A data management method, system and storage medium for intelligent manufacturing

Abstract

The present invention discloses a data management method, system and storage medium for intelligent manufacturing, which belongs to the field of factory production control technology. The method includes: calculating forecast data based on historical order data, the forecast data is the demand data of the product in a predetermined time period in the future; calculating out-of-stock data according to the inventory data and forecast data of the product, and defining the product whose out-of-stock data is greater than the first threshold as the target product; obtaining the production materials and production routes of each target product; if the production routes of multiple target products have the same production equipment, the target product is defined as a twin product, and the corresponding production route is a twin route; obtaining the production time of each twin product; combining the out-of-stock data and the production time to allocate the occupancy time of each twin product using the twin route, and generating production suggestions according to the occupancy time. The present invention can generate production suggestions by predicting future order situations, thereby assisting decision makers in formulating reasonable production plans.

Description

Technical Field

[0001] The present invention belongs to the technical field of factory production control, and specifically relates to a data management method, system and storage medium for intelligent manufacturing. Background Art

[0002] The core of intelligent manufacturing is to use the new generation of information technology to transform traditional manufacturing technology into intelligent, automated and digital ones. With the help of data analysis, factories can monitor and predict the production process in real time, discover and solve problems early, thereby ensuring product quality and improving production efficiency.

[0003] A variety of intelligent manufacturing solutions have been proposed in the prior art. For example, Chinese patent document CN117311299B discloses a factory management system and method based on multi-source heterogeneous data integration. The method includes: pre-processing and classifying the multi-source heterogeneous data to obtain production management data; monitoring changes in the production management data, and scheduling the production tasks of each production line when changes are found in the production task data; calculating the production pressure of each product type based on the production task schedule, production task data and production line data of the production line; calculating the utilization rate of the personnel of each production line, judging the production efficiency of each production personnel based on the utilization rate of the production personnel, calculating the production pressure on the production line of each product type based on the production line data, and finally generating a production personnel mobilization strategy. For example, Chinese patent document CN116976614A discloses an intelligent management method, system, equipment and storage medium for a digital factory. The method first obtains the production plan data of the current order to be processed and the current production capacity of the factory, plans the resource scheduling plan for the order to be processed according to the production plan data and the current production capacity, and performs processing technology planning on the order to be processed according to the resource scheduling plan to obtain order processing technology parameters that are compatible with the factory's current resources. In the process of processing the order according to the order processing technology parameters, the processing progress of the order to be processed is obtained, and the production plan data is fed back according to the processing progress to obtain production plan feedback data corresponding to the actual order progress and used to adjust the production plan of the order to be processed, thereby improving the coordination of resource scheduling.

[0004] The above-mentioned existing technologies are all aimed at existing orders and are intelligently scheduled in combination with the factory's production capacity; however, when a large number of orders appear in a short period of time, the factory's production capacity may be exceeded. At this time, even if the above-mentioned technical solutions are applied, there may be a situation where the goods cannot be delivered on time; with the development of big data technology, it has become possible to predict future order data based on past historical order data. Therefore, there is an urgent need for a technical solution that can generate factory production suggestions by predicting future order situations. Summary of the invention

[0005] To solve the above problems, the present invention provides a data management method, system and storage medium for intelligent manufacturing, so as to provide a technical solution that can generate factory production suggestions by predicting future order situations.

[0006] In order to achieve the above-mentioned object of the invention, the present invention proposes a data management method for intelligent manufacturing, comprising:

[0007] Calculate forecast data based on the historical order data of the product, where the forecast data is the demand data of the product in a predetermined time period in the future;

[0008] Calculate out-of-stock data according to the inventory data of the product and the forecast data, and define the product whose out-of-stock data is greater than a first threshold as a target product;

[0009] If there are multiple target products, obtain the production materials and production route of each target product, wherein the production route includes multiple production equipment, and the production materials pass through the production equipment in a preset order to obtain the target product;

[0010] If the production routes of multiple target products have the same production equipment, the target products are defined as twin products, and the corresponding production routes are twin routes;

[0011] Analyze the historical production data of the twin routes and obtain the production time of each of the twin products;

[0012] The out-of-stock data and the production duration are combined to allocate the occupancy time of each twin product using the twin route, and a production suggestion is generated according to the occupancy time.

[0013] Further, calculating the predicted data includes the following steps:

[0014] Acquire the daily order volume from the historical order data, generate a first sequence and a second sequence for each month based on the order volume, wherein the first sequence and the second sequence belonging to the same month have the same code, combine the first sequences with different codes in pairs to generate multiple combination results, and calculate a first mutual exclusion degree between two first sequences in the combination results;

[0015] Clustering the combination results based on the first mutual exclusion to obtain multiple first clustering results, wherein the first mutual exclusion of the combination results in the same first clustering result is less than a second threshold, calculating a second mutual exclusion between the first sequence and the second sequence in the first clustering result, and generating a second clustering result by clustering based on the second mutual exclusion;

[0016] Obtain an actual sequence before the predetermined time period, calculate a third mutual exclusion degree between the actual sequence and each of the first clustering results, take the first clustering result with the smallest third mutual exclusion degree as a matching result, count the number of second sequence aggregations included in the matching result corresponding to the second clustering result, generate a predicted sequence based on the second clustering result with the largest number of aggregations, add the order quantities of the predicted sequence, and obtain the predicted data.

[0017] Further, calculating the first mutual exclusion degree includes the following steps:

[0018] The first sequence for comparison is defined as a first object and a second object, the order volume with the largest value in the first object and the second object is defined as a first base value and a second base value respectively, the first base value and the second base value are reduced by a predetermined multiple to obtain a first value and a second value, the order volume in the first object that is greater than the first value is screened and defined as a first comparison value, and the order volume in the second object that is greater than the second value is defined as a second comparison value;

[0019] The first mutual exclusion degree is calculated based on the first formula , the first formula is:

[0020] ,

[0021] Wherein, I, J, and K are respectively the number of the first comparison values, the number of the second comparison values, and the number of the order quantities in the first object. is the i-th first comparison value, is the order quantity corresponding to the i-th order of the first comparison value in the second object, is the jth second comparison value, is the order quantity corresponding to the jth order of the second comparison value in the first object, and are the k-th order quantities in the first object and the second object respectively.

[0022] Furthermore, the first mutual exclusivity between the actual sequence and each of the first sequences in the first clustering result is calculated, and an average value of the first mutual exclusivity is used as the second mutual exclusivity with the first clustering result.

[0023] Further, calculating the production time of the twin product includes the following steps:

[0024] Obtaining the historical processing time of the production equipment for processing the same production material, clustering the historical processing time to obtain a third clustering result, wherein the difference between any two historical processing times in the same third clustering result is less than a third threshold, and taking the historical processing time as the center of the third clustering result as the representative time;

[0025] If there is only one third clustering result, the representative duration in the third clustering result is set as the standard duration of the production equipment; if there are multiple third clustering results, the production process data of the production equipment when processing the production raw materials is obtained, the production process data is analyzed to screen out the best result from the multiple third clustering results, and the representative duration in the best result is used as the standard duration;

[0026] Calculate the sum of the standard durations of all the production equipment in the twin route, and use the sum as the production duration of the twin product.

[0027] Further, analyzing the production process data comprises the following steps:

[0028] Based on the third clustering result, a plurality of environmental parameters when processing the production raw materials in each of the historical processing durations are obtained, and a coordinate system is established, wherein the coordinate system takes the historical processing duration as the end point of the horizontal axis, takes the first preset value and the second preset value as the end point scale of the vertical axis, and plots actual curves of the changes of different environmental parameters over time in the coordinate system;

[0029] A template curve is established for each of the environmental parameters, and the similarity between each of the actual curves and each of the template curves is identified. The average value of the similarities is used as the evaluation value corresponding to the historical processing time. The average value of the evaluation values ​​of all the historical processing times in the third clustering result is calculated and used as the state value of the third clustering result. The third clustering result with the largest state value that is greater than a fourth threshold is used as the best result.

[0030] Furthermore, after detecting a predetermined operation, the coordinate system shortens the endpoint scale of the horizontal axis to a second preset value, reduces the endpoint scale of the vertical axis to a third parameter value and a fourth parameter value, and increases the second preset value by a preset step size to dynamically display the dynamic change trend of each of the actual curves. Before increasing the second preset value, the maximum parameter value and the minimum parameter value of the environmental parameter in the future time period are obtained. If the maximum parameter value and the minimum parameter value are outside the endpoint scales of the vertical axis, the endpoint scales of the vertical axis are set to the maximum parameter value and the minimum parameter value, respectively.

[0031] Furthermore, the twin product with larger out-of-stock data is assigned a longer occupancy time.

[0032] The present invention also provides a data management system for intelligent manufacturing, which is used to implement the above-mentioned data management method for intelligent manufacturing, and the system includes:

[0033] A forecasting module, which calculates forecast data based on the historical order data of the product, wherein the forecast data is the demand data of the product in a predetermined time period in the future;

[0034] A screening module, which calculates out-of-stock data according to the inventory data of the product and the forecast data, and defines the product whose out-of-stock data is greater than a first threshold as a target product;

[0035] An analysis module, if there are multiple target products, then obtain the production materials and production routes of each target product, the production route includes multiple production equipment, the production materials pass through the production equipment in a preset order to obtain the target product, if the production routes of multiple target products have the same production equipment, then define the target product as a twin product, the corresponding production route is a twin route, analyze the historical production data of the twin route and obtain the production time of each twin product;

[0036] The suggestion module combines the out-of-stock data and the production time to allocate the occupancy time of each twin product using the twin route, and generates production suggestions according to the occupancy time.

[0037] The present invention also discloses a computer storage medium, characterized in that the computer storage medium stores program instructions, wherein when the program instructions are executed, the device where the computer storage medium is located is controlled to execute the above-mentioned method.

[0038] Compared with the prior art, the beneficial effects of the present invention are at least as follows:

[0039] The present invention first calculates the future demand data for the product based on historical data, then calculates the out-of-stock data of the product based on the inventory data and the demand data, and selects products with a large out-of-stock volume as target products; then analyzes the production process of the target product, and searches for products that cannot be produced at the same time as twin products. Such products cannot be produced at the same time due to conflicts in the production lines, and therefore need more attention; finally, by obtaining the production time of the twin products, combined with the out-of-stock data and the available time of the production equipment, production suggestions are generated to help managers formulate reasonable production plans. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 A flowchart of a data management method for intelligent manufacturing according to the present invention;

[0041] Figure 2 is a diagram showing the calculation principle of the first mutual exclusion degree of the present invention;

[0042] Figure 3 The present invention is a structural schematic diagram of a data management system for intelligent manufacturing. DETAILED DESCRIPTION

[0043] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0044] It is understood that the terms "first", "second", etc. used in this application may be used herein to describe various elements, but unless otherwise specified, these elements are not limited by these terms. These terms are only used to distinguish a first element from another element. For example, without departing from the scope of this application, a first xx script may be referred to as a second xx script, and similarly, a second xx script may be referred to as a first xx script.

[0045] like Figure 1 As shown, a data management method for intelligent manufacturing includes:

[0046] Step S1: Calculate forecast data based on the historical order data of the product, where the forecast data is the demand data of the product in a predetermined time period in the future.

[0047] The products in this embodiment include multiple types, which are set here to include product 1, product 2 and product 3. The historical order data of product 1, product 2 and product 3 are obtained. The historical order data is the order volume data received by the factory in the past period of time. Then, the future demand data of product 1, product 2 and product 3 is predicted based on the historical order data. This embodiment is based on the historical order data of the first half of each month to predict the forecast data for the second half of the month. The specific prediction method will be introduced in detail later.

[0048] Step S2: Calculate out-of-stock data based on the inventory data and forecast data of the product, and define the product whose out-of-stock data is greater than a first threshold as a target product.

[0049] In this embodiment, out-of-stock data is calculated by subtracting inventory data from demand data. If the out-of-stock data is greater than a preset first threshold, it indicates that the supply gap of the product is large, and such products are set as target products and paid attention to. If it is less than or equal to the first threshold, it indicates that the product gap is small and there is no need to make production plans in advance.

[0050] Step S3: If there are multiple target products, the production materials and production route of each target product are obtained. The production route includes multiple production equipment. The production materials pass through the production equipment in a preset order to obtain the target product.

[0051] Step S4: If the production routes of multiple target products have the same production equipment, the target products are defined as twin products, and the corresponding production routes are twin routes;

[0052] In this embodiment, products 1-3 are all set as target products, and the production route of each product is obtained. For example, the production route 1 of product 1 is: the production raw material 1 passes through the production equipment 1, the generation equipment 2 and the production equipment 3 to obtain product 1, and the production route 2 of product 2 is: the production raw material 2 passes through the production equipment 4, the production equipment 2 and the production equipment 5 to obtain product 2. It can be seen that the same production equipment 2 exists in the production route 1 and the production route 2, so product 1 and product 2 are set as twin products. In particular, since the same production equipment 2 exists in both, product 1 and product 2 cannot be produced at the same time, and at this time, the production capacity of both needs to be planned in advance.

[0053] Step S5: Analyze the historical production data of the twin route and obtain the production time of each twin product.

[0054] The production time here refers to the total time from the time the production materials enter the production route to the time the product is obtained. The specific details of calculating the product production time will be introduced later.

[0055] Step S6: Combine the out-of-stock data and production time to allocate the occupancy time of each twin product using the twin route, and generate production suggestions based on the occupancy time.

[0056] Specifically, when allocating occupancy time, it can be determined according to the size of the out-of-stock data. For example, the larger the out-of-stock data, the longer the occupancy time allocated to the product. In addition, the present invention also obtains the available time of production equipment 2. The available time of production equipment 2 can be determined according to the current and future production plans. For example, if the total available time of production equipment 2 is 100 and the ratio of the out-of-stock data of product 1 and product 2 is 3:1, then the occupancy time allocated to product 1 is 75 and the occupancy time allocated to product 2 is 25. Afterwards, the production suggestion generated according to the occupancy time is, for example, "Based on the historical order data, products 1-3 may be in large shortage in the future, and there is a conflict in the production of product 1 and product 2. It is recommended to give priority to the production of product 1, and the recommended generation time is set to 75".

[0057] The present invention first calculates the future demand data for the product based on historical data, then calculates the out-of-stock data of the product based on the inventory data and the demand data, and selects products with a large out-of-stock volume as target products; then analyzes the production process of the target product, and searches for products that cannot be produced at the same time as twin products. Such products cannot be produced at the same time due to conflicts in the production lines, and therefore need more attention; finally, by obtaining the production time of the twin products, combined with the out-of-stock data and the available time of the production equipment, production suggestions are generated to help managers formulate reasonable production plans.

[0058] It is particularly noteworthy that the present invention can generate production suggestions by predicting future order situations, thereby assisting decision makers in formulating reasonable production plans.

[0059] The calculation of the prediction data in this embodiment includes the following steps:

[0060] The daily order volume is obtained from the historical order data. The first and second sequences of each month are generated based on the order volume. The first and second sequences belonging to the same month have the same code. The first sequences with different codes are combined in pairs to generate multiple combination results. The first mutual exclusion degree between the two first sequences in the combination results is calculated.

[0061] The historical order data of this embodiment includes the order volume of each day in the previous year, and then the first sequence and the second sequence of each month are generated according to the order volume. The first sequence is a data sequence generated according to time. For example, the values ​​of the first sequence are 8, 7, 15, 9, 14, etc. The first sequence includes the order volume of the first 15 days of each month, and the second sequence includes the order volume of the last 15 days of each month. Among them, if each month is less than 30 days, such as February has 28 days, the order volume on the 28th is divided into three parts, which are used as the order volumes of the 28th to the 30th respectively. If each month exceeds 30 days, the order volume on the 31st is merged into the order volume on the 30th, so that the first sequence and the second sequence are aligned in quantity.

[0062] When encoding in this example, the first sequence and the second sequence belonging to the same month have the same code, for example, they belong to January and are encoded as first sequence 1 and second sequence 1. Then, the first sequences of different months are combined in pairs, such as first sequence 1 and first sequence 2, first sequence 1 and first sequence 3, etc., and a total of 66 combinations are obtained. The first mutual exclusion degree between the two sequences of each combination is calculated. The higher the first mutual exclusion degree, the less similar the two first sequences are. The calculation method will be introduced later.

[0063] The combination results are clustered based on the first mutual exclusion to obtain multiple first clustering results, the first mutual exclusion of the combination results in the same first clustering result are all less than the second threshold, the second mutual exclusion between the first sequence and the second sequence in the first clustering result is calculated, and the second clustering result is generated based on the second mutual exclusion clustering.

[0064] In this implementation, clustering is performed when the first mutual exclusion is less than the second threshold value. Then, the first clustering results include relatively similar first sequences. Since each first sequence has a corresponding second sequence, clustering is performed again using the same method based on the first clustering results. The calculation method of the second mutual exclusion is the same as that of the first mutual exclusion, and will not be repeated here. The second sequence is clustered into multiple second clustering results, and the second clustering results include relatively similar second sequences.

[0065] Get the actual sequence before the predetermined time period, calculate the third mutual exclusion between the actual sequence and each first clustering result, take the first clustering result with the smallest third mutual exclusion as the matching result, count the number of second sequence aggregations included in the second clustering result corresponding to the matching result, generate a predicted sequence based on the second clustering result with the largest number of aggregations, add up the order quantities of the predicted sequence, and obtain the predicted data.

[0066] Here, the scheduled time period is the last 15 days of this month, and the actual sequence before it is the sequence composed of the order volume of the first 15 days of this month; the actual sequence obtained is used to calculate the third mutual exclusion with each first clustering result. The smaller the third mutual exclusion, the more similar the actual sequence is to the first clustering result. Then the first clustering result with the smallest third mutual exclusion is the result most similar to the actual sequence, which means that the order pattern of the first sequence in the first clustering result is similar to the actual sequence. On this basis, when the second clustering result is obtained, it can be known that the more the number of second clustering results, the more likely it is that the second sequence under the clustering result will appear after the first sequence. Therefore, the predicted sequence is generated according to the second clustering result with the largest number of aggregations; finally, the 15 order volumes in the predicted sequence are added together to obtain the total data of 15 days as the predicted data.

[0067] In this embodiment, calculating the first mutual exclusion degree includes the following steps:

[0068] The first sequence for comparison is defined as the first object and the second object, the order quantities with the largest values ​​in the first object and the second object are defined as the first base value and the second base value respectively, the first value and the second value are obtained by reducing the first base value and the second base value by a predetermined multiple, the order quantity in the first object that is greater than the first value is screened and defined as the first comparison value, and the order quantity in the second object that is greater than the second value is defined as the second comparison value.

[0069] The first object and the second object both include 15 order quantities. The order quantity with the largest value is extracted from the first object as the first base value, and the order quantity with the largest value is extracted from the second object as the second base value. In this embodiment, the first base value is reduced to 0.75 times of the original value to obtain the first value, and the second base value is reduced to 0.75 times of the original value to obtain the second value. Then, from the 15 order quantities included in the first object, the order quantities with values ​​greater than the first value are screened as the first comparison value, and from the 15 order quantities included in the second object, the order quantities with values ​​greater than the second value are screened as the second comparison value.

[0070] Calculate the first mutual exclusion degree based on the first formula , the first formula is:

[0071] ,

[0072] Wherein, I, J, and K are respectively the number of the first comparison values, the number of the second comparison values, and the number of the order quantities in the first object. is the i-th first comparison value, is the order quantity corresponding to the i-th order of the first comparison value in the second object, is the jth second comparison value, is the order quantity corresponding to the jth order of the second comparison value in the first object, and are the k-th order quantities in the first object and the second object respectively.

[0073] Reference Figure 2 Here, the order quantities in the order of 2, 5, and 10 are extracted from the 15 order quantities of the first object X as the first comparison values ​​A1, A2, and A3, then I=3, and accordingly, the order quantities in the order of 2, 5, and 10 are extracted from the second object Y as B1, B2, and B3; the order quantities in the order of 4, 6, and 12 are extracted from the 15 order quantities of the second object Y as the second comparison values ​​C1, C2, and C3, then J=3, and accordingly, the order quantities in the order of 4, 6, and 12 are extracted from the first object X as D1, D2, and D3; finally, the 15 order quantities in the first object X are sequentially taken as E1~E15, and the 15 order quantities in the second object Y are sequentially taken as F1~F15, and the corresponding K=15. From the above calculation process, it can be seen that if the first object is similar to the second object, the calculated first mutual exclusion will be smaller. Therefore, in the case of similarity, the results of the three monomials will approach 0, while in the case of dissimilarity, the calculated first mutual exclusion will be larger.

[0074] In this embodiment, the first mutual exclusion degree between the actual sequence and each first sequence in the first clustering result is calculated, and the average value of the first mutual exclusion degree is used as the second mutual exclusion degree with the first clustering result.

[0075] For example, the first clustering result has three combination results, namely, the first sequence 1 and the first sequence 2, the first sequence 2 and the first sequence 3, and the first sequence 1 and the first sequence 3. Then there are first sequences 1-3. The first mutual exclusion degree 1-3 between the actual sequence and the first sequences 1-3 is calculated. The average value of the first mutual exclusion degree 1-3 is calculated as the second mutual exclusion value of the first clustering result. The second mutual exclusion value calculated in this way can represent the similarity between the actual sequence and the first clustering result as a whole.

[0076] In this embodiment, calculating the production time of twin products includes the following steps:

[0077] The historical processing time of production equipment for processing the same kind of production raw materials is obtained, and the historical processing time is clustered to obtain a third clustering result. If the difference between any two historical processing times in the same third clustering result is less than a third threshold, the historical processing time at the center of the third clustering result is used as the representative time.

[0078] In this embodiment, the historical processing time of the production material processed by the generating device 1 for the past 100 times is obtained, and then clustering is performed according to the difference between the historical processing time. The K-means clustering algorithm can be used to obtain the third clustering result by specifying the number of clusters in advance; by clustering in this way, for example, the processing time between 6-8s can be clustered into one category, the processing time between 8-10s can be clustered into another category, and so on. When determining the cluster center, a historical processing time can be extracted, and the comprehensive difference between the historical processing time and other historical processing time is calculated as the central value of the historical processing time. Based on this method, the central value of each historical processing time is calculated, and the historical processing time with the smallest central value is used as the center of the cluster.

[0079] If there is only one third clustering result, the representative duration in the third clustering result is set as the standard duration of the production equipment. If there are multiple third clustering results, the production process data of the production equipment when processing production materials is obtained, and the production process data is analyzed to screen out the best result from multiple third clustering results, and the representative duration in the best result is used as the standard duration.

[0080] Calculate the sum of the standard durations of all production equipment in the twin route, and use the sum as the production duration of the twin product.

[0081] If there is only one third clustering result, it indicates that production equipment 1 has been in a relatively stable state, so the representative duration is used as the standard duration; if there are multiple third clustering results, it indicates that the equipment state of production equipment 1 has changed, so that the time for production equipment 1 to process production materials at different stages is different. This may be due to equipment parameter adjustment or equipment abnormality. Therefore, it is necessary to determine the period when production equipment 1 is in the best state, that is, to determine the best third clustering result. Specifically, it can be determined through automatic analysis by the system, or by visualizing the process environment data of the production equipment and using manual analysis to determine it. The specific introduction will be made later. After determining the best third clustering result, the representative duration is used as the standard duration. This method is used to calculate the standard duration of each production equipment on the generation route, and the standard duration is added and summed as the production duration of the production material being completely processed into the product. The duration determined in this way is the production duration of all production equipment in the best state in the recent period of time, making the subsequent production suggestions more reasonable.

[0082] This embodiment analyzes the production process data including the following steps:

[0083] Based on the third clustering result, a variety of environmental parameters when processing production raw materials in each historical processing time are obtained, and a coordinate system is established. The coordinate system uses the historical processing time as the end point of the horizontal axis, and the first preset value and the second preset value as the end point scale of the vertical axis. The actual curves of different environmental parameters changing with time are plotted in the coordinate system.

[0084] A template curve for each environmental parameter is established, the similarity between each actual curve and each template curve is identified, the average value of the similarity is used as the evaluation value of the corresponding historical processing time, the average value of all historical processing time evaluation values ​​in the third clustering result is calculated, and used as the state value of the third clustering result, and the third clustering result with the largest state value that is greater than the fourth threshold is taken as the best result.

[0085] Specifically, the production process data corresponding to the historical processing time of 10s includes three environmental parameters: temperature, humidity, and pressure. The actual curves of the three environmental parameters changing with time are plotted in a coordinate system, where the scale of the coordinate system is predetermined, such as setting the minimum scale of the vertical axis to 0 and the maximum scale to 100. This facilitates the subsequent automatic comparison of curve similarity.

[0086] The template curve is a pre-set standard change curve, such as a horizontal line of 100°C for the temperature template curve and a parabola for the pressure template curve. The actual temperature curve is then compared with the template curve to obtain the similarity between the two. The similarity can be calculated using a deep learning algorithm, such as a CNN network, or using a simple trend comparison method, that is, by comparing the increase and decrease of the values, to see whether the rising and falling trends of the two curves at the same time point are the same, and combining the numerical fluctuation range of the two curves to calculate the similarity of the two curves. Here, since each third clustering result includes multiple environmental parameters, the average of the evaluation values ​​of different environmental parameters is taken as the state value. If the state value is the largest and is greater than the fourth threshold, it means that in the third clustering result, all actual curves are similar to the template curve, that is, the historical processing time included in the third clustering result is the processing time of the production equipment in the best state.

[0087] In addition, if the maximum state value is less than the fourth threshold, it indicates that the comparison result is poor and manual analysis is required.

[0088] After detecting the predetermined operation, the coordinate system shortens the endpoint scale of the horizontal axis to the second preset value, reduces the endpoint scale of the vertical axis to the third parameter value and the fourth parameter value, and increases the second preset value by the preset step size to dynamically display the dynamic change trend of each actual curve. Before increasing the second preset value, the maximum parameter value and the minimum parameter value of the environmental parameter in the future time length are obtained. If the maximum parameter value and the minimum parameter value are outside the endpoint scale of the vertical axis, the endpoint scale of the vertical axis is set to the maximum parameter value and the minimum parameter value, respectively.

[0089] Specifically, this solution is provided with a specific button on the graphical interface. When the user performs manual analysis, first click the button with the mouse, and the coordinate system will dynamically display the changes of each actual curve based on the passage of time, which can help the user analyze the production process. The second preset value is 2s. For example, if the historical processing time is 10s, then after detecting the predetermined operation, the end point of the horizontal axis of the coordinate axis is set to 2s, and it changes dynamically in steps of 1s. The future duration is set to 2s. When it is detected that the maximum parameter value will be greater than the maximum scale of the vertical axis, or the minimum parameter value will be less than the minimum scale of the vertical axis in the next 2s, the scale is adjusted according to the maximum or minimum parameter value, so that the curve of the environmental parameters can be fully displayed in the coordinate system.

[0090] In this embodiment, the twin product with larger out-of-stock data is assigned a longer occupancy time.

[0091] like Figure 3 As shown, the present invention also provides a data management system for intelligent manufacturing, which is used to implement the above-mentioned data management method for intelligent manufacturing, and the system includes:

[0092] The forecasting module calculates forecast data based on the historical order data of the product. The forecast data is the demand data of the product in a predetermined time period in the future.

[0093] A screening module, which calculates out-of-stock data based on the inventory data and forecast data of the product, and defines the product whose out-of-stock data is greater than a first threshold as a target product;

[0094] Analysis module: if there are multiple target products, the production materials and production routes of each target product are obtained. The production route includes multiple production equipment. The production materials pass through the production equipment in a preset order to obtain the target product. If the production routes of multiple target products have the same production equipment, the target product is defined as a twin product, and the corresponding production route is the twin route. The historical production data of the twin route is analyzed and the production time of each twin product is obtained.

[0095] The recommendation module combines the out-of-stock data and production time to allocate the occupancy time of each twin product using the twin route, and generates production recommendations based on the occupancy time.

[0096] The present invention also discloses a computer storage medium, characterized in that the computer storage medium stores program instructions, wherein when the program instructions are executed, the device where the computer storage medium is located is controlled to execute the above method.

[0097] It should be understood that, although each step in the flow chart of each embodiment of the present invention is shown in sequence according to the indication of the arrow, these steps are not necessarily performed in sequence according to the order indicated by the arrow. Unless there is a clear explanation in this article, the execution of these steps does not have a strict order restriction, and these steps can be performed in other orders. Moreover, at least a portion of the steps in each embodiment may include a plurality of sub-steps or a plurality of stages, and these sub-steps or stages are not necessarily performed at the same time, but can be performed at different times, and the execution order of these sub-steps or stages is not necessarily performed in sequence, but can be performed in turn or alternately with at least a portion of other steps or sub-steps or stages of other steps.

[0098] The technical features of the above-mentioned embodiments can be combined arbitrarily. In order to make the description concise, not all possible combinations of the technical features in the above-mentioned embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0099] The above embodiments only express several implementation methods of the present invention, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the patent of the present invention. It should be pointed out that, for those of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present invention, and these all belong to the protection scope of the present invention. Therefore, the protection scope of the patent of the present invention shall be subject to the attached claims.

[0100] The above are only preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A data management method for intelligent manufacturing, characterized in that: include: Calculate forecast data based on the historical order data of the product, where the forecast data is the demand data of the product in a predetermined time period in the future; Calculate out-of-stock data according to the inventory data of the product and the forecast data, and define the product whose out-of-stock data is greater than a first threshold as a target product; If there are multiple target products, obtain the production materials and production route of each target product, wherein the production route includes multiple production equipment, and the production materials pass through the production equipment in a preset order to obtain the target product; If the production routes of multiple target products have the same production equipment, the target products are defined as twin products, and the corresponding production routes are twin routes; Analyze the historical production data of the twin routes and obtain the production time of each of the twin products; The out-of-stock data and the production duration are combined to allocate the occupancy time of each twin product using the twin route, and a production suggestion is generated according to the occupancy time.

2. The method according to claim 1, characterized in that Calculating the predicted data includes the following steps: Acquire the daily order volume from the historical order data, generate a first sequence and a second sequence for each month based on the order volume, wherein the first sequence and the second sequence belonging to the same month have the same code, combine the first sequences with different codes in pairs to generate multiple combination results, and calculate a first mutual exclusion degree between two first sequences in the combination results; Clustering the combination results based on the first mutual exclusion to obtain multiple first clustering results, wherein the first mutual exclusion of the combination results in the same first clustering result is less than a second threshold, calculating a second mutual exclusion between the first sequence and the second sequence in the first clustering result, and generating a second clustering result by clustering based on the second mutual exclusion; Obtain an actual sequence before the predetermined time period, calculate a third mutual exclusion degree between the actual sequence and each of the first clustering results, take the first clustering result with the smallest third mutual exclusion degree as a matching result, count the number of second sequence aggregations included in the matching result corresponding to the second clustering result, generate a predicted sequence based on the second clustering result with the largest number of aggregations, add the order quantities of the predicted sequence, and obtain the predicted data.

3. The method according to claim 2, characterized in that Calculating the first mutual exclusion degree comprises the following steps: The first sequence for comparison is defined as a first object and a second object, the order volume with the largest value in the first object and the second object is defined as a first base value and a second base value respectively, the first base value and the second base value are reduced by a predetermined multiple to obtain a first value and a second value, the order volume in the first object that is greater than the first value is screened and defined as a first comparison value, and the order volume in the second object that is greater than the second value is defined as a second comparison value; The first mutual exclusion degree is calculated based on the first formula , the first formula is: ， Wherein, I, J, and K are respectively the number of the first comparison values, the number of the second comparison values, and the number of the order quantities in the first object. is the i-th first comparison value, is the order quantity corresponding to the i-th order of the first comparison value in the second object, is the jth second comparison value, is the order quantity corresponding to the jth order of the second comparison value in the first object, and are the k-th order quantities in the first object and the second object respectively.

4. The method according to claim 2, characterized in that: The first mutual exclusion degree between the actual sequence and each of the first sequences in the first clustering result is calculated, and an average value of the first mutual exclusion degrees is used as the second mutual exclusion degree with the first clustering result.

5. The method according to claim 1, characterized in that Calculating the production time of the twin product includes the following steps: Obtaining the historical processing time of the production equipment for processing the same production material, clustering the historical processing time to obtain a third clustering result, wherein the difference between any two historical processing times in the same third clustering result is less than a third threshold, and taking the historical processing time as the center of the third clustering result as the representative time; If there is only one third clustering result, the representative duration in the third clustering result is set as the standard duration of the production equipment; if there are multiple third clustering results, the production process data of the production equipment when processing the production raw materials is obtained, the production process data is analyzed to screen out the best result from the multiple third clustering results, and the representative duration in the best result is used as the standard duration; Calculate the sum of the standard durations of all the production equipment in the twin route, and use the sum as the production duration of the twin product.

6. The method according to claim 5, characterized in that Analyzing the production process data comprises the following steps: Based on the third clustering result, a plurality of environmental parameters when processing the production raw materials in each of the historical processing durations are obtained, and a coordinate system is established, wherein the coordinate system takes the historical processing duration as the end point of the horizontal axis, takes the first preset value and the second preset value as the end point scale of the vertical axis, and plots actual curves of the changes of different environmental parameters over time in the coordinate system; A template curve is established for each of the environmental parameters, and the similarity between each of the actual curves and each of the template curves is identified. The average value of the similarities is used as the evaluation value corresponding to the historical processing time. The average value of the evaluation values ​​of all the historical processing times in the third clustering result is calculated and used as the state value of the third clustering result. The third clustering result with the largest state value that is greater than a fourth threshold is used as the best result.

7. The method according to claim 6, characterized in that After detecting a predetermined operation, the coordinate system shortens the endpoint scale of the horizontal axis to a second preset value, reduces the endpoint scale of the vertical axis to a third parameter value and a fourth parameter value, increases the second preset value by a preset step size, so as to dynamically display the dynamic change trend of each of the actual curves, and before increasing the second preset value, obtains the maximum parameter value and the minimum parameter value of the environmental parameter in the future time length; if the maximum parameter value and the minimum parameter value are outside the endpoint scales of the vertical axis, sets the endpoint scales of the vertical axis to the maximum parameter value and the minimum parameter value, respectively.

8. The method according to claim 1, characterized in that The twin product with larger out-of-stock data is assigned a longer occupancy time.

9. A data management system for intelligent manufacturing, used to implement the method according to any one of claims 1 to 8, characterized in that: include: A forecasting module, which calculates forecast data based on the historical order data of the product, wherein the forecast data is the demand data of the product in a predetermined time period in the future; A screening module, which calculates out-of-stock data according to the inventory data of the product and the forecast data, and defines the product whose out-of-stock data is greater than a first threshold as a target product; An analysis module, if there are multiple target products, then obtain the production materials and production routes of each target product, the production route includes multiple production equipment, the production materials pass through the production equipment in a preset order to obtain the target product, if the production routes of multiple target products have the same production equipment, then define the target product as a twin product, the corresponding production route is a twin route, analyze the historical production data of the twin route and obtain the production time of each twin product; The suggestion module combines the out-of-stock data and the production time to allocate the occupancy time of each twin product using the twin route, and generates production suggestions according to the occupancy time.

10. A computer storage medium, characterized in that: The computer storage medium stores program instructions, wherein when the program instructions are executed, the device where the computer storage medium is located is controlled to execute the method according to any one of claims 1 to 8.

Images