusable surgical instruments, have greater fluctuations, while the usage of disposable supplies such as disinfectants and masks is relatively stable. Therefore, in order to accurately extract its fluctuation characteristics instead of extracting small fluctuations in the flat section, this application constructs the first usage speed mean reference line and the second use speed mean reference line , located at the first average speed reference line and the second use speed mean reference line The section between is a flat section. The extraction range of the two reference lines is related to their variance, so as to extract the fluctuation characteristics of the fluctuation curves with different fluctuation characteristics.

[0083] S1214: extracting the first usage speed mean reference line from the usage speed fluctuation curve The part above the speed fluctuation curve is obtained to obtain the peak section; and the second speed average reference line is extracted from the speed fluctuation curve. The lower part gets the trough section;

[0084] First use the speed mean reference line The upper part is the peak section, and the second part uses the speed mean reference line The lower part is the trough section, the first reference line is the speed mean Use the mean speed reference line with the second The section in between is a flat section.

[0085] S1215, extracting the upper inflection point of the peak section and the lower inflection point in the trough section ;

[0086] Inflection point extraction uses a sequential difference method to calculate the size relationship between any three adjacent sampling points. If the value of a middle sampling point is greater than the values ​​of the sampling points on the left and right sides, the middle sampling point is the upper inflection point. ; If the value of a middle sampling point is smaller than the values ​​of the sampling points on the left and right sides, the middle sampling point is the lower inflection point .

[0087] S1216, for all upper inflection points of each medical material Density-based clustering is performed to obtain multiple upper inflection point clusters; and all lower inflection points of each medical material are clustered. Performing density-based clustering to obtain a plurality of lower inflection point clusters; wherein the maximum intra-cluster distances between the upper inflection point cluster and the lower inflection point cluster are preset;

[0088] In this embodiment, all the upper inflection points and all the lower inflection points extracted in the above text are clustered respectively, and the clustering method is the DSCAN clustering algorithm. After the maximum distance within the cluster is set in advance, all the upper inflection points and all the lower inflection points can be clustered respectively. In the obtained cluster, all data points have similarities in time and inflection point values.

[0089] S1217, taking the upper inflection point cluster or the lower inflection point cluster whose number of data points in the cluster is greater than the preset number threshold as the target cluster, and calculating the time range of the data points in the target cluster and speed range , obtain regular characteristics; and take the medical supplies with regular characteristics as regular medical supplies, where, is the time average of the data points in the target cluster, is the time standard deviation of the data points in the target cluster, is the average usage speed of the data points in the target cluster, is the standard deviation of the speed of the data points in the target cluster, is the range adjustment factor.

[0090] If there are enough data points in the clusters obtained by clustering, it means that there are many points that are close in time and inflection point values. This means that the fluctuation curve of the use of medical supplies is within this time range and value range, thus reflecting its regularity.

[0091] This application uses the target clusters obtained by clustering to construct a time range that reflects regular characteristics and speed range .

[0092] S122, performing correlation analysis on a plurality of regular medical supplies based on the regular characteristics to obtain a plurality of groups of medical supplies having correlation characteristics, wherein the correlation characteristics include positive correlation and negative correlation;

[0093] After obtaining medical supplies with regular usage, their relevance is reflected through big data analysis, including:

[0094] S1221, dividing the time period into a plurality of time periods, and assigning values ​​to the plurality of time periods based on the regular characteristics of each regular medical resource, to obtain a regular code for each regular medical resource, wherein the value of each bit of the regular code is positively correlated with the usage speed range of the position corresponding to the regular characteristic;

[0095] The time period is one year, and this application divides the time period into 12 time periods, namely one time period for each month; then the regular characteristics are assigned values ​​corresponding to the 12 time periods to obtain a regular code that reflects the regularity of medical substance use.

[0096] For example, January to March each year is the peak season for influenza, and the usage rate of paracetamol reaches a peak during this period, while it reaches a trough in July and August.

[0097] In the present application, the peak area is assigned a value of 1, the trough area is assigned a value of -1, and the flat area is assigned a value of 0, and the regular code of paracetamol can be obtained, which is: 111000-1-1-1000.

[0098] In addition, if a time period includes both a peak segment and a flat segment, a value can be assigned based on the time proportion. For example, if the peak segment accounts for 80% and the flat segment accounts for 20%, the value is 0.8.

[0099] S1222, performing bitwise difference and absolute value calculation on the regular codes of any two regular medical supplies to obtain a difference code;

[0100] By using regular coding to calculate the difference, the difference code obtained can reflect the correlation between the two medical supplies.

[0101] For example: the regular code for paracetamol is: 111000-1-1-1000;

[0102] The regular code of masks is: 011000000001; that is, they are consumed at a higher rate in February, March, and December every year;

[0103] The difference between the two sets of codes is 100000-1-1-100-1;

[0104] S1223, calculating the sum of all values ​​in the difference code, and comparing the sum of all values ​​in the difference code with a pre-constructed first screening threshold and a second screening threshold, wherein the first screening threshold is greater than the second screening threshold;

[0105] The difference code obtained by the above process calculates the absolute value of each bit. If the regular characteristics of the two are completely consistent, showing a complete positive correlation, the value is 0; if they are completely opposite, showing a complete negative correlation, a larger value will appear.

[0106] Using the above principle, the present application compares the sum of all values ​​in the difference code with the first screening threshold and the second screening threshold to perform a correlation analysis.

[0107] S1224, when the sum of all values ​​in the difference code is less than the second screening threshold, it is determined that the two regular medical supplies are positively associated; when the sum of all values ​​in the difference code is greater than the first screening threshold, it is determined that the two regular medical supplies are negatively associated; otherwise, it is determined that the two regular medical supplies are not associated;

[0108] S1225, constructing a plurality of groups of medical supplies with correlation characteristics based on the correlation relationship between any two regular medical supplies.

[0109] The obtained association relationships are used to construct multiple groups of associated medical supplies.

[0110] S123, constructing a material usage speed correlation map based on all the multiple groups of medical materials with correlation characteristics.

[0111] Finally, with medical supplies as nodes and correlation relationships as connecting lines, we can obtain a correlation map of material usage speed.

[0112] Figure 3 Schematic diagram of a material usage speed correlation map in an embodiment of the present application. The constructed material usage speed correlation map is as follows: Figure 3 As shown, this application mainly seeks a positive correlation with the speed of use of medical supplies.

[0113] S130, extracting the change characteristics in the current usage speed information, and taking the medical supplies whose change characteristics satisfy the unidirectional change as the target medical supplies, and predicting the usage speed of the target medical supplies and the related medical supplies as well as the predicted usage speed at a future time point based on a pre-built statistical model, the change characteristics of the target medical supplies, and the change characteristics of the related materials of the target medical supplies;

[0114] This application uses a sliding window averaging method to extract the change characteristics in the current speed information, including:

[0115] S131, mapping the usage speeds at multiple current time points to a two-dimensional coordinate system, wherein the two-dimensional coordinate system includes a time axis and a usage speed axis;

[0116] S132, sliding along the time axis based on the pre-built sliding window, and calculating the average value of all data points in the sliding window during each sliding. ,in, is the number of sliding steps;

[0117] S133, determine whether multiple average values ​​meet and ,or, and ; If satisfied, it is determined that there is a unidirectional change trend and the change slope is calculated , ,in, For the The average value of sliding calculations, is the average value of the first calculation, The average value calculated from the last slide, For the The average value of sliding calculations, is the time interval between any two slides.

[0118] The overall usage speed level over a period of time can be calculated by using the sliding window to find the average value, and then the trend of change can be determined by using simple sequence comparison and difference comparison.

[0119] Specifically, in this application, the current time period is one week, and the usage of medical supplies for 7 days in a week is collected. Then the width of the sliding window is 3 and the step size is 2. That is to say, the usage speed data of the current time period can be converted into 3 overall levels. , , ;exist ,and When judging one-way increase.

[0120] Through the above analysis, if there is a one-way increasing trend of a medical material, it means that the increase in the use of medical materials may be caused by abnormal events (such as disasters in the region) or the annual flu season. At this time, the increase in the use of related medical materials can be predicted by the current use speed of medical materials.

[0121] Due to the different specific usage of different medical supplies, there may be differences in the time of collection, resulting in a sequence of one-way changes. However, by using the association analysis in this application, it is possible to find the associated medical supplies whose usage speed will increase in time, so as to timely discover the increase in the usage speed of all medical supplies as the basic data for inventory scheduling.

[0122] However, if we only predict the upward trend, we cannot use the data for scheduling planning. Therefore, this application also uses historical data to predict the upward amplitude. The constructed statistical model is used to predict the future predicted usage speed.

[0123] Methods for constructing statistical models include:

[0124] (1) Performing curve fitting on the usage speed of each medical material at multiple historical time points in each time period to obtain a usage speed fluctuation curve of each medical material in multiple time periods, wherein the usage speed fluctuation curve is located at In a two-dimensional coordinate system;

[0125] (2) Smoothing the usage speed fluctuation curve based on a sliding average to obtain a usage speed smooth curve;

[0126] The construction of the speed smoothing curve is consistent with the previous one and will not be repeated here.

[0127] (3) Use a sliding window along the two-dimensional coordinate system The axis slides, and at each slide, the historical change characteristics of the data points in the sliding window are extracted;

[0128] In this embodiment, a sliding window with a width of 7 days is used to perform sliding and extract historical change features. The extraction of historical change features can refer to the extraction process of change features in the previous text.

[0129] (4) When the historical change feature satisfies a unidirectional change, extract the slope of the change feature and extract the first inflection point sample after the historical change feature , inflection point samples A plurality of sampling points between the current position center of the sliding window and the inflection point sample, and a time interval between the current position center of the sliding window and the inflection point sample , wherein the inflection point sample is an upper inflection point or a lower inflection point, is the time range parameter, and the time interval between adjacent sampling points is fixed;

[0130] In the present application, the main principle used is to use the slope of the rising section at the initial stage of formation to predict the highest point of the rising section and the time required to rise to the highest point.

[0131] Specifically, the first inflection point sample after the change feature is the coordinate value of the highest point. Inflection point sample A plurality of sampling points between the current position center of the sliding window are used as data samples of the fitting curve, and the time interval between the current position center of the sliding window and the inflection point sample is the time required to rise to the highest point.

[0132] (5) Based on slope, multiple sampling points, inflection point samples and time interval Construct sample data to obtain sample data set; and based on the usage speed in the inflection point sample and time interval Cluster the sample data set to obtain multiple sample clusters;

[0133] Based on the usage speed in the inflection point sample and time interval Clustering the sample data set can be done using speed and time interval Similar sample data is divided into sample clusters. Since the main feature of the convex part of the curve is the inflection point value and time interval , so the samples with similar shape features of the convex part of the curve can be grouped together through preliminary clustering.

[0134] (6) Calculate the slope mean of the sample data in the target sample cluster whose sample data volume is greater than the preset quantity threshold , slope variance ; and in the slope variance When the variance is greater than the preset threshold, the sample cluster with the slope mean is removed. Sample data whose deviation rate is greater than the deviation rate threshold, and return to calculate the slope mean of the sample data in the target sample cluster , slope variance , until the slope variance When the variance is not greater than the preset threshold and the amount of sample data in the target sample cluster is greater than the preset threshold, the slope mean in the target sample cluster is used as the Construct a slope range that satisfies 3 times the standard deviation and calculate the average inflection point speed corresponding to the slope range and the average time interval ;

[0135] Figure 4 FIG. 1 is a schematic diagram of a target sample cluster screening process in an embodiment of the present application. Figure 4As shown, when clustering is initially completed, the sample cluster obtained may contain some samples whose slopes are different from other samples, mainly because there are sometimes too many materials. In order to screen out such extreme samples, this application sets up a loop and uses the variance to calculate the distribution uniformity of the slope. If the variance is large, it means that the cluster contains extreme samples. Then, after removing the extreme samples, the variance of the slope is calculated again, and the comparison and loop are performed again until the variance of the slope in the cluster is small.

[0136] The obtained clusters are multiple data samples with similar slopes. After calculating the mean and standard deviation, and constructing a slope range that satisfies 3 times the standard deviation, the slope range can be obtained.

[0137] (7) aligning the sampling points of different sample data in the target sample cluster, and calculating the average value of the aligned sampling points to obtain multiple reference sampling points;

[0138] Among the retained samples, the convex shapes are relatively similar and all contain multiple sampling points. The number of sampling points of different samples may be different. The present application is based on the base point for alignment, and the base point can be the midpoint of the position of the sliding window when the unidirectional change feature is first extracted. After alignment, the sampling points in different samples are aligned according to the position, and then the average of each sampling point is calculated to obtain multiple reference sampling points.

[0139] For example:

[0140] Sample A: sampling point A1, sampling point A2, sampling point A3, sampling point A4, sampling point A5;

[0141] Sample B: sampling point B1, sampling point B2, sampling point B3, sampling point B4;

[0142] Sample C: sampling point C1, sampling point C2, sampling point C3, sampling point C4, sampling point C5;

[0143] The sampling samples are: (A1+B1+C1) / 3, (A2+B2+C2) / 3, (A3+B3+C3) / 3, (A4+B4+C4) / 3, (A5+C5) / 2;

[0144] Since the time can be aligned, we only need to calculate the average of the usage speed.

[0145] (8) Construct multiple reference sampling points and inflection point samples corresponding to multiple slope ranges Constructing a speed change reference curve; and constructing a statistical model based on the speed change reference curves in multiple slope ranges.

[0146] Finally, multiple reference sampling points corresponding to each slope range are used to fit the reference curve as the typical reference curve corresponding to the initial slope of the rise. In other words, at the beginning of the epidemic season, the rate of increase in the use of materials at the beginning can be used to evaluate the rate of use of materials in the subsequent process. If the rate of increase is faster, the corresponding peak value will be larger and last longer.

[0147] S140, issuing an early warning based on the predicted usage speed of the target medical supplies at a future time point, the current inventory of the target medical supplies, the predicted usage speed of the related supplies at a future time point, and the current inventory of the related supplies.

[0148] Finally, the process of predicting usage velocity and current inventory levels for early warning includes:

[0149] S141, substituting the slope of the target medical material or the slope of the associated medical material into the statistical model to obtain a corresponding speed change reference curve; and reading the predicted usage speed at a future time point from the speed change reference curve.

[0150] S142, when the associated medical supplies do not show a one-way change feature, mark the associated medical supplies, and calculate the available time of the target medical supplies based on the predicted usage speed of the target medical supplies at a future time point and the current inventory of the target medical supplies;

[0151] The marked materials may not have seen an increase in usage rate for the time being, but according to big data statistics, there is a high probability that the rate will continue to increase in the subsequent process.

[0152] S143, when there is a one-way change feature in the associated medical supplies, calculating the available time of the target medical supplies based on the predicted usage speed of the target medical supplies at a future time point and the current inventory of the target medical supplies, and calculating the available time of the associated medical supplies based on the predicted usage speed of the associated medical supplies at a future time point and the current inventory of the associated medical supplies;

[0153] The calculation method is integration, that is, the area within the curve is used as the material usage in the future time period.

[0154] S144, when the available time is less than the first warning value, generate warning information, and output the warning information and the marked associated medical supplies to the target object.

[0155] If the calculated available time of medical supplies is less than the first warning value, a warning message is generated to prompt personnel to make purchases and allocation plans.

[0156] In addition, if it is necessary to wait until the usage rate of the associated materials increases before making a prediction, it may not be possible to schedule and plan the target medical materials at the same time. Therefore, when the current inventory of the marked associated medical materials is less than the second warning value, a warning message is generated, wherein the first warning value is less than the second warning value.

[0157] Therefore, the procurement plan or scheduling plan can be carried out synchronously with the target medical supplies, avoiding frequent triggering of the warning line and frequent formulation of procurement plans or scheduling plans, which greatly facilitates the medical supplies management work of managers.

[0158] The present invention provides a dynamic early warning method for medical supplies based on dynamic scheduling, which is mainly aimed at reducing scheduling planning. By analyzing historical usage speed information, medical supplies with usage associations are obtained. Then, when analyzing any medical supplies, if it is found that the usage speed of the medical supplies has increased in a short period of time, it means that the usage speed of the medical supplies with usage associations may also increase. Therefore, a pre-built statistical model is used to predict the usage speed at future time points, and the associated medical supplies are also predicted and warned. This avoids the situation where the inventory of multiple medical supplies triggers the warning line multiple times in a row within a period of time, thereby helping relevant personnel to make reasonable scheduling plans.

[0159] like Figure 5 As shown, the present application also provides a dynamic early warning system for medical supplies based on dynamic scheduling, including:

[0160] An acquisition module, used to acquire historical usage speed information of multiple medical supplies in multiple time periods, and to acquire current usage speed information and inventory of multiple medical supplies in a current time period, wherein the historical usage speed information includes usage speeds at multiple historical time points, and the current usage speed information includes usage speeds at multiple current time points;

[0161] A correlation analysis module, used to perform correlation analysis based on historical usage speed information of multiple medical supplies in multiple time periods to obtain a material usage speed correlation map, wherein the material usage speed correlation map includes a correlation relationship between the usage speeds of any two medical supplies;

[0162] A feature extraction module is used to extract the change features in the current usage speed information, and take the medical supplies whose change features satisfy the unidirectional change as the target medical supplies, and predict the usage speed of the target medical supplies and the related medical supplies as well as the predicted usage speed at a future time point based on a pre-built statistical model, the change features of the target medical supplies, and the change features of the related materials of the target medical supplies;

[0163] The early warning module is used to issue an early warning based on the predicted usage speed of the target medical supplies at a future time point, the current inventory of the target medical supplies, the predicted usage speed of the related supplies at a future time point, and the current inventory of the related supplies.

[0164] The present invention provides a dynamic early warning system for medical supplies based on dynamic scheduling, which is mainly aimed at reducing scheduling planning. By analyzing historical usage speed information, medical supplies with associated usage are obtained. Then, when analyzing any medical supply, if it is found that the usage speed of the medical supply has increased in a short period of time, it means that the usage speed of the associated medical supplies may also increase. Therefore, a pre-built statistical model is used to predict the usage speed at future time points, and the associated medical supplies are also predicted and warned. This avoids the situation where the inventory of multiple medical supplies triggers the warning line multiple times in a row within a period of time, thereby helping relevant personnel to make reasonable scheduling plans.

[0165] This embodiment also provides an electronic terminal, including: a processor and a memory;

[0166] The memory is used to store a computer program, and the processor is used to execute the computer program stored in the memory, so that the terminal executes any one of the methods in this embodiment.

[0167] The computer-readable storage medium in this embodiment can be understood by ordinary technicians in this field: all or part of the steps of implementing the above-mentioned method embodiments can be completed by hardware related to the computer program. The aforementioned computer program can be stored in a computer-readable storage medium. When the program is executed, the steps of the above-mentioned method embodiments are executed; and the aforementioned storage medium includes: ROM, RAM, magnetic disk or optical disk and other media that can store program codes.

[0168] The electronic terminal provided in this embodiment includes a processor, a memory, a transceiver and a communication interface. The memory and the communication interface are connected to the processor and the transceiver and complete communication with each other. The memory is used to store computer programs, the communication interface is used to communicate, and the processor and the transceiver are used to run computer programs so that the electronic terminal executes each step of the above method.

[0169] In this embodiment, the memory may include a random access memory (RAM), and may also include a non-volatile memory (non-volatile memory), such as at least one disk memory.

[0170] The above-mentioned processor can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components.

[0171] In the above-mentioned embodiments, although the present invention has been described in conjunction with the specific embodiments of the present invention, many replacements, modifications and variations of these embodiments will be apparent to those of ordinary skill in the art based on the foregoing description. The embodiments of the present invention are intended to cover all such replacements, modifications and variations falling within the broad scope of the appended claims.

[0172] The above embodiments are merely illustrative of the principles and effects of the present invention, and are not intended to limit the present invention. Anyone familiar with the art may modify or alter the above embodiments without departing from the spirit and scope of the present invention. Therefore, all equivalent modifications or alterations made by a person of ordinary skill in the art without departing from the spirit and technical concept disclosed by the present invention shall still be covered by the claims of the present invention.

Claims

1. A dynamic early warning method for medical supplies based on dynamic scheduling, characterized in that: Includes steps: Obtaining historical usage speed information of multiple medical supplies in multiple time periods, and obtaining current usage speed information and inventory of multiple medical supplies in a current time period, wherein the historical usage speed information includes usage speeds at multiple historical time points, and the current usage speed information includes usage speeds at multiple current time points; Performing correlation analysis based on historical usage speed information of multiple medical supplies in multiple time periods to obtain a material usage speed correlation map, wherein the material usage speed correlation map includes a correlation relationship between the usage speeds of any two medical supplies; Extract the change characteristics in the current usage speed information, and take the medical supplies whose change characteristics satisfy the unidirectional change as the target medical supplies, and predict the usage speed of the target medical supplies and the related medical supplies as well as the predicted usage speed at future time points based on the pre-constructed statistical model, the change characteristics of the target medical supplies, and the change characteristics of the related materials of the target medical supplies; the construction process of the statistical model includes: performing curve fitting on the usage speed of each medical supply at multiple historical time points in each time period to obtain the usage speed fluctuation curve of each medical supply in multiple time periods, wherein the usage speed fluctuation curve is located at In a two-dimensional coordinate system; based on a sliding average, the speed fluctuation curve is smoothed to obtain a speed smooth curve; a sliding window is used along the two-dimensional coordinate system The axis slides, and at each slide, the historical change characteristics of the data points in the sliding window are extracted; when the historical change characteristics satisfy a unidirectional change, the slope in the change characteristics is extracted, and the first inflection point sample after the historical change characteristics is extracted , inflection point samples A plurality of sampling points between the current position center of the sliding window and the inflection point sample, and a time interval between the current position center of the sliding window and the inflection point sample , wherein the inflection point sample is an upper inflection point or a lower inflection point, It is a time range parameter, and the time interval between adjacent sampling points is fixed; based on slope, multiple sampling points, inflection point samples and time interval Construct sample data to obtain sample data set; and based on the usage speed in the inflection point sample and time interval Cluster the sample data set to obtain multiple sample clusters; calculate the slope mean of the sample data in the target sample cluster whose sample data volume is greater than the preset quantity threshold , slope variance ; and in the slope variance When the variance is greater than the preset threshold, the sample cluster with the slope mean is removed. Sample data whose deviation rate is greater than the deviation rate threshold, and return to calculate the slope mean of the sample data in the target sample cluster , slope variance , until the slope variance When the variance is not greater than the preset threshold and the amount of sample data in the target sample cluster is greater than the preset threshold, the slope mean in the target sample cluster is used as the Construct a slope range that satisfies 3 times the standard deviation and calculate the average inflection point speed corresponding to the slope range and the average time interval ; Align the sampling points of different sample data in the target sample cluster, and calculate the average value of the aligned sampling points to obtain multiple reference sampling points; construct multiple reference sampling points and inflection point samples corresponding to multiple slope ranges Constructing a speed change reference curve; and constructing a statistical model based on the speed change reference curves with multiple slope ranges; An early warning is issued based on the predicted usage speed of the target medical supplies at a future time point, the current inventory of the target medical supplies, the predicted usage speed of the related supplies at a future time point, and the current inventory of the related supplies.

2. According to claim 1, a dynamic early warning method for medical supplies based on dynamic scheduling is characterized in that: Based on the historical usage speed information of various medical supplies in multiple time periods, a correlation analysis is performed to obtain a material usage speed correlation map, including: Conduct regularity analysis on the historical usage speed information of each medical material in multiple time periods to obtain regular medical materials with regular characteristics; Based on the regular characteristics, correlation analysis is performed on multiple regular medical supplies to obtain multiple groups of medical supplies with correlation characteristics, wherein the correlation characteristics include positive correlation and negative correlation; A material usage speed correlation map is constructed based on all groups of medical materials with correlation characteristics.

3. A dynamic early warning method for medical supplies based on dynamic scheduling according to claim 2, characterized in that: The regularity analysis of the historical usage speed information of each medical material in multiple time periods is carried out to obtain regular medical materials with regular characteristics, including: The usage speed of each medical material at multiple historical time points in each time period is curve fitted to obtain the usage speed fluctuation curve of each medical material in multiple time periods, wherein the usage speed fluctuation curve is located at In a two-dimensional coordinate system; Smoothing the usage speed fluctuation curve based on a sliding average to obtain a usage speed smooth curve; Calculate the average speed of the speed fluctuation curve and using speed variance , and construct a first usage speed mean reference line based on the usage speed average and the usage speed variance and the second use speed mean reference line ,in, To adjust the parameters; Extract the first usage speed mean reference line in the usage speed fluctuation curve The part above the speed fluctuation curve is obtained to obtain the peak section; and the second speed average reference line is extracted from the speed fluctuation curve. The lower part gets the trough section; Extract the upper inflection point of the peak segment and the lower inflection point in the trough section ; For all upper inflection points of each medical material Density-based clustering is performed to obtain multiple upper inflection point clusters; and all lower inflection points of each medical material are clustered. Performing density-based clustering to obtain a plurality of lower inflection point clusters; wherein the maximum intra-cluster distances between the upper inflection point cluster and the lower inflection point cluster are preset; The upper inflection point cluster or the lower inflection point cluster whose number of data points in the cluster is greater than the preset number threshold is taken as the target cluster, and the time range of the data points in the target cluster is calculated and speed range , obtain regular characteristics; and take the medical supplies with regular characteristics as regular medical supplies, where, is the time average of the data points in the target cluster, is the time standard deviation of the data points in the target cluster, is the average usage speed of the data points in the target cluster, is the standard deviation of the speed of the data points in the target cluster, is the range adjustment factor.

4. According to the method for dynamic early warning of medical supplies based on dynamic scheduling according to claim 3, a correlation analysis is performed on a plurality of regular medical supplies based on the regular characteristics to obtain a plurality of groups of medical supplies with correlation characteristics, including: The time period is divided into a plurality of time periods, and the plurality of time periods are assigned values ​​based on the regular characteristics of each regular medical resource to obtain a regular code for each regular medical resource, wherein the value of each bit of the regular code is positively correlated with the usage speed range of the position corresponding to the regular characteristic; The regular codes of any two regular medical supplies are subjected to bitwise difference and absolute value calculation to obtain the difference code; Calculating the sum of all values ​​in the difference code, and comparing the sum of all values ​​in the difference code with a pre-constructed first screening threshold and a second screening threshold, wherein the first screening threshold is greater than the second screening threshold; When the sum of all values ​​in the difference code is less than the second screening threshold, it is determined that the two regular medical supplies are positively associated; when the sum of all values ​​in the difference code is greater than the first screening threshold, it is determined that the two regular medical supplies are negatively associated; otherwise, it is determined that the two regular medical supplies are not associated; Based on the association relationship between any two regular medical supplies, multiple groups of medical supplies with associated characteristics are obtained.

5. According to claim 1, a dynamic early warning method for medical supplies based on dynamic scheduling is characterized in that: Extracting the change feature in the current speed information includes: And at each slide, calculate the average value of all data points in the sliding window ,in, is the number of sliding steps; Determine whether multiple averages meet and ,or, and ; If satisfied, it is determined that there is a unidirectional change trend and the change slope is calculated , ,in, For the The average value calculated by sliding, is the average value of the first calculation, The average value calculated from the last slide, For the The average value calculated by sliding, is the time interval between any two slides.

6. A dynamic early warning method for medical supplies based on dynamic scheduling according to claim 1, characterized in that: Based on the pre-built statistical model, the change characteristics of the target medical supplies, and the change characteristics of the associated materials of the target medical supplies, the target medical supplies and the associated medical supplies are predicted to be used at a future time point, including: Substitute the slope of the target medical material or the slope of the associated medical material into the statistical model to obtain a corresponding speed change reference curve; and read the predicted usage speed at a future time point from the speed change reference curve.

7. A dynamic early warning method for medical supplies based on dynamic scheduling according to claim 6, characterized in that: Providing an early warning based on the predicted usage speed of the target medical supplies at a future time point, the current inventory of the target medical supplies, the predicted usage speed of the associated supplies at a future time point, and the current inventory of the associated supplies, includes: When the associated medical supplies do not show a unidirectional change feature, the associated medical supplies are marked, and the available time of the target medical supplies is calculated based on the predicted usage speed of the target medical supplies at a future time point and the current inventory of the target medical supplies; When there is a one-way change feature in the associated medical supplies, the available time of the target medical supplies is calculated based on the predicted usage speed of the target medical supplies at a future time point and the current inventory of the target medical supplies, and the available time of the associated medical supplies is calculated based on the predicted usage speed of the associated medical supplies at a future time point and the current inventory of the associated medical supplies; When the available time is less than the first warning value, a warning message is generated, and the warning message and the marked associated medical supplies are output to the target object.

8. A dynamic early warning method for medical supplies based on dynamic scheduling according to claim 7, characterized in that: Also includes: When the current inventory of the marked associated medical supplies is less than a second warning value, a warning message is generated, wherein the first warning value is less than the second warning value.

9. A dynamic early warning system for medical supplies based on dynamic scheduling, characterized in that: include: An acquisition module, used to acquire historical usage speed information of multiple medical supplies in multiple time periods, and to acquire current usage speed information and inventory of multiple medical supplies in a current time period, wherein the historical usage speed information includes usage speeds at multiple historical time points, and the current usage speed information includes usage speeds at multiple current time points; A correlation analysis module, used to perform correlation analysis based on historical usage speed information of multiple medical supplies in multiple time periods to obtain a material usage speed correlation map, wherein the material usage speed correlation map includes a correlation relationship between the usage speeds of any two medical supplies; The feature extraction module is used to extract the change characteristics in the current usage speed information, and take the medical supplies whose change characteristics satisfy the unidirectional change as the target medical supplies, and predict the usage speed of the target medical supplies and the related medical supplies as well as the predicted usage speed at a future time point based on the pre-constructed statistical model, the change characteristics of the target medical supplies, and the change characteristics of the related materials of the target medical supplies; the construction process of the statistical model includes: curve fitting of the usage speed of each medical supply at multiple historical time points in each time period to obtain the usage speed fluctuation curve of each medical supply in multiple time periods, wherein the usage speed fluctuation curve is located at In a two-dimensional coordinate system; based on a sliding average, the speed fluctuation curve is smoothed to obtain a speed smooth curve; a sliding window is used along the two-dimensional coordinate system The axis slides, and at each slide, the historical change characteristics of the data points in the sliding window are extracted; when the historical change characteristics satisfy a unidirectional change, the slope in the change characteristics is extracted, and the first inflection point sample after the historical change characteristics is extracted , inflection point samples A plurality of sampling points between the current position center of the sliding window and the inflection point sample, and a time interval between the current position center of the sliding window and the inflection point sample , wherein the inflection point sample is an upper inflection point or a lower inflection point, It is a time range parameter, and the time interval between adjacent sampling points is fixed; based on slope, multiple sampling points, inflection point samples and time interval Construct sample data to obtain sample data set; and based on the usage speed in the inflection point sample and time interval Cluster the sample data set to obtain multiple sample clusters; calculate the slope mean of the sample data in the target sample cluster whose sample data volume is greater than the preset quantity threshold , slope variance ; and in the slope variance When the variance is greater than the preset threshold, the sample cluster with the slope mean is removed. Sample data whose deviation rate is greater than the deviation rate threshold, and return to calculate the slope mean of the sample data in the target sample cluster , slope variance , until the slope variance When the variance is not greater than the preset threshold and the amount of sample data in the target sample cluster is greater than the preset threshold, the slope mean in the target sample cluster is used as the Construct a slope range that satisfies 3 times the standard deviation and calculate the average inflection point speed corresponding to the slope range and the average time interval ; Align the sampling points of different sample data in the target sample cluster, and calculate the average value of the aligned sampling points to obtain multiple reference sampling points; construct multiple reference sampling points and inflection point samples corresponding to multiple slope ranges Constructing a speed change reference curve; and constructing a statistical model based on the speed change reference curves with multiple slope ranges; The early warning module is used to issue an early warning based on the predicted usage speed of the target medical supplies at a future time point, the current inventory of the target medical supplies, the predicted usage speed of the related supplies at a future time point, and the current inventory of the related supplies.

Images