Intelligent Scheduling System for Customer Orders Based on Multi-Source Data Analysis

Abstract

The present application provides a customer order intelligent scheduling system based on multi-source data analysis. The product production line is used to produce multiple products and includes a central control device and M production devices. The customer order intelligent scheduling system is installed in the central control device and includes: a data acquisition module for acquiring multi-source data for order scheduling, where the multi-source data for order scheduling includes N scheduling order data; a data processing module for preprocessing and feature extraction of the multi-source data for order scheduling to generate input data for optimizing scheduling; an intelligent scheduling module for constructing an intelligent scheduling model for optimization and solution, determining an order scheduling queue, and controlling the product production line to produce products based on the order scheduling queue. By introducing the reusability of production equipment, parallel scheduling of multiple order queues is realized based on multi-source data, so as to improve the working efficiency of production equipment while ensuring the completion of order task allocation, and achieve energy conservation and cost reduction while optimizing scheduling.

Description

Technical Field

[0001] This application relates to the field of intelligent order scheduling, and more particularly, to an intelligent customer order scheduling system based on multi-source data analysis. Background Art

[0002] With the deepening of globalization and digitalization, enterprises are facing increasingly fierce market competition. How to efficiently and accurately meet customer order requirements has become an important manifestation of enterprise competitiveness. Order scheduling, as the core link of production management, its optimization is directly related to production efficiency, resource utilization rate, and customer satisfaction. However, the complexity, diversity, and dynamics of the modern production environment pose severe challenges to traditional scheduling methods. In today's manufacturing environment, customer orders usually involve multiple product types and complex workloads. This complexity is mainly reflected in the diversity of order requirements, the heterogeneity of production equipment, etc. For example, the order requirements of different customers include significant differences in product types, production quantities, and delivery cycles; the equipment in the production system usually has different performances, making the equipment load allocation more complex.

[0003] Traditional scheduling methods such as the first-in-first-out rule or heuristic methods based on fixed allocation perform well in single-task types or simple production environments, but show obvious limitations in complex and dynamic production environments. Facing the requirements of modern production systems, there is an urgent need for new methods that can adapt to system state changes in real time and optimize the scheduling plan.

[0004] Applying multi-source data analysis to order scheduling is an important direction in scheduling optimization research. Through data-driven methods, the real-time state of the production system can be dynamically obtained and model parameters can be updated to enable it to reflect the dynamic changes of the system. This combined method has achieved initial results in some studies, but its application is still limited to specific scenarios, such as the optimization of a single factory or a single-stage production system. For more complex production systems, further development and optimization are still required. Summary of the Invention

[0005] The purpose of the embodiments of this application is to provide an intelligent customer order scheduling system based on multi-source data analysis. By introducing the reusability of production equipment, parallel scheduling of multiple order queues is achieved based on multi-source data, so as to improve the working efficiency of production equipment while ensuring the completion of order task allocation, and to achieve energy conservation and cost reduction while optimizing scheduling.

[0006] To achieve the above purpose, the embodiments of this application are implemented as follows:

[0007] In a first aspect, an embodiment of the present application provides an intelligent production scheduling system for customer orders based on multi-source data analysis. The product production line is used to produce multiple products and includes a central control device and M production devices. Each production device is a dedicated device or a shared device. A dedicated device only participates in a single production process of a product, and a shared device can participate in multiple production processes of multiple products. The intelligent production scheduling system for customer orders is installed in the central control device and includes: a data acquisition module, configured to acquire multi-source data for order production scheduling. The multi-source data for order production scheduling includes N production scheduling order data. Each production scheduling order data includes a product type, a product quantity, an order deadline, and the production equipment and production process required for product production. Each production scheduling order data is used for the production of a single product; a data processing module, configured to preprocess and extract features from the multi-source data for order production scheduling to generate input data for optimizing production scheduling; an intelligent production scheduling module, configured to build an intelligent production scheduling model based on the input data for optimization and solution, determine an order production scheduling queue, and control the product production line to produce products based on the order production scheduling queue.

[0008] In combination with the first aspect, in the first possible implementation manner of the first aspect, the data processing module is specifically configured to: preprocess the multi-source data for order production scheduling; extract features from each preprocessed production scheduling order data, and extract the order feature and production feature of each production scheduling order data. The order feature reflects the product production requirements for completing the production scheduling order data, and the production feature reflects the production equipment, product processes, process sequence, and production time required for completing the production scheduling order data; integrate the order feature and production feature of each production scheduling order data to construct an order feature matrix as the input data for optimizing production scheduling.

[0009] In combination with the first possible implementation manner of the first aspect, in the second possible implementation manner of the first aspect, the production process includes the product processes, process sequence, and production time of the production equipment corresponding to the product type. The data processing module is specifically configured to: for each preprocessed production scheduling order data: extract the order number, product type, product quantity, and order deadline in this production scheduling order data to generate a one-dimensional vector with a length of S as the order feature; perform one-hot encoding on the production equipment corresponding to the product type in this production scheduling order data, and expand the attributes corresponding to each production equipment in the one-hot encoding feature vector. The expanded attributes include product processes, process sequence, and production time, and complete the expanded one-hot encoding feature vector based on the product processes, process sequence, and production time of the production equipment to obtain a one-dimensional vector with a length of 4M as the production feature; splice the order feature and the production feature to obtain the integrated feature of each production scheduling order data; combine the integrated features of the N production scheduling order data into an order feature matrix of (4M + S) × N.

[0010] Combined with the first possible implementation manner of the first aspect, in the third possible implementation manner of the first aspect, the intelligent production scheduling module is specifically configured to: obtain the set production efficiency matrix of M production devices within a unit time period; define the production task allocation function of each production device based on the order feature matrix and the set production efficiency matrix of each production device; construct the objective function to be optimized based on the production task allocation function; generate constraint conditions based on the order feature matrix and the set production efficiency matrix of each production device; optimize and solve the objective function based on the constraint conditions to obtain the order production scheduling queue.

[0011] Combined with the third possible implementation manner of the first aspect, in the fourth possible implementation manner of the first aspect, the set production efficiency matrix of M production devices within a unit time period is:

[0012]

[0013] where, w 11 represents the set production efficiency of the first production device for the first product process in the set of all product processes within a unit time period, M is the total number of production devices, and V is the total number of processes of all product types supported in the product production line.

[0014] Combined with the fourth possible implementation manner of the first aspect, in the fifth possible implementation manner of the first aspect, it is specifically configured to: the set production efficiency of the kth production device is w k =[w k1 , w k2 , …, w ki , …, w kV , w ki is the set production efficiency of the kth production device for the ith product process in the set of all product processes within a unit time period, and define the production task allocation function δ k of the kth production device as:

[0015] δ k =[δ k1 , δ k2 , …, δ ki , …, δ kV ,

[0016] where, δ ki is the proportion of the ith product process in the set of all product processes allocated to the kth production device within a unit time period in the set production efficiency of the kth production device.

[0017] Combined with the fifth possible implementation manner of the first aspect, in the sixth possible implementation manner of the first aspect, the intelligent production scheduling module is specifically configured to: based on the order feature matrix, for the j-th production scheduling order data: determine the order start node ts of the j-th production scheduling order data based on the start time node of the unit period corresponding to the product process with the smallest process sequence number in which the order number of the j-th production scheduling order data first appears j , and, based on the end time node of the unit period corresponding to the product process with the largest process sequence number when the quantity of products with the order number of the j-th production scheduling order data is completed, determine the order end node te of the j-th production scheduling order data j ; construct the objective function to be optimized as:

[0018] minimize F = A + B + C,

[0019]

[0020] where F is the objective function to be optimized, A, B, and C are the three components of the function F respectively, M is the number of production equipment, δ ki is the ratio of the i-th product process in all product process sets to the set production efficiency of the k-th production equipment within a unit period, V is the total number of processes of all product types supported in the product production line, is the production efficiency of the k-th production equipment in the l-th effective unit period, T k is the total number of effective unit periods of the k-th production equipment. The effective unit period means the unit period from the start of the unit period when the production equipment participates in the production of the products of this N production scheduling order data to the end of the unit period when the production equipment finishes the production of the products of this N production scheduling order data, α k is the efficiency weight value of the k-th production equipment, N is the total number of production scheduling order data, ts j is the order start node of the j-th production scheduling order data, te j is the order end node of the j-th production scheduling order data, β is the adjustment factor, T is the total duration for completing the production of the products of N production scheduling order data in the case of single-order production, t j is the order deadline of the j-th production scheduling order data, tz j is the order entry time of the j-th production scheduling order data, γ j is the timeliness weight value of the j-th production scheduling order data.

[0021] Combined with the sixth possible implementation manner of the first aspect, in the seventh possible implementation manner of the first aspect, the intelligent production scheduling module is specifically configured to: generate the following constraint conditions based on the order feature matrix and the set production efficiency matrix of each production equipment:

[0022] Device efficiency constraint conditions:

[0023] Production time constraint conditions:

[0024] Combined with the seventh possible implementation manner of the first aspect, in the eighth possible implementation manner of the first aspect, for the product processes corresponding to the same order number, the product process i of the subsequent process order is produced by the production equipment x = {x1,..., x p}, p ∈ [1, k] in the first l effective unit time periods, and the product process r of the previous process order is produced by the production equipment y = {y1,..., y q}, q ∈ [1, k] in the previous (l - 1) effective unit time periods. The intelligent production scheduling module is further configured to: generate product process constraint conditions:

[0025]

[0026] Wherein, is the proportion of the set production efficiency of the production equipment x b occupied by the product process i within the unit time period, b when allocated to the production equipment x is the set production efficiency of the production equipment x b for the product process i within the unit time period, is the production efficiency of the product process i in the lth effective unit time period, is the proportion of the set production efficiency of the production equipment y c occupied by the product process r within the unit time period, c when allocated to the production equipment y is the set production efficiency of the production equipment y c for the product process r within the unit time period, is the total production efficiency of the product process r in the previous (l - 1) effective unit time periods, is the total production efficiency of the product process i in the previous (l - 1) effective unit time periods.

[0027] Combined with the third possible implementation manner of the first aspect, in the ninth possible implementation manner of the first aspect, a genetic algorithm or a particle swarm algorithm is used to optimize and solve the objective function.

[0028] Beneficial effects:

[0029] 1. In this solution, the product production line is used to produce multiple products and includes a central control device and M production devices. Each production device is a dedicated device or a shared device. The dedicated device only participates in the single production process of the product, and the shared device can participate in multiple production processes of multiple products. The intelligent production scheduling system for customer orders is installed in the central control device and includes: a data acquisition module for acquiring multi-source data for production scheduling. The multi-source data for production scheduling includes N production scheduling order data. Each production scheduling order data includes product type, product quantity, order deadline, and the production equipment and production process required for product production. Each production scheduling order data is used for the production of a single product (by pre-processing to split the customer's multi-product order into single-product orders to distinguish different product types, quantities, deadlines, etc.); a data processing module for pre-processing and feature extraction of the multi-source data for production scheduling to generate input data for optimizing production scheduling; an intelligent production scheduling module for constructing an intelligent production scheduling model based on the input data for optimization and solution, determining the production scheduling queue for orders, and controlling the product production line to produce products based on the production scheduling queue for orders. By introducing the reusability of production equipment and realizing parallel production scheduling of multiple order queues based on multi-source data, introducing multi-source data analysis can obtain the status of the production system in real time and dynamically adjust the production scheduling plan based on these data, so as to better adapt to the complexity, diversity, and dynamics of the modern production environment. By constructing an intelligent production scheduling model, it is possible to comprehensively consider various factors such as order requirements, production equipment performance, and production process to optimize the production task allocation. This not only helps to ensure the timely completion of orders, but also improves the working efficiency of production equipment, reduces resource waste, and achieves energy conservation and cost reduction. The system can process customer orders containing multiple product types and complex workloads. Through parallel production scheduling, the overall order processing cycle is effectively shortened, and customer satisfaction is improved. The system not only supports dedicated devices, but also makes full use of shared devices to participate in multiple production processes of multiple products, thus improving the flexibility and response speed of the production line.

[0030] 2. When preprocessing and extracting features from multi-source data for order scheduling, the system not only extracts basic information about the order (such as order number, product type, product quantity, order deadline), but also goes deep into the production level and extracts production features related to production equipment, product process, process sequence, and production time. This comprehensive feature extraction method helps the system to understand order requirements and production conditions more accurately, providing a solid foundation for subsequent optimization and scheduling. Processing the input data into an order feature matrix can ensure the consistency of the input data (N data can be changed, (3M+S) remains constant), and can represent a large amount of order data in an efficient and compact way, which is not only convenient for subsequent calculations and analysis, but also helps to improve the operating efficiency and response speed of the system. Based on this matrix, the system can use advanced optimization algorithms (such as genetic algorithms, particle swarm algorithms, etc.) to find the optimal scheduling plan. This data-driven optimization method can fully consider various constraints (such as equipment efficiency, production time, etc.) to ensure the effectiveness and feasibility of the scheduling plan.

[0031] 3. By setting the production efficiency matrix for M production equipment in a unit time period, the system can accurately reflect the production efficiency of each production equipment in different product processes. This helps the system to fully consider the actual capacity of the production equipment when allocating production tasks and ensure the rationality of the allocation. Setting the production efficiency matrix also makes it easier for the system to monitor and optimize the performance of production equipment, thereby improving overall production efficiency. Based on the order feature matrix and the set production efficiency matrix of each production equipment, the production task allocation function is defined (including three parts of the optimization function. By minimizing the three components A, B, and C, the system can comprehensively consider multiple factors such as production efficiency, production time, and order deadline, so as to find the optimal production scheduling plan. Part A in the objective function takes into account the efficiency balance of the production equipment, which helps to avoid equipment overload or idleness; Part B considers the minimization of production time, which helps to reduce production costs; Part C considers the timely delivery of orders, which helps to improve customer satisfaction. In this way, factors such as the production efficiency of the production equipment, the task processing time of each order scheduling data, and the completion cycle of the entire order scheduling task can be comprehensively considered. Since the supply can be supplemented according to the situation and is highly affected by external influences and human intervention, it is directly assumed that the supply is sufficient and not considered. On the production line of the fully automatic process, the raw material supply can also be taken into consideration to jointly construct the production task allocation function), so that the system can flexibly adjust the production task allocation according to the actual situation, which helps to respond quickly to customer needs. The production task allocation function can also ensure the balanced distribution of production tasks among multiple production equipment, avoid overloading or idleness of some equipment, thereby improving resource utilization, and can take into account the continuity between the various processes of the task as much as possible, shorten the production cycle of each product, and avoid the situation where a large number of process parts are accumulated.

[0032] 4. The equipment efficiency constraint ensures that the production task allocation does not exceed the actual capacity of the production equipment, avoiding the risks of equipment overload and damage. The production time constraint ensures that the start and end times of orders conform to the actual production process, avoiding production delays and order backlogs. The product process constraint ensures the sequentiality and continuity of the product processes with the same order number during the production process, helping to ensure product quality and production efficiency. By optimizing and solving the objective function based on the constraints, the system can quickly find the optimal production scheduling plan. This helps enterprises improve production efficiency and reduce production costs. The optimization and solution process can also consider multiple constraints and objective functions, thus generating a production scheduling plan that better meets the actual needs. Through the order production scheduling queue obtained by the optimization and solution, the system can guide the specific task allocation and time arrangement during the production process. This helps enterprises achieve visualization and controllability of the production process and improve the refinement level of production management. The production scheduling queue can also serve as an important basis for the production plan, helping enterprises reasonably arrange production resources and inventory, and reducing production risks and costs.

[0033] To make the above objects, features, and advantages of the present application more obvious and understandable, the following provides preferred embodiments in conjunction with the accompanying drawings and describes them in detail as follows. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] To more clearly illustrate the technical solutions of the embodiments of the present application, the following briefly introduces the accompanying drawings required for the embodiments of the present application. It should be understood that the following drawings only show some embodiments of the present application and should not be regarded as limiting the scope. For those of ordinary skill in the art, without creative efforts, other relevant drawings can also be obtained based on these drawings.

[0035] Figure 1 It is a schematic diagram of the intelligent production scheduling system for customer orders based on multi-source data analysis provided by the embodiments of the present application.

[0036] Reference numerals: 10 - intelligent production scheduling system for customer orders; 11 - data acquisition module; 12 - data processing module; 13 - intelligent production scheduling module; 20 - production equipment; 21 - dedicated equipment; 22 - shared equipment; 30 - central control device. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0037] The following will describe the technical solutions in the embodiments of the present application in conjunction with the accompanying drawings in the embodiments of the present application.

[0038] Please refer to Figure 1 , Figure 1This is a schematic diagram of the intelligent production scheduling system for customer orders based on multi-source data provided by the embodiments of this application. In this embodiment, the product production line is used to produce multiple products, including a central control device and M production devices. Each production device is a dedicated device or a shared device. A dedicated device only participates in the single production process of a product (i.e., only processes one product process), and a shared device can participate in multiple production processes of multiple products (i.e., can process multiple product processes, generally processes the product processes of multiple different products, rather than different product processes of the same product). The intelligent production scheduling system for customer orders is installed in the central control device.

[0039] The intelligent production scheduling system for customer orders includes a data acquisition module, a data processing module, and an intelligent production scheduling module.

[0040] The data acquisition module is used to acquire multi-source data for order production scheduling. The multi-source data for order production scheduling includes N production order data. Each production order data includes product type, product quantity, order deadline, and the production equipment and production process required for product production. Each production order data is used for the production of a single product.

[0041] In this embodiment, a customer order can be a single product order or a complex customer order that includes multiple products and corresponding delivery times. To handle complex customer orders, such customer orders need to be pre-processed and decomposed into single product orders. For example, a customer order requires 20,000 pieces of product A, with a delivery deadline before December 31, 2024, 50,000 pieces of product B, with a delivery deadline before December 31, 2024, and 100,000 pieces of product C, with a delivery deadline before March 31, 2025. It can be split into 3 single product orders. And each single product order can form a production order data (including product type, product quantity, order deadline, and the production equipment and production process required for product production), which is entered into the system. The system can differentially acquire order production according to the deadline. For example, the production schedule is divided by week (or day) for production scheduling, and the multi-source data for order production scheduling that needs to be completed in a week is acquired for task planning.

[0042] Accordingly, the data acquisition module can acquire multi-source data for order production scheduling, and the multi-source data for order production scheduling includes N production order data.

[0043] After that, the data processing module can pre-process and extract features from the multi-source data for order production scheduling to generate input data for optimizing production scheduling.

[0044] Exemplarily, the data processing module can pre-process the multi-source data for order production scheduling, such as removing duplicates and handling outliers, to ensure that each production order data in the multi-source data for order production scheduling is accurate and reliable.

[0045] After preprocessing, the data processing module can extract features from each scheduled production order data after preprocessing, extracting the order features and production features of each scheduled production order data. The order features reflect the product production requirements for completing the scheduled production order data, and the production features reflect the production equipment, product processes, process sequence, and production time required to complete the scheduled production order data.

[0046] Specifically, for each scheduled production order data after preprocessing:

[0047] The data processing module can extract the order number, product type, product quantity, and order deadline in this scheduled production order data to generate a one-dimensional vector of length S as the order features. For example, the order number occupies 1 dimension, the product type occupies 1 dimension, the product quantity occupies 1 dimension, the order deadline occupies 1 dimension, and the order entry time occupies 1 dimension. Then, the order features are a one-dimensional vector of length 5.

[0048] The data processing module can perform one-hot encoding on the production equipment corresponding to the product type in this scheduled production order data (for example, there are 46 production equipment in the entire production line that can participate in product production), and expand the attributes corresponding to each production equipment in the one-hot encoded feature vector (each original dimension is expanded to 4 dimensions). The expanded attributes include product processes (the processing required to complete the production of this product), process sequence (used to indicate the sequence of this product process in the production of products of this product type), and production time (the time required to complete a single product in this product process), and complete the expanded one-hot encoded feature vector based on the product processes, process sequence, and production time of the production equipment to obtain a one-dimensional vector of length 4M as the production features.

[0049] After that, the data processing module can concatenate the order features and production features to obtain the integrated features of each scheduled production order data (a one-dimensional vector of length (4M + S)), and then combine the integrated features of N scheduled production order data into an order feature matrix of (4M + S) × N. Thus, the input data for optimizing the production schedule can be obtained.

[0050] After obtaining the input data, the intelligent production scheduling module can build an intelligent production scheduling model based on the input data for optimization and solution, determine the order production scheduling queue, and control the product production line for product production based on the order production scheduling queue.

[0051] Exemplarily, the intelligent production scheduling module can obtain the set production efficiency matrix of M production devices within a unit time period. The set production efficiency matrix of M production devices within a unit time period (pre-set according to the efficiency of each production device in processing product processes. For example, the efficiency of a certain production device in a certain product process is to complete the processing of 15 products within each unit time period) is as follows:

[0052]

[0053] Among them, w 11 represents the set production efficiency of the first production device for the first product process in the set of all product processes within a unit time period. M is the total number of production devices, and V is the total number of processes of all product types supported in the product production line (for example, 278 product processes).

[0054] After obtaining the set production efficiency matrix of M production devices within a unit time period, the intelligent production scheduling module can define the production task allocation function of each production device based on the order feature matrix and the set production efficiency matrix of each production device.

[0055] Exemplarily, the intelligent production scheduling module can determine the set production efficiency parameters of each production device from the set production efficiency matrix. For example, the set production efficiency of the kth production device is:

[0056] w k =[w k1 , w k2 , …, w ki , …, w kV , (2)

[0057] Among them, w ki is the set production efficiency of the kth production device for the ith product process in the set of all product processes within a unit time period.

[0058] Accordingly, the intelligent production scheduling module can define the production task allocation function δ k of the kth production device as:

[0059] δ k =[δ k1 , δ k2 , …, δ ki , …, δ kV , (3)

[0060] Among them, δ ki is the proportion of the ith product process in the set of all product processes allocated to the kth production device within a unit time period to the set production efficiency of the kth production device.

[0061] After determining the production task allocation function, the intelligent production scheduling module can construct an objective function to be optimized based on the production task allocation function.

[0062] Exemplarily, based on the order feature matrix, for the j-th production scheduling order data:

[0063] The intelligent production scheduling module can determine the order start node ts of the j-th production scheduling order data based on the start time node of the unit time period corresponding to the product process with the smallest process sequence at the first occurrence of the order number of the j-th production scheduling order data. j And, based on the end time node of the unit time period corresponding to the product process with the largest process sequence when the number of products with the order number of the j-th production scheduling order data is completed, determine the order end node te of the j-th production scheduling order data. j .

[0064] Then, the objective function to be optimized is constructed as:

[0065] minimize F = A + B + C, (4)

[0066]

[0067] where minimize means minimizing the function, F is the objective function to be optimized, A, B, and C are the three components of the function F respectively, M is the number of production equipment, δ ki is the ratio of the i-th product process in all product process sets assigned to the k-th production equipment in a unit time period to the set production efficiency of the k-th production equipment, V is the total number of processes of all product types supported in the product production line, is the production efficiency of the k-th production equipment in the l-th effective unit time period, T k is the total number of effective unit time periods of the k-th production equipment. The effective unit time period means the unit time period from the start of the unit time period when the production equipment participates in the production of the products in this N production scheduling order data to the end of the unit time period when the production equipment finishes the production of the products in this N production scheduling order data. α k is the efficiency weight value of the k-th production equipment (used to measure the weight of the efficiency of different production equipment. The greater the power consumption and the more obvious the energy-saving effect of improving the efficiency on the equipment, the greater this weight), N is the total number of production scheduling order data, ts j is the order start node of the j-th production scheduling order data, te j is the order end node of the j-th production scheduling order data, β is an adjustment factor (used to set according to the production cycle of the production of various product types on the product production line to make part B occupy a reasonable weight. In this embodiment, β takes values between 0.1 and 0.7), T is the total duration of the production of the products in N production scheduling order data in the case of single-order production, tj is the order deadline of the j-th production scheduling order data, tz j is the order entry time of the j-th production scheduling order data, γ j is the timeliness weight value of the j-th production scheduling order data (which can measure the importance of the order. For example, if an order is urgent and important, an urgent order can be set so that the γ corresponding to the j-th production scheduling order data j has a relatively large value, for example, 10 times the weight of a regular order).

[0068] After constructing the objective function to be optimized, the intelligent production scheduling module can generate constraint conditions based on the order feature matrix and the set production efficiency matrix of each production equipment.

[0069] First, the intelligent production scheduling module can generate the following constraint conditions based on the order feature matrix and the set production efficiency matrix of each production equipment:

[0070] Equipment efficiency constraint condition:

[0071]

[0072] Production time constraint condition:

[0073]

[0074] The equipment efficiency constraint condition is used to constrain the tasks assigned to the production equipment not to exceed its set production efficiency in each unit time period, while the production time constraint condition is used to constrain the start time and end time of the product process to avoid task overrun.

[0075] In addition, complex product process constraints need to be considered. For the product processes corresponding to the same order number, the product process i of the subsequent process sequence is produced by the production equipment x = {x1,..., x p}, p ∈ [1, k] in the first l effective unit time periods, and the product process r of the previous process sequence is produced by the production equipment y = {y1,..., y q}, q ∈ [1, k] in the first (l - 1) effective unit time periods. Then, the intelligent production scheduling module can generate the product process constraint condition:

[0076]

[0077] Among them, is the proportion of the set production efficiency of the production equipment x occupied by the product process i in the unit time period b when the product process i is assigned to the production equipment x b of the set production efficiency of the production equipment x is the set production efficiency of the production equipment x b for the product process i in the unit time period, is the production efficiency of product process i in the l-th effective unit time period, is the proportion of the production equipment y allocated to product process r within a unit time period c when it accounts for the set production efficiency of production equipment y c ; is the set production efficiency of production equipment y c for product process r within a unit time period, is the total production efficiency of product process r in the previous (l - 1) effective unit time periods, is the total production efficiency of product process i in the previous (l - 1) effective unit time periods.

[0078] The product process constraint conditions are used to constrain the process sequence of product processes to avoid the situation of cross-process sequences.

[0079] Accordingly, the intelligent production scheduling module can optimize and solve the objective function based on the constraint conditions to obtain the order production scheduling queue. In this embodiment, this model is suitable for using a genetic algorithm (or a variant of the genetic algorithm) or a particle swarm algorithm for optimization and solution, and outputs the product processes and task allocations of each production equipment in each unit time period, so as to obtain the order production scheduling queue.

[0080] Since the process of the optimization algorithm is not the focus of this solution, only the general process is introduced here. Taking the genetic algorithm as an example, when optimizing and solving the objective function in the intelligent production scheduling system for customer orders based on multi-source data analysis, the specific process can be carried out according to the following steps:

[0081] (1) Initialize the population

[0082] Population size setting: First, determine the size of the population in the genetic algorithm, that is, the number of initial solutions.

[0083] Coding method: Select a suitable coding method to represent each solution (individual). In this embodiment, a real number coding method can be used to represent the production task allocation and product process information of each unit time period.

[0084] Initial population generation: Randomly generate an initial population that satisfies the constraint conditions, and each individual represents a possible production scheduling plan.

[0085] (2) Fitness evaluation

[0086] Objective function calculation: For each individual, calculate its corresponding objective function value according to the input data such as the order feature matrix and the production equipment set production efficiency matrix. The objective function value reflects the comprehensive performance of this production scheduling plan, including multiple aspects such as production efficiency, production time, and order deadline.

[0087] Fitness value determination: Determine the fitness value of each individual based on the objective function value. In the genetic algorithm, the fitness value is used to measure the quality of an individual in the population and is usually inversely proportional to the objective function value (i.e., the smaller the objective function value, the higher the fitness value).

[0088] (3) Selection operation

[0089] Selection strategy: Select some excellent individuals (individuals with high fitness values) as parents according to the fitness values of the individuals for generating the next generation.

[0090] (4) Crossover operation

[0091] Crossover method: Perform crossover operations (such as two-point crossover) on the selected parent individuals to generate new individuals (offspring).

[0092] Crossover probability: Set a crossover probability to control the execution frequency of the crossover operation.

[0093] (5) Mutation operation

[0094] Mutation method: Perform mutation operations (such as substitution) on the offspring individuals to introduce new gene information and increase the diversity of the population.

[0095] Mutation probability: Set a mutation probability to control the execution frequency of the mutation operation.

[0096] (6) Iterative update

[0097] Generate a new generation of population: Combine the offspring individuals obtained after selection, crossover, and mutation operations with the elite individuals to form a new generation of population.

[0098] Termination condition judgment: Check whether the termination conditions are met (such as reaching the maximum number of iterations, the fitness value reaching the preset threshold, etc.). If the termination conditions are met, output the optimal solution; otherwise, return to step 2 and continue the iterative update.

[0099] (7) Output the optimal solution

[0100] Optimal solution decoding: Convert the encoding of the optimal solution (individual) into a specific production scheduling plan and output it to obtain the order production scheduling queue. This order production scheduling queue can be based on production equipment or on product orders (i.e., production scheduling order data). In this embodiment, the former is taken as an example, but it is not limited.

[0101] In summary, the embodiment of the present application provides a customer order intelligent scheduling system based on multi-source data analysis. The product production line is used to produce multiple products and includes a central control device and M production devices. Each production device is a dedicated device or a shared device. The dedicated device only participates in the single production process of the product, and the shared device can participate in multiple production processes of multiple products. The customer order intelligent scheduling system is installed in the central control device and includes: a data acquisition module for acquiring multi-source data for order scheduling. The multi-source data for order scheduling includes N scheduling order data. Each scheduling order data includes product type, product quantity, order deadline, and production equipment and production process required for product production. Each scheduling order data is used for the production of a single product (by preprocessing to split the customer's multi-product order into single-product orders to distinguish different product types, quantities, deadlines, etc.); a data processing module for preprocessing and feature extraction of the multi-source data for order scheduling to generate input data for optimizing scheduling; an intelligent scheduling module for constructing an intelligent scheduling model based on the input data for optimization and solution, determining the order scheduling queue, and controlling the product production line to produce products based on the order scheduling queue. By introducing the reusability of production equipment and realizing parallel scheduling of multiple order queues based on multi-source data, introducing multi-source data analysis can obtain the status of the production system in real time and dynamically adjust the scheduling plan based on these data, so as to better adapt to the complexity, diversity, and dynamics of the modern production environment. By constructing an intelligent scheduling model, it is possible to comprehensively consider various factors such as order requirements, production equipment performance, and production process to optimize the production task allocation. This not only helps to ensure the timely completion of orders, but also improves the working efficiency of production equipment, reduces resource waste, and realizes energy conservation and cost reduction. The system can process customer orders containing multiple product types and complex workloads. Through parallel scheduling, the overall order processing cycle is effectively shortened, and customer satisfaction is improved. The system not only supports dedicated devices, but also can make full use of shared devices to participate in multiple production processes of multiple products, thus improving the flexibility and response speed of the production line.

[0102] When preprocessing and extracting features from multi-source data for order scheduling, the system not only extracts basic information about the order (such as order number, product type, product quantity, order deadline), but also goes deep into the production level and extracts production features related to production equipment, product process, process sequence, and production time. This comprehensive feature extraction method helps the system understand order requirements and production conditions more accurately, providing a solid foundation for subsequent optimization and scheduling. Processing the input data into an order feature matrix can ensure the consistency of the input data (N data items are variable, (3M+S) remains constant), and can represent a large amount of order data in an efficient and compact way, which is not only convenient for subsequent calculations and analysis, but also helps to improve the operating efficiency and response speed of the system. Based on this matrix, the system can use advanced optimization algorithms (such as genetic algorithms, particle swarm algorithms, etc.) to find the optimal scheduling plan. This data-driven optimization method can fully consider various constraints (such as equipment efficiency, production time, etc.) to ensure the effectiveness and feasibility of the scheduling plan.

[0103] By setting the production efficiency matrix for M production equipment in a unit time period, the system can accurately reflect the production efficiency of each production equipment in different product processes. This helps the system to fully consider the actual capacity of the production equipment when allocating production tasks and ensure the rationality of the allocation. Setting the production efficiency matrix also makes it easier for the system to monitor and optimize the performance of production equipment, thereby improving overall production efficiency. Based on the order feature matrix and the set production efficiency matrix of each production equipment, the production task allocation function is defined (including three parts of the optimization function. By minimizing the three components A, B, and C, the system can comprehensively consider multiple factors such as production efficiency, production time, and order deadline, so as to find the optimal production scheduling plan. Part A in the objective function takes into account the efficiency balance of the production equipment, which helps to avoid equipment overload or idleness; Part B considers the minimization of production time, which helps to reduce production costs; Part C considers the timely delivery of orders, which helps to improve customer satisfaction. In this way, factors such as the production efficiency of the production equipment, the task processing time of each order scheduling data, and the completion cycle of the entire order scheduling task can be comprehensively considered. Since the supply can be supplemented according to the situation and is highly affected by external influences and human intervention, it is directly assumed that the supply is sufficient and not considered. On the production line of the fully automatic process, the raw material supply can also be taken into consideration to jointly construct the production task allocation function), so that the system can flexibly adjust the production task allocation according to the actual situation, which helps to respond quickly to customer needs. The production task allocation function can also ensure the balanced distribution of production tasks among multiple production equipment, avoid overloading or idleness of some equipment, thereby improving resource utilization, and can take into account the continuity between the various processes of the task as much as possible, shorten the production cycle of each product, and avoid the situation where a large number of process parts are accumulated.

[0104] The equipment efficiency constraint ensures that the production task allocation does not exceed the actual capacity of the production equipment, avoiding the risks of equipment overload and damage. The production time constraint ensures that the start and end times of orders conform to the actual production process, avoiding production delays and order backlogs. The product process constraint ensures the sequentiality and continuity of the product processes with the same order number during the production process, helping to ensure product quality and production efficiency. By optimizing and solving the objective function based on the constraints, the system can quickly find the optimal production scheduling plan. This helps enterprises improve production efficiency and reduce production costs. The optimization and solution process can also consider various constraints and objective functions, thus generating a production scheduling plan that better meets the actual needs. Through the order production scheduling queue obtained by the optimization and solution, the system can guide the specific task allocation and time arrangement in the production process. This helps enterprises achieve visualization and controllability of the production process and improve the refinement level of production management. The production scheduling queue can also be used as an important basis for the production plan, helping enterprises reasonably arrange production resources and inventory and reduce production risks and costs.

[0105] In this article, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations.

[0106] The above are only embodiments of the present application and are not used to limit the protection scope of the present application. For those skilled in the art, the present application can have various changes and modifications. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A customer order intelligent production scheduling system based on multi-source data analysis, characterized in that: The product production line is used to produce a variety of products, including a central control device and M production equipment. Each production equipment is a dedicated device or a shared device. Dedicated equipment only participates in a single production process of a product, and shared equipment can participate in multiple production processes of multiple products. The customer order intelligent scheduling system is installed in the central control device, including: The data acquisition module is used to acquire multi-source data of order scheduling. The multi-source data of order scheduling includes N pieces of scheduling order data. Each piece of scheduling order data includes product type, product quantity, order deadline, and production equipment and production process required for product production. Each piece of scheduling order data is used for the production of a single product. The data processing module is used to preprocess and extract features from multi-source data of order scheduling and generate input data for optimizing scheduling; The intelligent production scheduling module is used to build an intelligent production scheduling model based on input data for optimization and solution, determine the order scheduling queue, and control the product production line to produce products based on the order scheduling queue; The data processing module is specifically used to: pre-process the multi-source data of order scheduling; extract the features of each scheduled order data after pre-processing, and extract the order features and production features of each scheduled order data, where the order features reflect the product production requirements for completing the scheduled order data, and the production features reflect the production equipment, product process and production time required to complete the scheduled order data; integrate the order features and production features of each scheduled order data, and construct an order feature matrix as input data for optimizing scheduling; The intelligent production scheduling module is specifically used to: obtain the set production efficiency matrix of M production equipment in a unit time period; define the production task allocation function of each production equipment based on the order feature matrix and the set production efficiency matrix of each production equipment; construct the objective function to be optimized based on the production task allocation function; generate constraint conditions based on the order feature matrix and the set production efficiency matrix of each production equipment; optimize and solve the objective function based on the constraint conditions to obtain the order production scheduling queue; The set production efficiency matrix of M production equipment in a unit time period is: ， in, It represents the set production efficiency of the first production equipment for the first product process in the set of all product processes within a unit time period. is the total number of production equipment, The total number of processes for all product types supported by the product production line; Intelligent production scheduling module, specifically used for: The set production efficiency of each production equipment is , For the The production equipment is responsible for the first The production efficiency of each product process in a unit time period is defined as The production task allocation function of each production equipment for: ， in, For all product process sets The product process is allocated to the first Production equipment accounts for The ratio of the set production efficiency of each production equipment; Intelligent production scheduling module, specifically used for: based on the order feature matrix, Production order data: based on The order number of the production order data is determined at the start time node of the unit time period corresponding to the product process with the smallest process sequence that appears for the first time. The order start node of the production order data , and, based on the completion of The end time node of the unit time period corresponding to the product process with the largest process sequence when the product quantity of the order number of the production order data is determined. The order end node of the production order data ; The objective function to be optimized is constructed as: ， ， ， ， in, is the objective function to be optimized, , , Function The three components of is the number of production equipment, For all product process sets The product process is allocated to the first Production equipment accounts for The ratio of the set production efficiency of each production equipment, is the total number of operations for all product types supported by the product production line. For the The production equipment is The production efficiency per effective unit period, For the The total number of valid unit time periods of the production equipment. The valid unit time period indicates the unit time period from the time when the production equipment participates in the production of the products of these N production scheduling order data to the time when the production of the products of these N production scheduling order data ends. For the The efficiency weight value of each production equipment, is the total number of production order data. For the The order start node of the production order data. For the The order end node of the production order data. is the regulating factor, Completed for single order production The total production time of the products in the production order data. For the The order deadline of the production order data. For the The order entry time of the production order data, For the The real-time weight value of the production order data.

2. The customer order intelligent production scheduling system based on multi-source data analysis according to claim 1 is characterized in that: The production process includes the product process, process sequence and production time of the production equipment corresponding to the product type, and the data processing module is specifically used for: For each production order data after preprocessing: Extract the order number, product type, product quantity, and order deadline from this production order data to generate a A one-dimensional vector of , as the order feature; Perform one-hot encoding on the production equipment corresponding to the product type in this production order data, and expand the attributes corresponding to each production equipment in the one-hot encoding feature vector. The extended attributes include product process, process sequence, and production time. The extended one-hot encoding feature vector is completed based on the product process, process sequence, and production time of the production equipment to obtain a length of A one-dimensional vector of , as the production feature; Combine order features and production features to obtain the integrated features of each production order data; Will The integrated feature combination of production order data is The order feature matrix of .

3. The customer order intelligent production scheduling system based on multi-source data analysis according to claim 1 is characterized in that: Intelligent production scheduling module, specifically used for: Based on the order feature matrix and the set production efficiency matrix of each production equipment, the following constraints are generated: Equipment efficiency constraints: , Production time constraints: .

4. The customer order intelligent production scheduling system based on multi-source data analysis according to claim 3 is characterized in that: The product process corresponding to the same order number, the product process of the next process order in front The effective unit time period is determined by the production equipment Participate in production, product process of the previous process sequence in front The effective unit time period is determined by the production equipment Participate in production, intelligent scheduling module, also used for: Generate product process constraints: ， in, For product process Allocate to production equipment within a unit period Shizhan production equipment The ratio of setting production efficiency, For production equipment Product Process The set production efficiency within a unit period, For product process In the The production efficiency per effective unit period, For product process Allocate to production equipment within a unit period Shizhan production equipment The ratio of setting production efficiency, For production equipment Product process The set production efficiency within a unit period, For product process in front The sum of production efficiency per effective unit period, For product process in front The sum of production efficiency per effective unit period.

5. The customer order intelligent production scheduling system based on multi-source data analysis according to claim 1 is characterized in that: Genetic algorithm or particle swarm algorithm is used to optimize the objective function.

Images