Intelligent production scheduling method and system

Abstract

The invention belongs to the technical field of factory production scheduling processing, and discloses an intelligent production scheduling method and system, and the method comprises the steps: generating at least two first production scheduling schemes according to historical production scheduling data and an intelligent process model; performing distribution processing on all the first production scheduling schemes based on preset personnel skill data to obtain at least two second production scheduling schemes; processing each second production scheduling scheme based on an intelligent matching model to obtain a corresponding production index, and sorting the corresponding second production scheduling schemes based on all the production indexes to obtain a target production scheduling scheme; the historical production scheduling data and the preset personnel skill data are processed in combination with an artificial intelligence processing technology, so that a production scheduling scheme which is more efficient and reasonable in distribution is obtained, and the problem that the capacity expansion requirement cannot be met due to the limitation of the number of workers can be effectively solved.

Description

Technical Field

[0001] The present application belongs to the field of factory production scheduling technology, and in particular, relates to an intelligent production scheduling method and system. Background Art

[0002] In the garment factories of traditional garment manufacturing companies, the assembly line layout needs to comprehensively consider the mutual influence of multiple factors such as process rules and personnel arrangements, which makes the assembly line layout mostly rely on experience or manual adjustments, which in turn causes the output of the assembly line to be overly dependent on the production scheduling level of the on-site team leader.

[0003] When factories continue to expand the production capacity of existing garment workshops, they are easily limited by the number of on-site team leaders, resulting in the existing layout being unable to meet the needs of capacity expansion and unable to effectively guarantee the production efficiency of the production lines. Summary of the invention

[0004] In order to solve the above-mentioned technical problems, this application proposes an intelligent production scheduling method and system, and its technical solution is as follows: In a first aspect, an embodiment of the present application provides an intelligent production scheduling method, including: Generate at least two first production scheduling plans according to the historical production scheduling data and the intelligent process model; wherein the historical production scheduling data includes at least two historical production scheduling plans, each historical production scheduling plan includes at least two standard processes arranged in a scheduling order, and each first production scheduling plan includes at least two standard processes arranged in a specified order; Allocate and process all first production scheduling plans based on preset personnel skill data to obtain at least two second production scheduling plans; wherein each second production scheduling plan includes at least two standard processes arranged in a specified order and a configuration personnel corresponding to each standard process; Each second production schedule is processed based on the intelligent matching model to obtain corresponding production indicators, and the corresponding second production schedules are sorted based on all production indicators to obtain a target production schedule.

[0005] In an optional solution of the first aspect, the historical production scheduling data further includes equipment parameters corresponding to each standard process; Generate at least two first production scheduling plans based on historical production scheduling data and intelligent process models, including: Based on a preset process coding table, all standard processes included in each historical production scheduling scheme in the historical production scheduling data are coded to obtain corresponding first coding values; Based on a preset parameter coding table, the equipment parameters corresponding to all standard processes included in each historical production scheduling plan are coded to obtain corresponding second coding values; The first coding value and the second coding value corresponding to each historical production scheduling plan are integrated to obtain at least two process coding values, and at least two first production scheduling plans are generated based on all process coding values ​​and the intelligent process model.

[0006] In another optional solution of the first aspect, the first code value and the second code value corresponding to each historical production scheduling plan are integrated to obtain at least two process code values, including: Integrate the first code value and the second code value corresponding to each historical production scheduling plan to obtain a corresponding third code value; Performing feature extraction processing on any at least two third code values ​​to obtain at least two difference feature codes; A preset classification model is trained based on all difference feature codes, and all third code values ​​are input into the preset classification model to obtain at least two process code values.

[0007] In yet another optional solution of the first aspect, at least two first production scheduling solutions are generated according to all process code values ​​and the intelligent process model, including: All process code values ​​are marked in production order, and the intelligent process model is trained based on all the marked process code values; Based on the preset process coding table, the target process table is coded to obtain the target coding value, and the target coding value is input into the trained intelligent process model to obtain the process matrix; The process matrix is ​​processed based on a deep search algorithm to obtain at least two first production scheduling plans.

[0008] In another optional solution of the first aspect, the preset personnel skill data includes process skill parameters corresponding to at least two standard processes for each process personnel; All first production scheduling plans are allocated and processed based on the preset personnel skill data to obtain at least two second production scheduling plans, including: Determine the similar processes corresponding to each standard process of each process personnel, and count the process skill parameters of all process personnel corresponding to the similar processes; Perform standard deviation calculation on the process skill parameters of all process personnel corresponding to similar processes to obtain the corresponding parameter standard deviation, and clean the process skill parameters of the corresponding standard process based on the parameter standard deviation; Based on the process skill parameters of each standard process of each process personnel after the cleaning process, all the first production scheduling plans are allocated and processed to obtain at least two second production scheduling plans.

[0009] In another optional solution of the first aspect, all first production scheduling plans are allocated based on the process skill parameters of each standard process of each process personnel after cleaning to obtain at least two second production scheduling plans, including: Based on the preset bandwidth parameters and the process skill parameters of each process personnel after each standard process has been cleaned, a Gaussian kernel density function is constructed, and the corresponding process skill efficiency is generated according to the Gaussian kernel density function; All first production scheduling plans are allocated and processed based on preset personnel skill data and all process skill efficiencies to obtain at least two second production scheduling plans.

[0010] In another optional solution of the first aspect, each second production scheduling plan is processed based on the intelligent matching model to obtain corresponding production indicators, including: Input the process personnel information configured for each standard process in each second production scheduling plan into the intelligent matching model to obtain the corresponding personnel skill simulation index; Generate corresponding production indicators based on the personnel skill simulation indicators of each second production scheduling plan.

[0011] In another optional solution of the first aspect, after the corresponding second production scheduling solutions are sorted based on all production indicators to obtain the target production scheduling solution, the method further includes: The target production schedule is executed based on user feedback information, and the target production schedule is updated according to the production line operation data corresponding to the target production schedule.

[0012] In a second aspect, an embodiment of the present application provides an intelligent production scheduling system, including: A data acquisition module, used to generate at least two first production scheduling plans according to historical production scheduling data and an intelligent process model; wherein the historical production scheduling data includes at least two historical production scheduling plans, each historical production scheduling plan includes at least two standard processes arranged in a scheduling order, and each first production scheduling plan includes at least two standard processes arranged in a specified order; A plan generation module is used to allocate and process all first production scheduling plans based on preset personnel skill data to obtain at least two second production scheduling plans; wherein each second production scheduling plan includes at least two standard processes arranged in a specified order and a configuration personnel corresponding to each standard process; The plan determination module is used to process each second production scheduling plan based on the intelligent matching model to obtain corresponding production indicators, and to sort the corresponding second production scheduling plans based on all production indicators to obtain a target production scheduling plan.

[0013] In a third aspect, an embodiment of the present application provides a computer storage medium, which stores a computer program. The computer program includes program instructions. When the program instructions are executed by a processor, the intelligent production scheduling method provided by the first aspect of the embodiment of the present application or any one of the implementation methods of the first aspect can be implemented.

[0014] In an embodiment of the present application, when performing intelligent production scheduling, at least two first production scheduling plans can be generated based on historical production scheduling data and intelligent process models; all first production scheduling plans are allocated and processed based on preset personnel skill data to obtain at least two second production scheduling plans; each second production scheduling plan is processed based on the intelligent matching model to obtain corresponding production indicators, and the corresponding second production scheduling plans are sorted based on all production indicators to obtain a target production scheduling plan; historical production scheduling data and preset personnel skill data are processed by combining artificial intelligence processing technology to obtain a more efficient and reasonably allocated production scheduling plan, and it can also effectively overcome the problem of being unable to meet production capacity expansion needs due to labor restrictions. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0016] Figure 1 An overall flow chart of an intelligent production scheduling method provided in an embodiment of the present application; Figure 2 A schematic diagram of the structure of an intelligent production scheduling system provided in an embodiment of the present application; Figure 3 A schematic diagram of the structure of another intelligent production scheduling system provided in an embodiment of the present application. DETAILED DESCRIPTION

[0017] The technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application.

[0018] In the following introduction, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance. The following introduction provides multiple embodiments of the present application, and different embodiments can be replaced or combined, so the present application can also be considered to include all possible combinations of the same and / or different embodiments recorded. Therefore, if one embodiment includes features A, B, and C, and another embodiment includes features B and D, then the present application should also be considered to include embodiments containing one or more of all other possible combinations of A, B, C, and D, although the embodiment may not be clearly recorded in the following text.

[0019] The following description provides examples and does not limit the scope, applicability or examples set forth in the claims. Changes may be made to the functions and arrangements of the elements described without departing from the scope of the present application. Various processes or components may be appropriately omitted, substituted or added to each example. For example, the described method may be performed in an order different from the order described, and various steps may be added, omitted or combined. In addition, features described in some examples may be combined in other examples.

[0020] See also Figure 1 , Figure 1 An overall flow chart of an intelligent production scheduling method provided in an embodiment of the present application is shown.

[0021] like Figure 1 As shown, the intelligent production scheduling method may at least include the following steps: Step 102: Generate at least two first production scheduling plans based on historical production scheduling data and the intelligent process model.

[0022] In an embodiment of the present application, the intelligent production scheduling method can be but is not limited to being applied to a control terminal in a garment factory. The control terminal can store and process all historical production scheduling data of each garment workshop in the garment factory according to a preset hierarchical structure. The preset hierarchical structure can be but is not limited to product types, for example, historical production scheduling data corresponding to different product types such as crew neck sweaters or hoodies can be stored and processed. Here, the historical production scheduling data may include at least two production scheduling plans, each of which may specifically include order number, order receipt date, order opening date, number of pieces produced, product description, process table, process production sequence, process configuration personnel, number of pieces produced in each process, production time for each piece, running time, daily output, number of defective products and reasons, or line balance and at least one or more data, wherein the process table may include all standard processes for preparing products, for example, it may include standard processes of splicing, reinforcement, piping, nailing or template types, and a set of standard action flows are set for each standard process, for example, it may include standard actions such as picking, placing, flattening, cutting, aligning, sewing or thread cutting, so as to improve the production efficiency and quality of each standard process through the standard actions.

[0023] It is understandable that in order to avoid the problem that the existing layout cannot meet the needs of capacity expansion due to the limitation on the number of on-site team leaders, the control terminal can combine the clothing needs and artificial intelligence processing technology to process the historical production scheduling data and preset personnel skill data to obtain a more efficient and reasonably allocated production scheduling plan, and it can also effectively overcome the problem of not being able to meet the needs of capacity expansion due to the limitation on the number of labor.

[0024] Specifically, when performing intelligent production scheduling, the control terminal can, but is not limited to, determine the process table corresponding to each historical production scheduling plan from the historical production scheduling data, and the process table records all the standard processes corresponding to the corresponding historical production scheduling plan when it is executed, and the recording method can be the production order specified when all standard processes are executed. It can be understood that after determining the process table corresponding to each historical production scheduling plan, it can also be, but is not limited to, query all the standard actions corresponding to each standard process according to the preset process action table, so as to facilitate the subsequent rapid encoding processing. Here, the preset process action table may include all the standard processes for the garment making area and all the standard actions corresponding to each standard process, and all the standard actions can be arranged in a specified execution order. For example, in the preset process action table, the standard action sequence corresponding to the block process can be queried, which is pick-up-move-align-place-sew-cut thread.

[0025] Further, after determining the process table corresponding to each historical production schedule, the control terminal can convert all standard processes included in each historical production schedule into corresponding code values ​​based on the preset process code table, and can splice the code values ​​corresponding to each standard process according to the arrangement order of all standard processes to obtain the process code values ​​corresponding to each historical production schedule. Among them, the preset process code table may include all standard processes in the garment making area and the code values ​​corresponding to each standard process. The code values ​​corresponding to each standard process can be generated, but not limited to, according to the code values ​​corresponding to all standard tasks included in the standard process. For example, the standard action sequence corresponding to the block process mentioned above is pick-up-move-align-place-sew-trim, and each standard action can correspond to a fixed code value. Then the code value corresponding to the block process can be understood as the code value composed of pick-up-move-align-place-sew-trim.

[0026] As an option in the embodiment of the present application, the historical production scheduling data also includes equipment parameters corresponding to each standard process; Generate at least two first production scheduling plans based on historical production scheduling data and intelligent process models, including: Based on a preset process coding table, all standard processes included in each historical production scheduling scheme in the historical production scheduling data are coded to obtain corresponding first coding values; Based on a preset parameter coding table, the equipment parameters corresponding to all standard processes included in each historical production scheduling plan are coded to obtain corresponding second coding values; The first coding value and the second coding value corresponding to each historical production scheduling plan are integrated to obtain at least two process coding values, and at least two first production scheduling plans are generated based on all process coding values ​​and the intelligent process model.

[0027] In order to more clearly distinguish the process code values ​​generated by each historical production schedule, so as to facilitate subsequent data processing, the equipment parameters corresponding to each historical production schedule can also be, but not limited to, encoded, so as to combine the code values ​​corresponding to the standard processes to obtain the process code values ​​corresponding to each historical production schedule. Among them, the equipment parameters can be, but not limited to, the equipment type, the processing part of the garment piece, and the abbreviation corresponding to each standard process in each historical production schedule. The equipment type can be specifically divided into flat car, cock car or bar tack car, etc., the processing part of the garment piece can be specifically divided into waist, hat or pocket, etc., and the abbreviation can be specifically divided into side pocket opening, body or folded hem, etc.

[0028] Specifically, when generating the process code value corresponding to each historical production scheduling plan, all standard processes included in each historical production scheduling plan can be converted into corresponding code values ​​based on the preset process code table mentioned above, and the code value corresponding to each standard process can be spliced ​​according to the arrangement order of all standard processes to obtain the first code value corresponding to each historical production scheduling plan. The first code value can be understood as the code value corresponding to the action sequence of all standard processes in each historical production scheduling plan.

[0029] Then, based on the preset parameter coding table, the equipment parameters corresponding to each standard process in each historical production scheduling plan can be converted into corresponding coding values, and the converted coding values ​​can be spliced ​​in the specified order to obtain the second coding value corresponding to each production scheduling plan. Among them, the preset parameter coding table can at least include the three equipment parameters mentioned above, and the coding values ​​corresponding to all the numerical values ​​contained in each equipment parameter. Then, after obtaining the first coding value and the second coding value corresponding to each historical production scheduling plan, each first coding value and the corresponding second coding value can be integrated to generate the process code corresponding to each historical production scheduling plan. Here, the integration processing method can be but not limited to dividing each first coding value into a standard action coding value corresponding to each standard process, and dividing each second coding value into an equipment parameter coding value corresponding to each standard process, and then the standard action coding value and the equipment parameter coding value corresponding to the same standard process can be spliced, and the spliced ​​coding values ​​of all standard processes are spliced ​​again according to the arrangement order corresponding to all standard processes to generate the process code corresponding to each historical production scheduling plan.

[0030] As another option of the embodiment of the present application, the first coding value and the second coding value corresponding to each historical production scheduling plan are integrated to obtain at least two process coding values, including: Integrate the first code value and the second code value corresponding to each historical production scheduling plan to obtain a corresponding third code value; Performing feature extraction processing on any at least two third code values ​​to obtain at least two difference feature codes; A preset classification model is trained based on all difference feature codes, and all third code values ​​are input into the preset classification model to obtain at least two process code values.

[0031] Since the amount of data in historical production scheduling plans is too large, in order to ensure the validity of the overall data and simplify the amount of data for subsequent processing, the process code values ​​corresponding to all historical production scheduling plans can also be classified through data classification to obtain process code values ​​with relatively small data volume and reliability.

[0032] Specifically, each first code value and the corresponding second code value can be integrated to generate a third code value corresponding to each historical production scheduling plan. The integration processing method can be to divide each first code value into a standard action code value corresponding to each standard process, and to divide each second code value into an equipment parameter code value corresponding to each standard process. Then, the standard action code value and the equipment parameter code value corresponding to the same standard process can be spliced, and the spliced ​​code values ​​of all standard processes can be spliced ​​again according to the arrangement order corresponding to all standard processes to generate a process code corresponding to each historical production scheduling plan.

[0033] Next, after obtaining the third coding value corresponding to each historical production scheduling plan, the bag-of-words algorithm can be used to extract at least two standard action coding value sets from the third coding values ​​corresponding to all historical production scheduling plans, and the equipment parameter coding value corresponding to each standard action coding value in each standard action coding value set can be extracted based on all third coding values, and each standard action coding value set and the corresponding equipment parameter coding value can be used as a difference feature code. Here, each standard action coding value set can include at least two standard action coding values ​​with high similarity or the same. For example, standard process A includes the abcde action sequence, and standard process B includes the acdfg action sequence. There are three identical standard actions between the two standard processes. The standard action coding values ​​corresponding to the two standard processes can be regarded as standard action coding values ​​with greater similarity, and the standard action coding values ​​corresponding to the two standard processes can be used together as a standard action coding value set. Next, after determining a plurality of difference feature codes, some difference feature codes can be divided out from all the difference feature codes, and the equipment parameter code values ​​corresponding to the part of the difference feature codes can be manually labeled by the on-site team leader, and the preset classification model can be trained in combination with the equipment parameter code values ​​in the part of the difference feature codes, the corresponding standard action code values, and the labeling processing results. Here, the labeling processing results may include the labeled standard action code values, and the similarity between the corresponding standard action code values ​​and the labeled standard action code values. It can be understood that the preset classification model is a conventional machine learning model, such as a decision tree classification model, a random forest classification model, or a K nearest neighbor classification model, etc., but will not be elaborated here.

[0034] Next, the equipment parameter coding values ​​of all the third coding values ​​mentioned above can be input into the trained preset classification model respectively to obtain the corresponding classified standard action coding values ​​and similarities, and all standard action coding values ​​whose similarity exceeds the set threshold can be used as at least two process coding values, but are not limited to.

[0035] As another option of the embodiment of the present application, at least two first production scheduling plans are generated according to all process code values ​​and the intelligent process model, including: All process code values ​​are marked in production order, and the intelligent process model is trained based on all the marked process code values; Based on the preset process coding table, the target process table is coded to obtain the target coding value, and the target coding value is input into the trained intelligent process model to obtain the process matrix; The process matrix is ​​processed based on a deep search algorithm to obtain at least two first production scheduling plans.

[0036] Specifically, after obtaining multiple process code values, the control terminal can mark the standard process production arrangement sequence corresponding to each process code value, and train the intelligent process model according to all process code values ​​and the arrangement sequence marking results corresponding to each process code value. Here, the intelligent process model can be a recurrent neural network, which predicts the corresponding process matrix according to the input process code value. The horizontal and vertical directions of the process matrix can include all standard processes included in the input process code value. For example, when the input process code value includes n standard processes, the output process matrix can be an n*n structure, and the process matrix can include all feasible production arrangement sequences of the n standard processes.

[0037] Then, after training the intelligent process model, the control terminal can input the process code value corresponding to the target process table into the intelligent process model to obtain the corresponding process matrix, and use the deep search algorithm to traverse the corresponding process matrix to generate all theoretically feasible first production scheduling plans.

[0038] It is understandable that in the process of building an intelligent process model, it is also possible but not limited to using a convolutional neural network trained with a historical process library as a process encoder to extract and produce process code values; wherein, when training the convolutional neural network, the correct process classification code can be obtained from the historical vehicle scheduling plan and on-site investigation and sorting as training data, and the key fields describing the name of the process (such as cock, flat car, collar, waist wrap, etc.) are normalized and formatted as the convolutional network input tensor. Then, the order of the processes can also be reordered by a process sequencer composed of a long short-term memory network or a recurrent neural network; wherein, when training the long short-term memory network, the correct process sequence in the historical vehicle scheduling plan can be converted by the process encoder as the coding sequence as the training data, and reasonable LSTM units and reasonable LSTM network layers can be set to achieve training.

[0039] Of course, in the embodiment of the present application, in order to further ensure the rationality and effectiveness of all process code values, it is also possible but not limited to cleaning all process code values ​​through recording data such as operating time, number of configured personnel and daily output, so as to further control the amount of data corresponding to all process code values ​​while facilitating the improvement of the training accuracy of subsequent intelligent process models.

[0040] Step 104: Allocate and process all first production scheduling plans based on preset personnel skill data to obtain at least two second production scheduling plans.

[0041] Specifically, after generating all theoretically feasible first production scheduling plans, the control terminal can perform personnel configuration processing on each first production scheduling plan according to the preset personnel skill data, so as to obtain multiple second production scheduling plans in which each standard process corresponds to a configured personnel, and each first production scheduling plan may correspond to one or more second production scheduling plans, and the configured personnel corresponding to each standard process in each second production scheduling plan must meet the skill attribute requirements of the corresponding standard process, or must meet the skill efficiency requirements of the corresponding standard process. Among them, the preset personnel skill data can be the personnel data in the historical production scheduling data mentioned above, for example, divided into process skill parameters and process skill attributes corresponding to each process personnel and at least two standard processes, and the process skill parameters can be used to judge whether the corresponding process personnel meet the skill efficiency requirements of the standard process, or the process skill attributes can be used to judge whether the corresponding process personnel meet the skill attribute requirements of the standard process. Here, the process skill parameters can be any one of manually reported skills, metered timing, working time per piece, or work output. The manually reported skills can be understood as the process proficiency or process completion efficiency manually filled in by the on-site team leader based on his own understanding of the capabilities of employees in each process; the process skill attributes can be understood as the skill attribute requirements corresponding to each standard process, such as the gender of the process personnel or the process mastery certificate.

[0042] As another option of the embodiment of the present application, the preset personnel skill data includes process skill parameters corresponding to at least two standard processes for each process personnel; All first production scheduling plans are allocated and processed based on the preset personnel skill data to obtain at least two second production scheduling plans, including: Determine the similar processes corresponding to each standard process of each process personnel, and count the process skill parameters of all process personnel corresponding to the similar processes; Perform standard deviation calculation on the process skill parameters of all process personnel corresponding to similar processes to obtain the corresponding parameter standard deviation, and clean the process skill parameters of the corresponding standard process based on the parameter standard deviation; Based on the process skill parameters of each standard process of each process personnel after the cleaning process, all the first production scheduling plans are allocated and processed to obtain at least two second production scheduling plans.

[0043] In order to improve the efficiency and accuracy of personnel allocation for each standard process, the process skill efficiency of the corresponding standard process can be calculated based on the process skill parameters of each process personnel, so as to combine the preset personnel skill data and all process skill efficiencies to quickly and efficiently allocate and process all standard processes in the first production schedule.

[0044] Specifically, based on all the standard processes configured by each process personnel, similar processes with high similarity to each standard process can be determined. The method for determining the similarity can refer to the similarity calculation method or similarity prediction model well known in the art, which will not be elaborated here. After determining the similar processes corresponding to each standard process of each process personnel, the process skill parameters of all process personnel corresponding to the similar processes are counted. Since the process skill parameters of all process personnel satisfy the rules of normal distribution, it is possible to judge whether there are any abnormalities in the process skill parameters corresponding to the corresponding process personnel based on the process skill parameters of all process personnel, and the abnormal parameters can be cleaned up in time.

[0045] Next, after counting the process skill parameters of all process personnel corresponding to similar processes, the standard deviation of the process skill parameters of all process personnel can be calculated to obtain the corresponding parameter standard deviation, and all parameters that are greater than or less than the parameter standard deviation (or the product of the parameter standard deviation and the specified multiple) in the process skill parameters corresponding to the corresponding process personnel will be regarded as abnormal values ​​and cleaned in time.

[0046] Next, after cleaning the process skill parameters corresponding to each standard process of each process personnel, a Gaussian kernel density function can be constructed in combination with the preset bandwidth parameters and the Gaussian kernel density function formula, and based on the Gaussian kernel density function, the range of specified values ​​that differ from the highest peak value is used as the corresponding process skill efficiency.

[0047] Next, after determining the process skill efficiency corresponding to each standard process of each process personnel, all first production scheduling plans can be allocated and processed according to all process skill efficiencies and the process skill attributes in the preset personnel skill data, so as to effectively ensure that the process personnel configured for each standard process can meet the corresponding process requirements.

[0048] It should be noted that after obtaining multiple second production scheduling plans, in order to effectively ensure the rationality of each second production scheduling plan, all second production scheduling plans can also be screened. Here, when screening all second production scheduling plans, the line balance rate of the corresponding second production scheduling plan can be calculated based on, but not limited to, the process skill efficiency of the process personnel configured for each standard process included in each second production scheduling plan, and when it is detected that the line balance rate is lower than 30% of the target line balance rate, the second production scheduling plan can also be determined as an unreasonable plan, and then the second production scheduling plan can be cleaned. Here, the method of calculating the line balance rate of the corresponding second production scheduling plan can be specifically to obtain the third ratio corresponding to each standard process according to the standard working time per piece corresponding to each standard process included in the second production scheduling plan and the process skill efficiency of the configured process personnel, and then calculate the mean of all third ratios to obtain the corresponding average third ratio, and calculate the ratio of the average third ratio to the maximum third ratio, so as to compare the ratio calculation result with 30% of the target line balance rate.

[0049] Step 106: Process each second production schedule based on the intelligent matching model to obtain corresponding production indicators, and sort the corresponding second production schedules based on all production indicators to obtain a target production schedule.

[0050] Specifically, after obtaining multiple second production scheduling plans, the control terminal can process each second production scheduling plan based on the intelligent matching model, for example, but not limited to, inputting the process personnel information configured for each standard process in each second production scheduling plan into the intelligent matching model to predict the personnel skill simulation index corresponding to each second production scheduling plan. Here, the personnel skill simulation index can at least include two indicators: the simulated skill efficiency value and the simulated off-duty probability of the process personnel configured for each standard process in each second production scheduling plan. The process personnel information can be understood as the identity information of each process personnel under the corresponding standard process, such as the name abbreviation or the corresponding work number; the intelligent matching model can be understood as a normal distribution probability prediction model, which is trained by the identity information of multiple process personnel under different standard processes, the historical single-piece working time, and the manually marked indicator results.

[0051] It is understandable that the intelligent matching model can also achieve accurate matching of processes with personnel or equipment through a variety of methods, but is not limited to it, such as establishing a rule base through process experts and on-site management experience, and defining the skills, experience, equipment and personnel related information required for the process, and matching through matching algorithms to ensure the matching degree between personnel and processes. Among them, in the matching process, matching can be carried out according to preset rules. If the personnel have the required skills, the match is successful. If they have similar skills, downgrade matching is performed; at the same time, through machine learning classification algorithms and historical reimbursement data, personnel are divided into different categories such as senior workers, intermediate workers, and junior workers, and the corresponding personnel categories are selected for matching according to the skill categories required for the process; of course, matching can also be based on cosine similarity or Jaccard similarity, and genetic algorithms can be used to achieve the optimal planning problem of process personnel matching, which is not limited to this.

[0052] In addition, the control terminal can also, but is not limited to, input the equipment information corresponding to each standard process in each second production schedule into a preset equipment probability distribution model to predict the probability of equipment simulated failure as an equipment simulation indicator, and can use the equipment simulation preparation together with the above-mentioned personnel skill simulation indicator as a production indicator. Here, the equipment information can be understood as all equipment types included in each standard process in the second production schedule and the size information of each equipment, and the preset equipment probability distribution model can be understood as a Weibull distribution probability prediction model for simulating the probability of equipment failure. The Weibull distribution probability prediction model can be trained by all equipment types included in multiple standard processes and historical fault detection data corresponding to each equipment, and is not limited to this.

[0053] Furthermore, after predicting the production indicators corresponding to each second production schedule, the control terminal can also, but is not limited to, input the production indicators corresponding to all second production schedules into a preset evaluation prediction model to generate a sorting result including all second production schedules, and can use one or more second production schedules that are at the top of the sorting result as the target production schedule, so that the on-site team leader can quickly select or confirm the final production schedule. Here, the preset evaluation prediction model can be understood as a conventional evaluator or a neural network prediction model, which can be trained by all production indicators corresponding to multiple historical production schedules and the sorting results after sorting by the on-site team leader, and is not limited to this.

[0054] In an embodiment of the present application, after obtaining the target production schedule, the control terminal may also, but is not limited to, determine the display format with the most records based on the historical display records of the user's PC terminal or tablet computer, and convert the target production schedule according to the display format to display the target production schedule corresponding to the display format on the user's PC terminal or tablet computer. Here, the display format may be, but is not limited to, a sand table display format, a simulated production animation display format, and an estimated line balance chart display format, wherein the sand table display format is a virtual mapping of the actual production line on the user's electronic device, restoring the actual production situation in proportion, such as, but not limited to, including all equipment length and width, equipment direction, equipment relative position, obstacles, process personnel names, seat orientation, or piece logistics direction and other information in the display content.

[0055] As another option of the embodiment of the present application, after the corresponding second production scheduling plans are sorted based on all production indicators to obtain the target production scheduling plan, the method further includes: The target production schedule is executed based on user feedback information, and the target production schedule is updated according to the production line operation data corresponding to the target production schedule.

[0056] Specifically, when receiving the modification information fed back by the user's PC terminal or tablet computer, the control terminal can modify the target production scheduling plan according to the modification information, and send the modified target production scheduling plan to the garment factory, so that the garment factory can perform actual production processing according to the modified target production scheduling plan. Here, when the information fed back by the user's PC terminal or tablet computer is confirmation information, the control terminal can directly send the final confirmed target production scheduling plan to the garment factory, so that the garment factory can perform actual production processing according to the final confirmed target production scheduling plan.

[0057] Next, when the actual production process is carried out according to the target production scheduling plan, the control terminal can also, but is not limited to, control the hanging system and the workstation monitoring system set up in the production line of the garment factory to collect data such as production line production speed, assembly line process bottleneck and equipment status in real time, and analyze and process the collected data to obtain the actual production efficiency of the personnel of each standard process, and the actual output can be calculated based on the actual production efficiency of the personnel of each standard process. It can be understood that when the actual output is detected to be lower than the target output, it indicates that the actual operation of the production scheduling plan has deviated, such as temporary employee leave or equipment damage, and then it can be, but is not limited to, collecting and recording the production scheduling data corresponding to the current production scheduling plan (for example, it can include the process personnel, personnel skill data, operating time, number of configured personnel and daily output corresponding to each standard process), that is, the production scheduling data corresponding to the current production scheduling plan is used as historical production scheduling data, and combined with one or more of the above-mentioned embodiments, production bottlenecks can be eliminated purposefully, and a better target production scheduling plan can be generated.

[0058] See also Figure 2 , Figure 2 A structural schematic diagram of an intelligent production scheduling system provided in an embodiment of the present application is shown.

[0059] like Figure 2 As shown, the intelligent production scheduling system may at least include a data acquisition module 201, a solution generation module 202, and a solution determination module 203, wherein: The data acquisition module 201 is used to generate at least two first production scheduling plans according to the historical production scheduling data and the intelligent process model; wherein the historical production scheduling data includes at least two historical production scheduling plans, each of which includes at least two standard processes arranged in a scheduling order, and each first production scheduling plan includes at least two standard processes arranged in a specified order; The plan generation module 202 is used to allocate and process all the first production scheduling plans based on the preset personnel skill data to obtain at least two second production scheduling plans; wherein each second production scheduling plan includes at least two standard processes arranged in a specified order and the configuration personnel corresponding to each standard process; The plan determination module 203 is used to process each second production scheduling plan based on the intelligent matching model to obtain corresponding production indicators, and to sort the corresponding second production scheduling plans based on all production indicators to obtain a target production scheduling plan.

[0060] In some possible embodiments, the historical production scheduling data further includes equipment parameters corresponding to each standard process; Generate at least two first production scheduling plans based on historical production scheduling data and intelligent process models, including: Based on a preset process coding table, all standard processes included in each historical production scheduling scheme in the historical production scheduling data are coded to obtain corresponding first coding values; Based on a preset parameter coding table, the equipment parameters corresponding to all standard processes included in each historical production scheduling plan are coded to obtain corresponding second coding values; The first coding value and the second coding value corresponding to each historical production scheduling plan are integrated to obtain at least two process coding values, and at least two first production scheduling plans are generated based on all process coding values ​​and the intelligent process model.

[0061] In some possible embodiments, the first coding value and the second coding value corresponding to each historical production scheduling plan are integrated to obtain at least two process coding values, including: Integrate the first code value and the second code value corresponding to each historical production scheduling plan to obtain a corresponding third code value; Performing feature extraction processing on any at least two third code values ​​to obtain at least two difference feature codes; A preset classification model is trained based on all difference feature codes, and all third code values ​​are input into the preset classification model to obtain at least two process code values.

[0062] In some possible embodiments, at least two first production scheduling plans are generated according to all process code values ​​and the intelligent process model, including: All process code values ​​are marked in production order, and the intelligent process model is trained based on all the marked process code values; Based on the preset process coding table, the target process table is coded to obtain the target coding value, and the target coding value is input into the trained intelligent process model to obtain the process matrix; The process matrix is ​​processed based on a deep search algorithm to obtain at least two first production scheduling plans.

[0063] In some possible embodiments, the preset personnel skill data includes process skill parameters corresponding to at least two standard processes for each process personnel; All first production scheduling plans are allocated and processed based on the preset personnel skill data to obtain at least two second production scheduling plans, including: Determine the similar processes corresponding to each standard process of each process personnel, and count the process skill parameters of all process personnel corresponding to the similar processes; Perform standard deviation calculation on the process skill parameters of all process personnel corresponding to similar processes to obtain the corresponding parameter standard deviation, and clean the process skill parameters of the corresponding standard process based on the parameter standard deviation; Based on the process skill parameters of each standard process of each process personnel after the cleaning process, all the first production scheduling plans are allocated and processed to obtain at least two second production scheduling plans.

[0064] In some possible embodiments, all first production scheduling plans are allocated based on the process skill parameters of each standard process of each process personnel after the cleaning process to obtain at least two second production scheduling plans, including: Based on the preset bandwidth parameters and the process skill parameters of each process personnel after each standard process has been cleaned, a Gaussian kernel density function is constructed, and the corresponding process skill efficiency is generated according to the Gaussian kernel density function; All first production scheduling plans are allocated and processed based on preset personnel skill data and all process skill efficiencies to obtain at least two second production scheduling plans.

[0065] In some possible embodiments, each second production scheduling plan is processed based on the intelligent matching model to obtain corresponding production indicators, including: Input the process personnel information configured for each standard process in each second production scheduling plan into the intelligent matching model to obtain the corresponding personnel skill simulation index; Generate corresponding production indicators based on the personnel skill simulation indicators of each second production scheduling plan.

[0066] In some possible embodiments, after the corresponding second production scheduling plans are sorted based on all production indicators to obtain the target production scheduling plan, the method further includes: The target production schedule is executed based on user feedback information, and the target production schedule is updated according to the production line operation data corresponding to the target production schedule.

[0067] See also Figure 3 , Figure 3 A structural schematic diagram of another intelligent production scheduling system provided in an embodiment of the present application is shown.

[0068] like Figure 3 As shown, the intelligent production scheduling system 300 may include at least one processor 301 , at least one network interface 304 , a user interface 303 , a memory 305 , and at least one communication bus 302 .

[0069] The communication bus 302 may be used to realize the connection and communication of the above-mentioned components; the user interface 303 may include buttons, and the optional user interface may also include a standard wired interface or a wireless interface; the network interface 304 may include, but is not limited to, a Bluetooth module, an NFC module, a Wi-Fi module, etc.; the processor 301 may include one or more processing cores; the memory 305 may include a RAM or a ROM to store instructions, programs, codes, code sets or instruction sets. Figure 3 As shown, the memory 305 as a computer storage medium may include an operating system, a network communication module, a user interface module, and an intelligent production scheduling application.

[0070] Specifically, the processor 301 may be used to call the intelligent production scheduling application stored in the memory 305, and specifically perform the following operations: Generate at least two first production scheduling plans according to the historical production scheduling data and the intelligent process model; wherein the historical production scheduling data includes at least two historical production scheduling plans, each historical production scheduling plan includes at least two standard processes arranged in a scheduling order, and each first production scheduling plan includes at least two standard processes arranged in a specified order; Allocate and process all first production scheduling plans based on preset personnel skill data to obtain at least two second production scheduling plans; wherein each second production scheduling plan includes at least two standard processes arranged in a specified order and a configuration personnel corresponding to each standard process; Each second production schedule is processed based on the intelligent matching model to obtain corresponding production indicators, and the corresponding second production schedules are sorted based on all production indicators to obtain a target production schedule.

[0071] In some possible embodiments, the historical production scheduling data further includes equipment parameters corresponding to each standard process; Generate at least two first production scheduling plans based on historical production scheduling data and intelligent process models, including: Based on a preset process coding table, all standard processes included in each historical production scheduling scheme in the historical production scheduling data are coded to obtain corresponding first coding values; Based on a preset parameter coding table, the equipment parameters corresponding to all standard processes included in each historical production scheduling plan are coded to obtain corresponding second coding values; The first coding value and the second coding value corresponding to each historical production scheduling plan are integrated to obtain at least two process coding values, and at least two first production scheduling plans are generated based on all process coding values ​​and the intelligent process model.

[0072] In some possible embodiments, the first coding value and the second coding value corresponding to each historical production scheduling plan are integrated to obtain at least two process coding values, including: Integrate the first code value and the second code value corresponding to each historical production scheduling plan to obtain a corresponding third code value; Performing feature extraction processing on any at least two third code values ​​to obtain at least two difference feature codes; A preset classification model is trained based on all difference feature codes, and all third code values ​​are input into the preset classification model to obtain at least two process code values.

[0073] In some possible embodiments, at least two first production scheduling plans are generated according to all process code values ​​and the intelligent process model, including: All process code values ​​are marked in production order, and the intelligent process model is trained based on all the marked process code values; Based on the preset process coding table, the target process table is coded to obtain the target coding value, and the target coding value is input into the trained intelligent process model to obtain the process matrix; The process matrix is ​​processed based on a deep search algorithm to obtain at least two first production scheduling plans.

[0074] In some possible embodiments, the preset personnel skill data includes process skill parameters corresponding to at least two standard processes for each process personnel; All first production scheduling plans are allocated and processed based on the preset personnel skill data to obtain at least two second production scheduling plans, including: Determine the similar processes corresponding to each standard process of each process personnel, and count the process skill parameters of all process personnel corresponding to the similar processes; Perform standard deviation calculation on the process skill parameters of all process personnel corresponding to similar processes to obtain the corresponding parameter standard deviation, and clean the process skill parameters of the corresponding standard process based on the parameter standard deviation; Based on the process skill parameters of each standard process of each process personnel after the cleaning process, all the first production scheduling plans are allocated and processed to obtain at least two second production scheduling plans.

[0075] In some possible embodiments, all first production scheduling plans are allocated based on the process skill parameters of each standard process of each process personnel after the cleaning process to obtain at least two second production scheduling plans, including: Based on the preset bandwidth parameters and the process skill parameters of each process personnel after each standard process has been cleaned, a Gaussian kernel density function is constructed, and the corresponding process skill efficiency is generated according to the Gaussian kernel density function; All first production scheduling plans are allocated and processed based on preset personnel skill data and all process skill efficiencies to obtain at least two second production scheduling plans.

[0076] In some possible embodiments, each second production scheduling plan is processed based on the intelligent matching model to obtain corresponding production indicators, including: Input the process personnel information configured for each standard process in each second production scheduling plan into the intelligent matching model to obtain the corresponding personnel skill simulation index; Generate corresponding production indicators based on the personnel skill simulation indicators of each second production scheduling plan.

[0077] In some possible embodiments, after the corresponding second production scheduling plans are sorted based on all production indicators to obtain the target production scheduling plan, the method further includes: The target production schedule is executed based on user feedback information, and the target production schedule is updated according to the production line operation data corresponding to the target production schedule.

Claims

1. An intelligent production scheduling method, characterized in that: include: Generate at least two first production scheduling plans according to the historical production scheduling data and the intelligent process model; wherein the historical production scheduling data includes at least two historical production scheduling plans, each of the historical production scheduling plans includes at least two standard processes arranged in a scheduling order, and each of the first production scheduling plans includes at least two standard processes arranged in a specified order; Allocate and process all the first production scheduling plans based on preset personnel skill data to obtain at least two second production scheduling plans; wherein each of the second production scheduling plans includes at least two standard processes arranged in a specified order and a configuration personnel corresponding to each of the standard processes; Each of the second production scheduling plans is processed based on the intelligent matching model to obtain corresponding production indicators, and the corresponding second production scheduling plans are sorted based on all the production indicators to obtain a target production scheduling plan.

2. The method according to claim 1, characterized in that The historical production scheduling data also includes equipment parameters corresponding to each of the standard processes; The generating at least two first production scheduling plans according to the historical production scheduling data and the intelligent process model includes: Based on a preset process coding table, all standard processes included in each historical production scheduling scheme in the historical production scheduling data are coded to obtain corresponding first coding values; Based on a preset parameter coding table, the equipment parameters corresponding to all standard processes included in each of the historical production scheduling plans are coded to obtain corresponding second coding values; The first coding value and the second coding value corresponding to each of the historical production scheduling plans are integrated to obtain at least two process coding values, and at least two first production scheduling plans are generated based on all the process coding values ​​and the intelligent process model.

3. The method according to claim 2, characterized in that The integrating the first coding value and the second coding value corresponding to each of the historical production scheduling plans to obtain at least two process coding values ​​includes: Integrate the first coding value and the second coding value corresponding to each of the historical production scheduling plans to obtain a corresponding third coding value; Performing feature extraction processing on any at least two of the third code values ​​to obtain at least two difference feature codes; A preset classification model is trained based on all the difference feature codes, and all the third code values ​​are input into the preset classification model to obtain at least two process code values.

4. The method according to claim 2, characterized in that: At least two first production scheduling plans are generated according to all the process code values ​​and the intelligent process model, including: Performing production sequence marking processing on all the process code values, and training the intelligent process model based on all the process code values ​​after the marking processing; Based on the preset process code table, the target process table is encoded to obtain a target code value, and the target code value is input into the trained intelligent process model to obtain a process matrix; The process matrix is ​​processed based on a deep search algorithm to obtain at least two first production scheduling plans.

5. The method according to claim 1, characterized in that The preset personnel skill data includes process skill parameters corresponding to each process personnel and at least two of the standard processes; The allocating and processing all the first production scheduling plans based on the preset personnel skill data to obtain at least two second production scheduling plans includes: Determine the similar process corresponding to each standard process of each process personnel, and count the process skill parameters of all process personnel corresponding to the similar processes; Performing standard deviation calculation processing on the process skill parameters of all the process personnel corresponding to the similar processes to obtain corresponding parameter standard deviations, and performing cleaning processing on the process skill parameters of the corresponding standard processes based on the parameter standard deviations; Based on the process skill parameters of each standard process of each process personnel after the cleaning process, all the first production scheduling plans are allocated and processed to obtain at least two second production scheduling plans.

6. The method according to claim 5, characterized in that The step of allocating all the first production scheduling plans based on the process skill parameters of each standard process of each process personnel after the cleaning process to obtain at least two second production scheduling plans includes: Based on the preset bandwidth parameters and the process skill parameters of each of the process personnel after each of the standard processes has been cleaned, a Gaussian kernel density function is constructed, and corresponding process skill efficiency is generated according to the Gaussian kernel density function; All of the first production scheduling plans are allocated and processed based on preset personnel skill data and all of the process skill efficiencies to obtain at least two second production scheduling plans.

7. The method according to claim 1, characterized in that The intelligent matching model is used to process each of the second production scheduling plans to obtain corresponding production indicators, including: Inputting the process personnel information configured for each standard process in each of the second production scheduling plans into the intelligent matching model to obtain corresponding personnel skill simulation indicators; Generate corresponding production indicators based on the personnel skill simulation indicators of each of the second production scheduling plans.

8. The method according to claim 1, characterized in that After the corresponding second production scheduling plans are sorted based on all the production indicators to obtain a target production scheduling plan, the method further includes: The target production scheduling plan is executed based on user feedback information, and the target production scheduling plan is updated according to the production line operation data corresponding to the target production scheduling plan.

9. An intelligent production scheduling system, characterized in that: include: A data acquisition module, used to generate at least two first production scheduling plans according to historical production scheduling data and an intelligent process model; wherein the historical production scheduling data includes at least two historical production scheduling plans, each of the historical production scheduling plans includes at least two standard processes arranged in a scheduling order, and each of the first production scheduling plans includes at least two standard processes arranged in a specified order; A plan generation module, used for allocating and processing all the first production scheduling plans based on preset personnel skill data to obtain at least two second production scheduling plans; wherein each of the second production scheduling plans includes at least two standard processes arranged in a specified order and a configuration personnel corresponding to each of the standard processes; The plan determination module is used to process each of the second production scheduling plans based on the intelligent matching model to obtain corresponding production indicators, and to sort the corresponding second production scheduling plans based on all the production indicators to obtain a target production scheduling plan.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: The computer-readable storage medium stores instructions, and when the instructions are executed on a computer or a processor, the computer or the processor executes the steps of the method according to any one of claims 1 to 8.

Images