Shoe production scheduling task allocation algorithm based on big data

Abstract

The present application relates to the field of data processing technology, and specifically to a shoe type production scheduling task allocation algorithm based on big data, the algorithm comprising: obtaining the historical production data set of each production line, and each shoe type and order production quantity corresponding to the current existing production order; obtaining the production speed sequence of each shoe type of each production line; determining the production speed weight, determining the production rate of each shoe type of each production line; determining the historical production density of each shoe type of each production line; determining the production priority of each shoe type in the current existing production order on each production line; constructing a scheduling vector, obtaining the initial spark of the fireworks algorithm; calculating the fitness of each spark, combining the fireworks algorithm for iterative calculation, obtaining the optimal scheduling vector, and determining the shoe type to be produced next time on the production line. The present application can reduce the number of model changes of the production line, enhance the stability of shoe type production on the production line, and improve production efficiency.

Description

Technical Field

[0001] The present application relates to the field of data processing technology, and in particular to a shoe production scheduling task allocation algorithm based on big data. Background Art

[0002] Production scheduling is the process of balancing the production load of machinery and workers and arranging the production order of different production tasks under the condition of certain production resources. A scientific and reasonable production scheduling task allocation algorithm can not only ensure stable and orderly production and improve production efficiency, but also effectively respond to the market environment and enhance the flexibility of production adjustment.

[0003] As the market environment changes, footwear production is constantly shifting towards a personalized and diversified model, and has the characteristics of large orders, small batches, multiple shoe types, and strict delivery requirements, which requires a more flexible and efficient production scheduling task allocation algorithm. At present, the traditional production scheduling task allocation algorithm used by shoemaking companies only independently allocates production tasks based on the current shoe type, and does not make sufficient use of historical data from past production, resulting in an increase in the number of model changes on the production line, insufficient stability in the production of shoe types on the production line, and a decrease in production efficiency. Summary of the invention

[0004] In order to solve the above technical problems, a shoe production scheduling task allocation algorithm based on big data is provided to solve the existing problems.

[0005] The solution to the technical problem of this application is to provide a shoe production scheduling task allocation algorithm based on big data, including the following steps:

[0006] Get the shoe type, order production quantity, and order completion time of each production line after each order is completed, and form the production data vector of each production line at each time. Build a historical production data set based on the production data vector, and get each shoe type and order production quantity corresponding to the current existing production order;

[0007] Based on the order production quantity and order completion time in the production data vector of all the same shoe types in the historical production data set of each production line, the production speed sequence of each shoe type of each production line is obtained;

[0008] Analyze the offset of elements in the neighborhood of each element in the production speed sequence and the production order corresponding to each element, determine the production speed weight of each element in the production speed sequence, and determine the production rate of each shoe type on each production line according to the concentrated distribution of elements in the neighborhood of each element in the production speed sequence and the production speed weight;

[0009] Determine the historical production density of each shoe type on each production line according to the change trend of the production quantity of the same shoe type in the historical production data set of each production line, the number of production data vectors of the same shoe type and the discrete degree of the corresponding production order, combined with the production rate;

[0010] Based on the historical production density, the production priority of each shoe type in the current existing production orders on each production line is determined to screen the shoe types on each production line and construct a scheduling vector as the initial spark of the fireworks algorithm; according to the difference between the shoe type of the component in the scheduling vector corresponding to each spark and the shoe type being produced by the corresponding production line, the fitness of each spark is calculated, and iterative calculation is performed in combination with the fireworks algorithm to obtain the optimal scheduling vector and determine the shoe type to be produced next on the production line.

[0011] Preferably, the step of obtaining the production speed sequence of each shoe type of each production line comprises:

[0012] Calculate the production speed of each production data vector based on the order production quantity and order completion time in each production data vector in the historical production data set of each production line;

[0013] The production speeds of all production data vectors corresponding to the same shoe type in the historical production data set of each production line are combined to form a production speed sequence for each shoe type of each production line.

[0014] Preferably, the step of determining the production speed weight of each element in the production speed sequence of each shoe type on each production line includes:

[0015] Selecting each element in the production speed sequence and multiple elements in its neighborhood to form a local subsequence of each element;

[0016] Calculate the mean of the production speed sequence; record the difference between each element in the local subsequence and the mean as a first difference; record the ratio of the production order corresponding to each element in the local subsequence in the production speed sequence and the first difference as a first ratio;

[0017] The sum of the first ratios of all elements in the local subsequence is used as the production speed weight of each element in the production speed sequence of each shoe type of each production line.

[0018] Preferably, the production rate is expressed as: ,in, For the Production line No. The production rate of the shoe type, , Respectively Production line No. The production speed sequence of the shoe type The production speed weight of each element and the The mean of a local subsequence of elements, For the Production line No. The number of all elements in the production speed sequence of the shoe type, is the normalization function.

[0019] Preferably, the historical production intensity is expressed as: ,in, For the Production line No. The historical production intensity of the shoe type, For the Production line No. The production rate of the shoe type, For the Production line No. The slope of the fitted straight line corresponding to the shoe type, For the Production line No. The number of all production data vectors of shoe types, For the Production line No. The mean of the order production quantity in all production data vectors of the shoe type, For the Production line No. The discrete degree of the production order corresponding to all production data vectors of the shoe type, is an exponential function with a natural constant as base, The default value is greater than 0.

[0020] Preferably, the slope of the fitted straight line includes:

[0021] The production sequence and order production quantity corresponding to each production data vector of the same shoe type in the historical production data set of each production line are formed into a two-dimensional array;

[0022] Linear fitting is performed on all two-dimensional arrays of the same shoe type in the historical production data set of each production line to obtain the slope of the fitting line.

[0023] Preferably, determining the production priority of each shoe type in the current existing production order on each production line includes:

[0024] The normalized result of the sum of the historical production intensities of each shoe type in all production lines is used as the production ratio of each shoe type; the sum of the production ratio and the preset priority value is used as the relative production ratio of each shoe type in the current existing production order;

[0025] If each shoe type in the current existing production order is the same as the shoe type currently being produced by any production line, the maximum value of the historical production density of all shoe types on any production line is used as the relative production density of each shoe type in the current existing production order on any production line; otherwise, the relative production density of each shoe type in the current existing production order on any production line is the historical production density;

[0026] The product of the relative production ratio and the relative production density is used as the production priority of each shoe type in each production line in the current existing production order.

[0027] Preferably, the scheduling vector is constructed by:

[0028] Obtaining a segmentation threshold of the production priority of each shoe type in each production line in the current existing production order; using the shoe type with a production priority greater than the segmentation threshold as a candidate shoe type for each production line;

[0029] A shoe type is randomly selected from the alternative shoe types of each production line as the component of the corresponding dimension in the scheduling vector, where the dimension of the scheduling vector is the same as the number of production lines.

[0030] Preferably, the fitness of each spark is calculated as follows: ,in, For the The fitness of a spark, For the The number of shoe types of all components in the scheduling vector corresponding to the spark that are different from the shoe types being produced by the production line corresponding to the component, For the The first spark in the schedule vector The estimated production time of the corresponding production line for each component to produce the corresponding shoe type, is the number of all production lines, is a value greater than 0, and the estimated production time is The ratio of the order production quantity of the shoe type corresponding to the component of each component in the scheduling vector corresponding to the spark to the production rate.

[0031] Preferably, the step of obtaining the optimal scheduling vector and determining the shoe type to be produced next time on the production line includes:

[0032] Based on the initial spark and the fitness, the fireworks algorithm is used to calculate the upward rounded result of the explosion radius of each spark as the explosion radius of each spark, which is recorded as ;

[0033] Arrange the production priorities of all shoe types on each production line in descending order; for each component of the scheduling vector corresponding to each spark, randomly select the production line that is ranked first. Select any one of the shoe types corresponding to the production priority and perform a displacement operation to obtain a scheduling vector after the displacement operation;

[0034] Calculate the fitness of all scheduling vectors after the displacement operation, arrange them in ascending order, use the scheduling vectors corresponding to the top fitness as the next generation sparks of the fireworks algorithm, and iterate in sequence to obtain the optimal scheduling vector; use the shoe type of each component in the optimal scheduling vector as the shoe type to be produced on the production line next time for each component.

[0035] This application has at least the following beneficial effects:

[0036] The present application analyzes the offset of elements in the neighborhood of each element in the production speed sequence, as well as the production order corresponding to each element, determines the production speed weight, and obtains the production rate of each shoe type on each production line. The beneficial effect is to balance the impact of different historical production data on the production rate calculation in the production line, and improve the accuracy of the production rate calculation of different shoe types on the production line. The application determines the historical production density of each shoe type on each production line based on the changing trend of the order production quantity of the same shoe type in the historical production data set of each production line, as well as the number of production data vectors of the same shoe type and the corresponding production order discreteness, combined with the production rate. The beneficial effect is to take into account the probability of each production line producing each shoe type in the historical orders, so as to reflect the order density of different shoe types, and provide guidance for subsequent production. The scheduling extracts more accurate production information; determines the production priority of each shoe type in each production line in the current existing production order, which has the beneficial effect of considering the idle time of the subsequent production line waiting and the number of times the production line is changed in the production scheduling, thereby improving production efficiency; based on the production priority, constructs a scheduling vector, and obtains the initial spark of the fireworks algorithm; calculates the fitness of each spark, combines the fireworks algorithm for iterative calculation, obtains the optimal scheduling vector, and determines the shoe type to be produced on the production line next time. The beneficial effect is that by determining the initial spark in the fireworks algorithm, the randomness of spark selection is reduced, and the optimization effect of the scheduling vector is improved. Secondly, the fitness of the scheduling vector is calculated to obtain the optimal scheduling vector, which can reduce the number of changes of the production line, enhance the stability of shoe type production on the production line, and improve production efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] The shoe type production scheduling task allocation algorithm based on big data of the present application is further described in detail below in conjunction with the accompanying drawings.

[0038] Figure 1 A flowchart of the steps of the shoe type production scheduling task allocation algorithm based on big data provided in an embodiment of the present application;

[0039] Figure 2 A flowchart of the steps of a method for obtaining the production speed weight of each element in the production speed sequence of each shoe type of each production line provided in an embodiment of the present application;

[0040] Figure 3 A flowchart of the steps of a method for obtaining the production priority of each shoe type in each production line in the current existing production order provided in an embodiment of the present application;

[0041] Figure 4 and Figure 5 This is a Gantt chart of production scheduling before and after the fireworks algorithm improvement provided in the embodiment of the present application. DETAILED DESCRIPTION

[0042] In order to make the purpose, technical solution and advantages of this application more clear, the shoe type production scheduling task allocation algorithm based on big data proposed in this application is further described in detail below in conjunction with the accompanying drawings and implementation examples. It should be understood that the specific embodiments described here are only used to explain this application and are not used to limit this application.

[0043] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs.

[0044] See also Figure 1 , which shows a flowchart of a shoe type production scheduling task allocation algorithm based on big data provided by an embodiment of the present application, the algorithm includes the following steps:

[0045] Step 1: Obtain the historical production data set of each production line, as well as each shoe type and order production quantity corresponding to the current existing production orders.

[0046] Because shoemaking companies only consider the optimization scheduling of the current production line in shoe production scheduling, and ignore the historical information contained in the production data during the historical production process, the production line will change the shoe model more frequently, resulting in a decrease in production efficiency.

[0047] The shoe type, order production quantity, and order completion time are combined to form the production data vector of each production line in each production. The production data vectors after N historical production completions of each production line are selected and arranged in production order to form the historical production data set of each production line.

[0048] Preferably, in this embodiment, the production data vectors after 1000 historical production completions of each production line are selected to form the historical production data set of each production line. As other implementation methods, the implementer can set them according to the actual situation.

[0049] Secondly, based on the shoe types and order production quantities of the existing orders that need to be delivered, each shoe type and order production quantity corresponding to the current existing production orders are obtained.

[0050] At this point, the historical production data set of each production line is obtained, as well as each shoe type and order production quantity corresponding to the current existing production orders.

[0051] Step 2, obtain the production speed sequence of each shoe type for each production line; analyze the offset of elements in the neighborhood of each element in the production speed sequence, and the corresponding production order of each element, determine the production speed weight of each element in the production speed sequence, and determine the production rate of each shoe type for each production line based on the concentrated distribution of elements in the neighborhood of each element in the production speed sequence combined with the production speed weight.

[0052] Since the production efficiency of different production equipment is different, and the production speed of the same production equipment producing the same shoe type at different time stages is also different, the production speed weight is determined by analyzing the production speed of each production line producing different shoe types in the historical production data, specifically:

[0053] Calculate the ratio of the order production quantity to the order completion time in each production data vector in the historical production data set of each production line, and record it as the production speed of each production data vector;

[0054] The production speeds of all production data vectors corresponding to the same shoe type in the historical production data set of each production line are used to form a production speed sequence of each shoe type on each production line;

[0055] Selecting each element in the production speed sequence and multiple elements in its neighborhood to form a local subsequence of each element;

[0056] Preferably, in this embodiment, each element in the production speed sequence is taken as the center, and 15 elements in each element and its neighborhood are selected to form a local subsequence of each element in the production speed sequence. As other implementation methods, the implementer can set it according to the actual situation. It should be noted that if the length of a local subsequence is less than 15, linear interpolation is used to fill it. Assuming that the production speed sequence is , with elements Center, select element and its 15 neighbourhood elements, where the 15 elements are , its constituent elements A local subsequence of The local subsequence is .

[0057] Calculate the mean of the production speed sequence of each shoe type of each production line; record the difference between each element in the local subsequence and the mean as the first difference;

[0058] Recording the ratio of the production order corresponding to each element in the local subsequence in the production speed sequence to the first difference as a first ratio;

[0059] The sum of the first ratios of all elements in the local subsequence of each element in the production speed sequence is used as the production speed weight of each element in the production speed sequence of each shoe type of each production line;

[0060] Preferably, in this embodiment, the method for calculating the production speed weight of each element in the production speed sequence of each shoe type of each production line is: ,in, For the Production line No. The production speed sequence of the shoe type The production speed weight of each element, For the Production line No. The production speed sequence of the shoe type The local subsequence of elements The elements in the production speed sequence correspond to the production order, For the Production line No. The production speed sequence of the shoe type The local subsequence of elements The element value of the element, For the Production line No. The mean of the production speed series of the shoe type, For the Production line No. The production speed sequence of the shoe type The number of all elements in a local subsequence of elements, To preset a value greater than 0 to avoid the denominator being 0, in this embodiment, The value is 0.01; secondly, For the first difference, is the first ratio.

[0061] It should be noted that, in the historical production data set of the production line, the larger the production order corresponding to the production data vector, the shorter the interval between the historical production data and the current production, the more it can reflect the production status of the production line, and the greater the impact on the subsequent predicted production speed. Therefore, the production speed weight is greater; the mean reflects the average speed of each production line producing each type of shoe. The greater the fluctuation of the production speed of each production line, the greater the first difference, and the smaller the effect on predicting the subsequent production speed of the production line. Therefore, the smaller the obtained production speed weight is, and the production speed weight reflects the degree of influence in predicting the production speed of the production line later.

[0062] Further, the flowchart of the steps of the method for obtaining the production speed weight of each element in the production speed sequence of each shoe type of each production line provided in the embodiment of the present application is as follows: Figure 2 shown.

[0063] Further, based on the production speed weight, the production rate of each shoe type on each production line is determined, specifically:

[0064] The production rate of each shoe type on each production line is calculated as: ,in, For the Production line No. The production rate of the shoe type, For the Production line No. The production speed sequence of the shoe type The production speed weight of each element, For the Production line No. The production speed sequence of the shoe type The mean of a local subsequence of elements, For the Production line No. The number of all elements in the production speed sequence of the shoe type, is the normalization function.

[0065] It should be noted that by calculating the mean of the local subsequence and determining the production rate, the influence of different historical production data of the production line on the production rate can be reduced, and the accuracy of the production rate calculation for each shoe type can be improved.

[0066] It should be noted that if a certain shoe type is a new shoe type and does not exist in the historical production data of the production line, the average production rate of all shoe types in the historical production data set of each production line will be used as the production rate of the new shoe type for each production line.

[0067] At this point, the production rate of each shoe type on each production line is obtained.

[0068] Step 3: Determine the historical production density of each shoe type on each production line based on the changing trend of the order production quantity of the same shoe type in the historical production data set of each production line, the number of production data vectors of the same shoe type and the discrete degree of the corresponding production order, combined with the production rate.

[0069] Furthermore, due to the large number of shoe types in the order, the production line needs to constantly adjust the types of shoes it produces. At the same time, the rapid changes in market demand lead to different trends in the change of the order volume of different shoe types, which in turn causes the production line to produce different shoe type orders in the historical production data set at different densities, that is, the production line concentrates on producing a certain shoe type at some time, and reduces the frequency of producing the shoe type at other times.

[0070] Based on the above analysis, the production intensity of each shoe type produced by each production line in the historical production stage is analyzed to determine the historical production intensity, which is as follows:

[0071] Each production data vector of the same shoe type in the historical production data set of each production line corresponds to the production sequence and order production quantity to form a two-dimensional array;

[0072] Perform linear fitting on the two-dimensional array corresponding to all production data vectors of the same shoe type in the historical production data set of each production line to obtain the slope of the fitting line;

[0073] Preferably, in this embodiment, the least squares method is used for linear fitting, wherein the least squares method is a well-known technology and will not be described in detail here. As other implementation methods, the implementer can adopt other methods of the prior art, such as the maximum likelihood estimation method, etc. This embodiment does not impose any special restrictions on this.

[0074] The historical production intensity of each shoe type on each production line is calculated as: ,in, For the Production line No. The historical production intensity of the shoe type, For the Production line No. The production rate of the shoe type, For the Production line No. The slope of the fitted straight line corresponding to the shoe type, For the Production line No. The number of all production data vectors of shoe types, For the Production line No. The mean of the order production quantity in all production data vectors of the shoe type, For the Production line No. The discrete degree of the production order corresponding to all production data vectors of the shoe type, is an exponential function with a natural constant as base, To preset a value greater than 0 to avoid the denominator being 0, in this embodiment, The value is 0.01.

[0075] Preferably, in this embodiment, the standard deviation of the production order corresponding to all production data vectors of the same shoe type in the historical production data set of each production line is calculated. As other implementation methods, the implementer can use other methods of the prior art to measure the degree of discreteness, such as variance, coefficient of variation, information entropy, etc. This embodiment does not impose any special restrictions on this.

[0076] It should be noted that the slope reflects the changing trend of the historical order production quantity of each shoe type produced by each production line. The larger the historical order quantity and the single order production quantity of each shoe type in each production line, the greater its historical production density. Secondly, the greater the production rate, the more shoe types are produced by the corresponding production line in the same period of time, the more orders are completed, and the greater the corresponding historical production density. At the same time, the shorter the interval between different orders of each shoe type in each production line, that is, the smaller the difference between the production sequences, the smaller the obtained standard deviation, and the stronger the order density of the corresponding shoe type, which is reflected in the greater probability of an increase in the order volume of this shoe type in the future, and the greater the obtained historical production density.

[0077] It should be noted that if a certain shoe type is a new shoe type and does not exist in the historical production data of the production line, the average of the historical production density of all shoe types in the historical production data set of each production line will be used as the historical production density of each production line when the new shoe type is added.

[0078] At this point, the historical production intensity of each shoe type on each production line is obtained.

[0079] Step 4: Based on the historical production density, determine the production priority of each shoe type in each production line in the current existing production orders.

[0080] Furthermore, based on the historical production density, the production density of each shoe type in all production lines is analyzed, and the production ratio of each shoe type is determined to reflect the density of each shoe type in the production data of historical orders, so as to extract more accurate production information for subsequent production scheduling, specifically:

[0081] The normalized result of the sum of the historical production intensities of each shoe type in all production lines is taken as the production proportion of each shoe type;

[0082] Preferably, in this embodiment, maximum and minimum value normalization is used for normalization processing.

[0083] It should be noted that the production ratio reflects the probability of orders for each shoe type appearing in the future.

[0084] Furthermore, when a production line completes the production of an order, in order to reduce the completion time of the shoe type order to be produced and reduce the number of production line changes, the production priority of each shoe type in the current existing production order on different production lines is analyzed, specifically:

[0085] The sum of the production ratio of each shoe type in the current existing production order and the preset priority value is used as the relative production ratio of each shoe type in the current existing production order;

[0086] Preferably, in this embodiment, the preset priority value is 1. As other implementation modes, the implementer can set it according to the actual situation. It should be noted that the purpose of setting the preset priority value is to increase the production priority of each shoe type in the existing production order. For other shoe types that are not in the current existing production order, the relative production ratio is the same as the production ratio.

[0087] If each shoe type in the current existing production order is the same as the shoe type currently being produced by any production line, the maximum value of the historical production density of all shoe types on any production line is used as the relative production density of each shoe type in the current existing production order on any production line; if each shoe type in the current existing production order is different from the shoe type currently being produced by any production line, the relative production density of each shoe type in the current existing production order on any production line is the historical production density;

[0088] The product of the relative production ratio and the relative production density is used as the production priority of each shoe type in each production line in the current existing production order;

[0089] It should be noted that the relative production intensity is mainly achieved by adjusting the production intensity of different shoe types in the current existing production orders on the production line to reduce the number of model changes for each production line in the subsequent production schedule; the production priority reflects the production priority of each shoe type in the current existing production orders when the production line is scheduled later. The greater the production proportion of each shoe type, the greater the probability that orders for the corresponding shoe type will appear in the future, the smaller the probability of model change of the production line, and the greater the resulting production priority. Secondly, the greater the historical production intensity of each shoe type in each production line, the greater the frequency of each production line producing the corresponding shoe type in historical orders, the fewer the number of model changes for the production line, the shorter the time to complete the same order production quantity, and the greater the resulting production priority.

[0090] Furthermore, the flowchart of the steps of the method for obtaining the production priority of each shoe type in each production line in the current existing production order provided by the embodiment of the present application is as follows: Figure 3 shown.

[0091] At this point, the production priority of each shoe type in each production line in the current existing production orders is obtained.

[0092] Step 5, based on the production priority, the shoe types of each production line are screened to construct a scheduling vector as the initial spark of the fireworks algorithm; according to the difference between the shoe type of the component in the scheduling vector corresponding to each spark and the shoe type currently being produced by the corresponding production line, the fitness of each spark is calculated, and iterative calculation is performed in combination with the fireworks algorithm to obtain the optimal scheduling vector and determine the shoe type to be produced next time on the production line.

[0093] There are multiple production lines in shoemaking enterprises. When changing the production line, not only the mold needs to be replaced, but also the parameters of each device on the production line need to be re-adjusted, which leads to a decrease in production efficiency. In addition, there are many types of shoes. In order to meet market demand, reasonable production scheduling is needed to reduce the number of times the shoe production line changes and improve production efficiency. In the traditional shoe production scheduling task allocation algorithm, only the completion time of the shoe order is considered. Therefore, in order to improve the overall efficiency of shoe production and reduce the number of production line changes, the fireworks algorithm is used to obtain the optimal scheduling vector.

[0094] First, based on the production priority, the initial spark of the fireworks algorithm is determined, specifically:

[0095] The maximum inter-class variance algorithm is used to obtain the segmentation threshold of the production priority of each shoe type in each production line in the current existing production orders;

[0096] It should be noted that the maximum inter-class variance algorithm is a well-known technology and will not be described in detail here.

[0097] The shoe types with production priority greater than the segmentation threshold are used as candidate shoe types for each production line;

[0098] It should be noted that, assuming that the number of all production lines of the enterprise is N, the scheduling vector is N-dimensional, wherein each dimension of the scheduling vector represents each corresponding production line. In this embodiment, the number of production lines N is 5.

[0099] Randomly select a shoe type from the candidate shoe types of each production line as the component of the corresponding dimension in the scheduling vector, where the scheduling vector dimension is the same as the number of production lines, and obtain a scheduling vector; use the same method to obtain multiple scheduling vectors as the initial sparks of the fireworks algorithm;

[0100] Preferably, in this embodiment, 20 scheduling vectors are used as the initial sparks of the fireworks algorithm. As other implementation methods, implementers can set them according to actual conditions.

[0101] It should be noted that, for the sake of ease of understanding, assume that there are 3 production lines, numbered 1, 2, and 3, and there are 5 types of shoes in the current production orders, namely: A, B, C, D, and E. The alternative shoe types for the first production line are: B, D, and E, the alternative shoe types for the second production line are: A, C, and D, and the alternative shoe types for the third production line are: A, B, and C; randomly select a shoe type from each of the three production lines, and convert the resulting 6 scheduling vectors , , , , , As the initial spark of the fireworks algorithm, each dimension of the schedule vector represents the corresponding numbered production line.

[0102] Furthermore, based on the scheduling vector corresponding to each spark, the fitness of each spark is determined, specifically:

[0103] The ratio of the order production quantity of the shoe type of the corresponding component produced by the production line corresponding to each component in the scheduling vector corresponding to each spark to the production rate is used as the estimated production time of the shoe type of the corresponding component produced by the production line corresponding to each component in the scheduling vector corresponding to each spark;

[0104] The fitness of each spark is calculated as: ,in, For the The fitness of a spark, For the The number of shoe types of all components in the scheduling vector corresponding to the spark that are different from the shoe types being produced by the production line corresponding to the component, To preset a value greater than 0, For the The first spark in the schedule vector The estimated production time of the corresponding production line for each component to produce the corresponding shoe type, is the number of all production lines.

[0105] It should be noted that, assuming that the scheduling vector corresponding to one of the initial sparks , assuming that the current 1st production line is producing D types of shoes, the current 2nd production line is producing C types of shoes, and the current 3rd production line is producing C types of shoes. The dimension corresponding to each component in the scheduling vector represents the 1st, 2nd, and 3rd production lines. Each component in the scheduling vector is B, A, and C respectively. Then the number of different shoe types corresponding to each component and the shoe types being produced by the corresponding production line is 2, that is, The value is 2, which means that when the production line starts the next production according to the shoe type of the scheduling vector, the two production lines need to be changed.

[0106] It should be noted that the fewer the number of model changes in the production line, the higher the production efficiency and the smaller the fitness. At the same time, the shorter the expected production time, the smaller the probability of order overdue and the smaller the resulting fitness.

[0107] Based on the initial spark and the fitness, the fireworks algorithm is used to calculate the upward rounded result of the explosion radius of each spark as the explosion radius of each spark. ;

[0108] It should be noted that the method of calculating the explosion radius by the fireworks algorithm is a well-known technique and will not be described in detail here. For the The explosion radius of the sparks.

[0109] Arrange the production priorities of all shoe types on each production line in descending order;

[0110] For each production line corresponding to each component in the scheduling vector corresponding to each spark, randomly select the production line with the highest order. Select any one of the shoe types corresponding to the production priority and perform a displacement operation to obtain a scheduling vector after the displacement operation;

[0111] It should be noted that the displacement operation in the fireworks algorithm is a well-known technique and will not be described in detail here. Assuming that the scheduling vector corresponding to an initial spark is , the production priorities of all shoe types in the first production line are arranged in descending order: D, B, E, C, A, the production priorities of all shoe types in the second production line are arranged in descending order: C, A, D, B, E, and the production priorities of all shoe types in the third production line are arranged in descending order: C, B, A, E, D. Assuming that the explosion radius R corresponding to the initial spark is 2, randomly select a D from the top two shoe types D and B in the first production line, randomly select an A from the top two shoe types C and A in the second production line, and randomly select a B from the top two shoe types C and B in the third production line. As the scheduling vector after the shift operation.

[0112] Calculate the fitness of all scheduling vectors after the shift operation, arrange them in ascending order, use the scheduling vectors corresponding to the top fitness as the next generation of sparks of the fireworks algorithm, and iterate them in sequence to obtain the optimal scheduling vector;

[0113] Preferably, in this embodiment, the scheduling vectors corresponding to the top 20 fitness values ​​are used as the next generation sparks of the fireworks algorithm. Secondly, the maximum number of iterations of the fireworks algorithm is set to 20. As other implementation methods, implementers can set them according to actual conditions.

[0114] Furthermore, the Gantt chart of the production schedule before and after the fireworks algorithm improvement provided in the embodiment of the present application is as follows: Figure 4 and Figure 5 As shown, Figure 4 To improve the previous production schedule Gantt chart, Figure 5 This is the Gantt chart of the production schedule before improvement. The horizontal axis is the time required to complete the existing production orders, and the vertical axis is the equipment code corresponding to each production line. Each black bar represents each shoe type order, and the number corresponding to each black bar represents each shoe type. Figure 4 and Figure 5 , within the same production time, Figure 5 The production schedule shown can produce more orders while reducing the number of changeovers on each production line and improving production efficiency.

[0115] According to the optimal scheduling vector, the shoe type corresponding to each production line in the optimal scheduling vector is used as the shoe type to be produced next time by the corresponding production line, thereby completing the production scheduling of different shoe types.

[0116] It should be understood that although Figure 1 The steps in the flowchart are shown in sequence as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified in this document, there is no strict order restriction for the execution of these steps, and these steps can be executed in other orders. Moreover, Figure 1 At least part of the steps may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily executed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be executed in turn or alternately with other steps or at least part of the sub-steps or stages of other steps.

[0117] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above 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.

[0118] The above-mentioned embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the present application. It should be pointed out that, for those of ordinary skill in the art, several modifications and improvements can be made without departing from the concept of the present application. Therefore, any simple modification, equivalent change and modification made to the above embodiments according to the technical essence of the present application without departing from the content of the technical solution of the present application, shall fall within the protection scope of the technical solution of the present application.

Claims

1. The shoe production scheduling task allocation algorithm based on big data is characterized by: The algorithm consists of the following steps: Get the shoe type, order production quantity, and order completion time of each production line after each order is completed, and form the production data vector of each production line at each time. Build a historical production data set based on the production data vector, and get each shoe type and order production quantity corresponding to the current existing production order; Based on the order production quantity and order completion time in the production data vector of all the same shoe types in the historical production data set of each production line, the production speed sequence of each shoe type of each production line is obtained; Analyze the offset of elements in the neighborhood of each element in the production speed sequence and the production order corresponding to each element, determine the production speed weight of each element in the production speed sequence, and determine the production rate of each shoe type on each production line according to the concentrated distribution of elements in the neighborhood of each element in the production speed sequence and the production speed weight; Determine the historical production density of each shoe type on each production line according to the change trend of the production quantity of the same shoe type in the historical production data set of each production line, the number of production data vectors of the same shoe type and the discrete degree of the corresponding production order, combined with the production rate; Based on the historical production density, the production priority of each shoe type in the current existing production order on each production line is determined to screen the shoe types of each production line and construct a scheduling vector as the initial spark of the fireworks algorithm; according to the difference between the shoe type of the component in the scheduling vector corresponding to each spark and the shoe type being produced by the corresponding production line, the fitness of each spark is calculated, and iterative calculation is performed in combination with the fireworks algorithm to obtain the optimal scheduling vector and determine the shoe type to be produced next time on the production line; The fitness of each spark is calculated as follows: ,in, For the The fitness of a spark, For the The number of shoe types of all components in the scheduling vector corresponding to the spark that are different from the shoe types being produced by the production line corresponding to the component, For the The first spark in the schedule vector The estimated production time of the corresponding production line for each component to produce the corresponding shoe type, is the number of all production lines, is a value greater than 0, and the estimated production time is The ratio of the order production quantity of the shoe type corresponding to the component of each component in the scheduling vector corresponding to the spark to the production rate.

2. The shoe type production scheduling task allocation algorithm based on big data as claimed in claim 1 is characterized in that: The step of obtaining the production speed sequence of each shoe type of each production line comprises: Calculate the production speed of each production data vector based on the order production quantity and order completion time in each production data vector in the historical production data set of each production line; The production speeds of all production data vectors corresponding to the same shoe type in the historical production data set of each production line are combined to form a production speed sequence for each shoe type of each production line.

3. The shoe type production scheduling task allocation algorithm based on big data as claimed in claim 1 is characterized in that: The step of determining the production speed weight of each element in the production speed sequence of each shoe type of each production line includes: Selecting each element in the production speed sequence and multiple elements in its neighborhood to form a local subsequence of each element; Calculate the mean of the production speed sequence; record the difference between each element in the local subsequence and the mean as a first difference; record the ratio of the production order corresponding to each element in the local subsequence in the production speed sequence and the first difference as a first ratio; The sum of the first ratios of all elements in the local subsequence is used as the production speed weight of each element in the production speed sequence of each shoe type of each production line.

4. The shoe type production scheduling task allocation algorithm based on big data as claimed in claim 3 is characterized in that: The production rate is expressed as: ,in, For the Production line No. The production rate of the shoe type, , Respectively Production line No. The production speed sequence of the shoe type The production speed weight of each element and the The mean of a local subsequence of elements, For the Production line No. The number of all elements in the production speed sequence of the shoe type, is the normalization function.

5. The shoe type production scheduling task allocation algorithm based on big data as claimed in claim 1 is characterized in that: The historical production intensity is expressed as: ,in, For the Production line No. The historical production intensity of the shoe type, For the Production line No. The production rate of the shoe type, For the Production line No. The slope of the fitted straight line corresponding to the shoe type, For the Production line No. The number of all production data vectors of shoe types, For the Production line No. The mean of the order production quantity in all production data vectors of the shoe type, For the Production line No. The discrete degree of production order corresponding to all production data vectors of the shoe type, is an exponential function with a natural constant as base, The default value is greater than 0.

6. The shoe type production scheduling task allocation algorithm based on big data as claimed in claim 5 is characterized in that: The slope of the fitted straight line includes: The production sequence and order production quantity corresponding to each production data vector of the same shoe type in the historical production data set of each production line are formed into a two-dimensional array; Linear fitting is performed on all two-dimensional arrays of the same shoe type in the historical production data set of each production line to obtain the slope of the fitting line.

7. The shoe type production scheduling task allocation algorithm based on big data as claimed in claim 1 is characterized in that: Determining the production priority of each shoe type in the current existing production order on each production line includes: The normalized result of the sum of the historical production intensities of each shoe type in all production lines is used as the production ratio of each shoe type; the sum of the production ratio and the preset priority value is used as the relative production ratio of each shoe type in the current existing production order; If each shoe type in the current existing production order is the same as the shoe type currently being produced by any production line, the maximum value of the historical production density of all shoe types on any production line is used as the relative production density of each shoe type in the current existing production order on any production line; otherwise, the relative production density of each shoe type in the current existing production order on any production line is the historical production density; The product of the relative production ratio and the relative production density is used as the production priority of each shoe type in each production line in the current existing production order.

8. The shoe type production scheduling task allocation algorithm based on big data as claimed in claim 1 is characterized in that: The scheduling vector is constructed as follows: Obtaining a segmentation threshold of the production priority of each shoe type in each production line in the current existing production order; using the shoe type with a production priority greater than the segmentation threshold as a candidate shoe type for each production line; A shoe type is randomly selected from the alternative shoe types of each production line as the component of the corresponding dimension in the scheduling vector, where the dimension of the scheduling vector is the same as the number of production lines.

9. The shoe type production scheduling task allocation algorithm based on big data as claimed in claim 1, characterized in that: The step of obtaining the optimal scheduling vector and determining the shoe type to be produced next on the production line includes: Based on the initial spark and the fitness, the fireworks algorithm is used to calculate the upward rounded result of the explosion radius of each spark as the explosion radius of each spark, which is recorded as ; Arrange the production priorities of all shoe types on each production line in descending order; for each component of the scheduling vector corresponding to each spark, randomly select the production line that is ranked first. Select any one of the shoe types corresponding to the production priority and perform a displacement operation to obtain a scheduling vector after the displacement operation; Calculate the fitness of all scheduling vectors after the displacement operation, arrange them in ascending order, use the scheduling vectors corresponding to the top fitness as the next generation of sparks of the fireworks algorithm, and iterate in sequence to obtain the optimal scheduling vector; use the shoe type of each component in the optimal scheduling vector as the shoe type to be produced on the production line next time for each component.

Images