Liquid crystal display module production line multi-variety order dynamic scheduling management method and system

Abstract

The application provides a liquid crystal display module production line multi-variety order dynamic scheduling management method and system, relates to the technical field of industrial automation, and the method comprises the following steps: collecting the size, resolution, interface type, batch, and delivery date of all orders, simultaneously acquiring production line process equipment, jig configuration, material inventory state, and material supply plan data, obtaining an order data set and a resource data set; based on the order data set and the resource data set, an order and resource associated data point set is constructed, a three-dimensional virtual voxel space is defined in a virtual structure, the order and resource associated data point set is mapped into corresponding voxels one by one, and the number of data points, corresponding order type, resource number, and material demand matching degree information of each voxel falling into the three-dimensional virtual voxel space are recorded, so that a voxel sub-region with data attribute labels is obtained. The application realizes dynamic optimization of a scheduling scheme.

Description

Technical Field

[0001] This invention relates to the field of industrial automation technology, and in particular to a method and system for dynamic scheduling management of multi-variety orders in liquid crystal display module production lines. Background Technology

[0002] As the demand for liquid crystal display modules (LCM) in consumer electronics, automotive displays and other fields rapidly evolves towards customization and differentiation, LCM production lines generally face production pressures of multiple varieties, small batches and tight delivery times. Moreover, the production lines involve multiple core processes such as bonding, FOG, and COG, with highly specialized equipment and fixtures, and complex material supply chains. Order scheduling must simultaneously take into account equipment load balancing, process changeover loss control and material availability requirements.

[0003] A certain LCM manufacturer once undertook three types of customized LCM orders for different application scenarios. The traditional scheduling method only statically sorted the orders according to the urgency of delivery, without considering the differences in changeover losses caused by the need to change special fixtures between some orders, and without linking the supply delay information of key materials. As a result, after the production line started production according to the static sequence, it was forced to stop due to the unexpected time consumption of fixture changeover and the shortage of key materials. The delivery of some orders was delayed, and some equipment was idle, resulting in insufficient capacity utilization. This exposed the technical defects of traditional static scheduling, such as the lack of dynamic correlation modeling between orders and resources, the failure to quantify process changeover costs, and the disconnect between materials and production plans. It is difficult to adapt to complex production scenarios with multiple intertwined variables. Summary of the Invention

[0004] The technical problem to be solved by this invention is to provide a method and system for dynamic scheduling management of multi-variety orders in liquid crystal display module production lines, so as to realize dynamic optimization of scheduling schemes.

[0005] To solve the above-mentioned technical problems, the technical solution of the present invention is as follows:

[0006] Firstly, a method for dynamic scheduling management of multi-variety orders in a liquid crystal display module production line, the method comprising:

[0007] Collect all order data including size, resolution, interface type, batch size, and delivery date. Simultaneously, acquire production line process equipment, fixture configuration, material inventory status, and material supply plan data to obtain order datasets and resource datasets.

[0008] Based on the order dataset and resource dataset, a set of data points associated with orders and resources is constructed, and a three-dimensional virtual voxel space is defined in the virtual structure. The set of data points associated with orders and resources is mapped one by one to the corresponding voxels, and the number of data points falling into the three-dimensional virtual voxel space, the corresponding order type, resource number and material requirement matching information of each voxel are recorded to obtain voxel sub-regions with data attribute labels.

[0009] The adjustment value is calculated based on the characteristics of the voxel sub-region;

[0010] Based on the adjustment value, order dataset, and resource dataset, the process changeover cost is quantified to obtain the process changeover cost matrix;

[0011] Based on the order dataset, process changeover cost matrix, and adjustment values, the priority of each order is dynamically calculated;

[0012] Based on the dynamic priority of orders, the process changeover cost matrix, and adjustment values, with the optimization objectives of minimizing total changeover cost and maximizing order delivery rate, orders are allocated to each production line process to obtain an initial scheduling sequence.

[0013] Based on the initial scheduling column, the system monitors order change events and material supply status in real time, recalculates the dynamic priority of orders based on the current resource dataset, and updates the scheduling column.

[0014] During the generation and adjustment of the scheduling sequence, the material demand planning data is linked in real time, and the scheduling order is adjusted according to the material arrival time.

[0015] Furthermore, data on the size, resolution, interface type, batch size, and delivery date of all orders are collected. Simultaneously, data on production line equipment, fixture configuration, material inventory status, and material supply plans are acquired to obtain order datasets and resource datasets, including:

[0016] Through the first data interface, the raw data of all orders to be scheduled is received synchronously;

[0017] Based on the production resource requirements corresponding to the received raw data, the system collects the associated production line resource status data in real time from the IoT platform and warehouse management via the second data interface.

[0018] The received raw data is integrated with the collected production line resource status data to obtain integrated data. The integrated data is then cleaned and standardized to eliminate outliers and format differences, resulting in order datasets and resource datasets.

[0019] Furthermore, based on the order dataset and resource dataset, a set of order and resource-related data points is constructed, and a three-dimensional virtual voxel space is defined in the virtual structure. The set of order and resource-related data points is mapped one by one to the corresponding voxels, and the number of data points falling into the three-dimensional virtual voxel space, the corresponding order type, resource number, and material requirement matching information for each voxel are recorded, resulting in voxel sub-regions with data attribute labels, including:

[0020] Based on the order dataset and resource dataset, we perform association matching to match the attribute features of each order with the required process equipment, fixtures and material resources, and construct a set of order and resource association data points.

[0021] Based on the core feature dimensions of orders and resources contained in the set of associated data points, a three-dimensional virtual voxel space is defined in the virtual structure;

[0022] Based on the defined three-dimensional virtual voxel space, the voxelization algorithm is invoked to map and assign each data point in the constructed set of order and resource-related data points to the corresponding voxel according to the attribute coordinates of each data point.

[0023] For each voxel containing data points after mapping, record the internal features of each voxel. The features include at least the number of data points falling into the voxel, the set of order types corresponding to the data points, the list of associated resource numbers, and the material demand matching degree calculated based on the material inventory status, thereby forming multiple voxel sub-regions with data attribute labels.

[0024] Furthermore, based on the characteristics of the voxel sub-regions, adjustment values ​​are calculated, including:

[0025] Based on the formed voxel sub-regions with data attribute labels, the key feature parameters of each voxel sub-region are extracted. The key feature parameters include at least the data point density, the real-time load status of the associated resources, and the material demand matching degree.

[0026] Based on the extracted key feature parameters, and according to the preset resource optimization and delivery urgency rules, a comprehensive initial adjustment value is calculated for each voxel sub-region.

[0027] Furthermore, based on the adjustment values, order datasets, and resource datasets, the process changeover cost is quantified to obtain a process changeover cost matrix, including:

[0028] Based on the initial adjustment values ​​of each element sub-region, and combined with the order attributes in the order dataset and the equipment and fixture configuration in the resource dataset, calculate the basic value of the process switching cost between any two order types.

[0029] Based on the real-time status data of centralized equipment and fixtures in the resource dataset, the calculated basic value of process changeover cost is dynamically corrected to obtain the corrected process changeover cost that reflects changes in the real-time production environment.

[0030] The corrected process switching costs between all pairs of order types are used to construct a complete process switching cost matrix.

[0031] Furthermore, based on the order dataset, process changeover cost matrix, and adjustment values, the priority of each order is dynamically calculated, including:

[0032] Based on the order delivery date and batch information in the order dataset, the delivery urgency and production load coefficient of each order are calculated to obtain the initial priority base value;

[0033] Based on the process changeover cost matrix and the current order status being executed on the production line, calculate the changeover cost impact factor of each pending order relative to the current production status.

[0034] Based on the initial adjustment values ​​of each voxel sub-region, obtain the resource optimization adjustment coefficients of the voxel sub-regions associated with each order;

[0035] The initial priority base value, switching cost impact factor, and resource optimization adjustment coefficient are weighted and integrated to dynamically calculate the real-time comprehensive priority of each order.

[0036] Furthermore, based on dynamic order priority, process changeover cost matrix, and adjustment values, with the optimization objectives of minimizing total changeover cost and maximizing order delivery rate, orders are allocated to each production line process, resulting in an initial scheduling sequence, including:

[0037] Based on the obtained real-time comprehensive priority score of the orders, an initial order processing sequence is generated as the basis for scheduling optimization.

[0038] Based on the process switching cost matrix, with the goal of minimizing the total switching cost, the basic scheme is sequentially optimized and adjusted to obtain the sequence optimization result.

[0039] Based on the initial adjustment values ​​of each element region, resource conflict detection and resolution are performed on the obtained preliminary optimization sequence to obtain the resource conflict resolution results.

[0040] By combining the sequence optimization results and resource conflict resolution results, and taking maximizing the order delivery rate as the final constraint, an executable initial scheduling sequence is obtained.

[0041] Furthermore, based on the initial scheduling column, order change events and material supply status are monitored in real time. The dynamic priority of orders is recalculated according to the current resource dataset, and the scheduling column is updated, including:

[0042] Based on an executable initial scheduling sequence, a real-time monitoring mechanism is established to continuously monitor updated data on order change events and material supply status.

[0043] When the monitoring mechanism identifies order change events and material supply status updates, it triggers the scheduling recalculation process and re-executes the dynamic priority calculation of orders based on the latest resource dataset.

[0044] Based on the recalculated dynamic priority of orders, the scheduling optimization algorithm is invoked to dynamically adjust and update the original initial scheduling sequence, resulting in the latest executable scheduling sequence for production status.

[0045] Furthermore, during the generation and adjustment of scheduling sequences, material requirements planning data is linked in real time, and the scheduling order is adjusted according to the material arrival time, including:

[0046] During the generation of the initial scheduling sequence and the dynamic adjustment of the sequence, the material requirements planning is linked in real time to obtain the latest material requirements planning data and the estimated arrival time;

[0047] Based on the latest material requirements planning data, determine the material availability status of each order in the scheduling sequence and identify order production conflicts caused by material delays.

[0048] Based on the expected arrival time of materials, the order of orders with material shortage risk in the scheduling sequence is adaptively adjusted to obtain a final executable scheduling sequence that matches the material supply plan, thus ensuring the feasibility of the production plan.

[0049] Secondly, the dynamic scheduling management system for multi-variety orders on the LCD module production line includes:

[0050] The acquisition module is used to collect data on the size, resolution, interface type, batch size, and delivery date of all orders. It also acquires data on production line process equipment, fixture configuration, material inventory status, and material supply plan to obtain order datasets and resource datasets.

[0051] The construction module is used to build a set of order and resource associated data points based on the order dataset and resource dataset, and define a three-dimensional virtual voxel space in the virtual structure. The set of order and resource associated data points is mapped one by one to the corresponding voxels, and the number of data points falling into the three-dimensional virtual voxel space, the corresponding order type, resource number and material requirement matching information of each voxel are recorded to obtain voxel sub-regions with data attribute labels.

[0052] The calculation module is used to calculate the adjustment value based on the characteristics of the voxel sub-region;

[0053] The quantification module is used to quantify process changeover costs based on adjustment values, order datasets, and resource datasets, and obtain a process changeover cost matrix.

[0054] The priority module is used to dynamically calculate the priority of each order based on the order dataset, the process changeover cost matrix, and adjustment values.

[0055] The optimization module is used to allocate orders to each production line process based on the order dynamic priority, process changeover cost matrix and adjustment value, with the optimization objectives of minimizing total changeover cost and maximizing order delivery rate, and to obtain the initial schedule.

[0056] The monitoring module is used to monitor order change events and material supply status in real time based on the initial scheduling sequence, recalculate the dynamic priority of orders according to the current resource dataset, and update the scheduling sequence.

[0057] The processing module is used to link material demand planning data in real time during the generation and adjustment of scheduling sequences, and adjust the scheduling order according to the material arrival time.

[0058] The above-described solution of the present invention has at least the following beneficial effects:

[0059] By employing a combination of technologies including dual-interface data acquisition and integration, 3D virtual voxel space correlation modeling, construction of a process changeover cost quantification matrix, dynamic priority weighted calculation of orders, real-time monitoring and scheduling updates, and linkage adjustment of material requirements, this approach effectively overcomes the technical problems of traditional static scheduling, such as the lack of deep dynamic correlation between orders and resources, unquantified process changeover costs, disconnect between material supply and production plans, and difficulty in responding to order changes and resource status fluctuations. This enables dynamic optimization of scheduling schemes, maximizing order delivery rates while reducing process changeover losses, balancing equipment load, improving overall production line capacity utilization and flexible manufacturing capabilities, and ensuring the feasibility and efficiency of production plans. Attached Figure Description

[0060] Figure 1 This is a flowchart illustrating the dynamic scheduling management method for multi-variety orders in a liquid crystal display module production line provided by an embodiment of the present invention.

[0061] Figure 2 This is a schematic diagram of a dynamic scheduling management system for multi-variety orders in a liquid crystal display module production line provided in an embodiment of the present invention. Detailed Implementation

[0062] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.

[0063] like Figure 1 As shown, embodiments of the present invention propose a dynamic scheduling management method for multi-variety orders in a liquid crystal display module production line, the method comprising the following steps:

[0064] Step 1: Collect data on size, resolution, interface type, batch size, and delivery date for all orders. At the same time, obtain data on production line process equipment, fixture configuration, material inventory status, and material supply plan to obtain order dataset and resource dataset.

[0065] Step 2: Based on the order dataset and resource dataset, construct a set of data points associated with orders and resources, and define a three-dimensional virtual voxel space in the virtual structure. Map the set of data points associated with orders and resources to the corresponding voxels one by one, and record the number of data points falling into the three-dimensional virtual voxel space, the corresponding order type, resource number and material requirement matching information for each voxel, to obtain voxel sub-regions with data attribute labels.

[0066] Step 3: Calculate the adjustment value based on the characteristics of the voxel sub-region;

[0067] Step 4: Based on the adjustment value, order dataset, and resource dataset, quantify the process changeover cost to obtain the process changeover cost matrix;

[0068] Step 5: Based on the order dataset, process changeover cost matrix, and adjustment values, dynamically calculate the priority of each order;

[0069] Step 6: Based on the dynamic priority of orders, the process changeover cost matrix, and the adjustment value, with the optimization objectives of minimizing the total changeover cost and maximizing the order delivery rate, allocate orders to each production line process to obtain the initial scheduling sequence;

[0070] Step 7: Based on the initial scheduling column, monitor order change events and material supply status in real time, recalculate the dynamic priority of orders according to the current resource dataset, and update the scheduling column;

[0071] Step 8: During the generation and adjustment of the scheduling sequence, the material demand plan data is linked in real time, and the scheduling order is adjusted according to the material arrival time.

[0072] In this embodiment of the invention, by employing a combination of technical means—collecting full-attribute data of orders and full-state data of production line resources, constructing a three-dimensional virtual voxel space to realize order and resource association modeling, calculating voxel sub-region adjustment values, quantifying process switching costs and constructing a matrix, dynamically calculating order priorities, generating initial scheduling sequences, and updating scheduling through real-time monitoring of order and material status, and adjusting scheduling order order in conjunction with material demand planning—the invention overcomes the technical problems of traditional scheduling, such as lack of precise association between orders and resources, lack of quantitative standards for process switching costs, static and fixed order priorities, difficulty in responding to order changes and material supply fluctuations during production, and disconnection between scheduling schemes and actual production conditions. This achieves the technical effects of minimizing total production line switching losses, maximizing on-time order delivery rate, balancing equipment load, improving overall production line capacity utilization and flexible manufacturing response speed, and ensuring that scheduling schemes are scientifically feasible and adaptable to dynamic production environments.

[0073] In a preferred embodiment of the present invention, step 1 above may include:

[0074] Step 1.1: Through the first data interface, synchronously receive the raw data of all orders to be scheduled. Specifically, this includes: deploying a dedicated first data interface, establishing a stable connection between the interface and the order management system, and continuously receiving the raw information of all customized LCD display module orders that need to be scheduled according to a preset synchronization frequency. This raw information covers key information such as the product size, resolution, interface type, production batch, and delivery deadline of the order, ensuring that complete basic data of each order to be scheduled is obtained without omission.

[0075] Step 1.2: Based on the production resource requirements corresponding to the received raw data, the production line resource status data associated with it is collected in real time from the IoT platform and warehouse management through the second data interface. Specifically, this includes: analyzing the production resource requirements of each order in the production process, such as process equipment, matching fixtures, and various materials, based on the received original order information; and establishing data interaction channels with the IoT platform and warehouse management through the pre-configured second data interface to collect production line resource status information associated with the order resource requirements in real time.

[0076] Step 1.3 involves integrating the received raw data with the collected production line resource status data to obtain integrated data. The integrated data is then cleaned and standardized to eliminate outliers and format differences, resulting in an order dataset and a resource dataset. Specifically, this includes: establishing a one-to-one correspondence between all received raw order data and the corresponding collected production line resource status data to form an integrated comprehensive dataset; inspecting the integrated comprehensive data to identify and remove outliers, while simultaneously standardizing the format of data from different sources to eliminate format differences caused by varying data sources, ultimately forming a unified, accurate, and complete order dataset and resource dataset.

[0077] In this embodiment of the invention, by employing a dual data interface for collaborative work and by cleaning and standardizing the two types of integrated data to eliminate outliers and format differences, the technical problems of unreliable datasets caused by asynchronous acquisition of order data and production line resource data, fragmented data sources, and inconsistent data quality in the traditional data acquisition process are overcome. This achieves the goal of quickly acquiring complete, accurate, and standardized order datasets and resource datasets.

[0078] In a preferred embodiment of the present invention, step 2 above may include:

[0079] Step 2.1: Based on the order dataset and resource dataset, perform association matching to map the attribute characteristics of each order to the required process equipment, fixtures, and material resources, constructing a set of order-resource association data points. Specifically, this includes: first, reviewing the various attribute characteristics contained in the order dataset, covering key information related to order production such as product size, resolution, interface type, production batch, and delivery deadline; and clarifying the details of various resources in the resource dataset, including the model, operating parameters, and availability of equipment for each process, the compatible product types and configuration quantities of various fixtures, and the name, specifications, inventory quantity, and supply plan of various materials; then, for each order to be scheduled, based on its attribute characteristics... The required production resources are matched item by item. For example, the appropriate COG fixture model is determined based on the size and resolution of the ordered products, the corresponding bonding process equipment is matched based on the interface type, the required quantity of materials such as panels and backlight modules is calculated based on the production batch, and the equipment resources that need to be prioritized in the production process are associated based on the delivery deadline. Through this item-by-item matching, a unique association is established between the complete attribute characteristics of each order and all process equipment, matching fixtures and various material resources required for the entire production process of that order. Each set of matched order attributes and resource requirements forms an independent associated data point. All associated data points are summarized and integrated to construct a complete set of order and resource associated data points.

[0080] Step 2.2: Based on the core feature dimensions of orders and resources contained in the associated data point set, a three-dimensional virtual voxel space is defined in the virtual structure. Specifically, this includes: comprehensively analyzing the completed set of associated data points for orders and resources, extracting the core feature dimensions of the included orders and resources, and determining three key dimensions as the coordinate references for the three-dimensional virtual voxel space based on the production characteristics and scheduling requirements of the LCD module production line. The first dimension is set as the core process requirement category corresponding to the order, covering the process requirements classification related to key processes such as Bonding, FOG, and COG. The second dimension is set as the matching model category of equipment and fixtures, classified according to the model specifications and matching range of existing equipment and fixtures in the production line. The third dimension is set as the requirement category of key materials, focusing on core material types that have a significant impact on production progress, such as panels, backlight modules, and driver ICs. In the virtual structure environment, a three-dimensional virtual voxel space framework is built according to the three core feature dimensions, clarifying the division criteria and value range of each dimension to ensure that the space fully covers the core process requirements of all orders, the matching characteristics of all production resources, and the requirement types of key materials.

[0081] Step 2.3: Based on the defined 3D virtual voxel space, the voxelization algorithm is invoked to map and assign each data point in the constructed set of order and resource-related data points to the corresponding voxel according to the attribute coordinates of each data point. Specifically, this includes: first, refining the defined 3D virtual voxel space by decomposing each dimension into several equal or adapted voxel units according to the set classification criteria, and clarifying the coordinate range of each voxel unit in the 3D space; then, starting and invoking the voxelization algorithm to process each data point in the set of order and resource-related data points one by one. For each data point, the core process requirement category of the order, the corresponding equipment and fixture model category, and the key material requirement category are analyzed first. Based on this, the specific attribute coordinates of the data point in the three-dimensional virtual voxel space are determined. According to the calculated attribute coordinates, the corresponding voxel unit is located in the three-dimensional virtual voxel space, and the data point is mapped and assigned to the voxel unit one by one. During the mapping process, the matching accuracy between the attribute coordinates of each data point and the voxel unit is continuously checked to ensure that all related data points can accurately fall into the corresponding voxel without mismatch or omission.

[0082] Step 2.4: For each voxel containing data points after mapping, record the internal characteristics of each voxel. These characteristics include at least the number of data points falling into the voxel, the set of order types corresponding to the data points, the list of associated resource numbers, and the material demand matching degree calculated based on the material inventory status. This forms multiple voxel sub-regions with data attribute labels. Specifically, this includes: traversing all voxel units in the 3D virtual voxel space, filtering out voxel units containing associated data points, and comprehensively recording the internal characteristics of each such voxel unit; first, counting the total number of associated data points falling into the voxel unit to reflect the concentration of order and resource association characteristics corresponding to that voxel; then, organizing the order types corresponding to all data points, based on the application scenario and product specifications of the orders. The data points are categorized by attributes such as grid to form a set of order types corresponding to each voxel unit. Then, all resource information associated with each data point is extracted, including the unique number of the process equipment, the unique number of the fixture, and the identification number of the material. These are then organized into a list of associated resource numbers by equipment, fixture, and material. Based on the real-time inventory status of various materials centrally recorded in the resource dataset, the material demand quantity corresponding to all data points in the voxel unit is compared to calculate the material demand satisfaction level of each data point. This is then summarized to obtain the overall material demand matching degree of the voxel unit. The statistically calculated feature information is bound to the corresponding voxel unit so that each voxel unit containing data points has complete attribute labels, ultimately forming multiple voxel sub-regions with clear data attribute labels.

[0083] In this embodiment of the invention, the technical means of constructing a set of order and resource associated data points by performing association matching based on order datasets and resource datasets, defining a three-dimensional virtual voxel space according to the core feature dimension of the associated data point set, calling the voxelization algorithm to map the associated data points one by one to the corresponding voxels and recording the internal features of each voxel to form a combination of voxel sub-regions with data attribute labels are used. Therefore, the technical problems of lack of accurate dynamic association between orders and required process equipment, fixtures, and material resources in traditional scheduling, and the difficulty in intuitively presenting the association relationship are overcome. Thus, the visualization modeling of the association relationship between orders and resources is realized, and the order clustering characteristics and resource matching status are clearly presented.

[0084] In a preferred embodiment of the present invention, step 3 above may include:

[0085] Step 3.1: Based on the formed voxel sub-regions with data attribute labels, extract the key feature parameters for each voxel sub-region. These key feature parameters include at least data point density, real-time load status of associated resources, and material demand matching degree. Specifically, this includes: first, identifying the core information related to scheduling optimization within the voxel sub-regions with data attribute labels; then, extracting key feature parameters for each voxel sub-region. For the data point density parameter, first determine the specific spatial range of each voxel sub-region in the 3D virtual voxel space, count the total number of mapped order and resource-related data points within the spatial range, and then calculate the data point density of each voxel sub-region by the ratio of this number to the spatial range of the voxel sub-region. Data point density directly reflects the concentration of order and resource-related features within the region; the more concentrated the associated data points for the same type of order, the higher the data point density. For the real-time load status parameter of associated resources, extract all relevant process equipment numbers and fixture numbers from the associated resource number list recorded in the voxel sub-region. Based on these numbers, the system retrieves the corresponding equipment's real-time operating status, cumulative operating time, remaining available time, and the number of occupied and idle fixtures from the resource dataset. The equipment load rate is calculated by the ratio of cumulative operating time to total available time, and the fixture load rate is calculated by the ratio of occupied fixtures to total available fixtures. Combining the equipment load rate and fixture load rate, the system forms the real-time load status parameters for the associated resources of each voxel sub-region, clarifying the current resource busyness and remaining capacity. For the material demand matching parameter, based on the material demand information recorded in the voxel sub-region and the material inventory status and material supply plan data in the resource dataset, the system calculates the total demand for key materials corresponding to all associated data points within the voxel sub-region. This is compared with the quantity that can be directly called upon in the current material inventory, and then combined with the quantity expected to arrive in the material supply plan. The system calculates the proportion that the existing inventory plus supply can meet the material demand of the voxel sub-region. This proportion is the material demand matching parameter; a higher proportion indicates a stronger material guarantee capability for orders within the region.

[0086] Step 3.2: Based on the extracted key feature parameters, and according to the preset resource optimization and delivery urgency rules, calculate a comprehensive initial adjustment value for each voxel sub-region. Specifically, this includes: first, combining the production management goals of the LCD module production line, preset resource optimization and delivery urgency rules. The resource optimization rules explicitly require prioritizing balanced equipment load to avoid some equipment being overworked while others are idle for extended periods, while also improving the utilization rate of key equipment and fixtures. The delivery urgency rules explicitly require prioritizing orders with tight delivery schedules, while also considering the impact of order batches on production efficiency, to ensure that key orders are delivered on time. Based on this preset rule, reasonable evaluation criteria and weight allocation logic are set for the three key feature parameters extracted. The data point density weight focuses on reflecting the efficiency advantage of centralized production of similar orders, the real-time load status weight of associated resources focuses on reflecting the need for balanced resource utilization, and the material demand matching degree weight focuses on ensuring the feasibility of production. Then, the three key feature parameters of each voxel sub-region are quantitatively evaluated. For example, a high score is given when the data point density reaches the preset high concentration standard, a high score is given when the associated resource load is within the preset reasonable range, and a high score is given when the material demand matching degree reaches the preset compliance ratio. Then, according to the preset weight, the quantitative evaluation results of the three parameters are comprehensively calculated. A higher comprehensive score is given to voxel sub-regions with high data point density, moderate resource load, and high material demand matching degree; a lower comprehensive score is given to voxel sub-regions with low data point density, excessively high or low resource load, and insufficient material demand matching degree. The comprehensive score is the comprehensive initial adjustment value for each voxel sub-region.

[0087] In this embodiment of the invention, by employing the technical means of extracting key feature parameters such as data point density, real-time load status of associated resources, and material demand matching degree from voxel sub-regions with data attribute labels, and calculating a comprehensive initial adjustment value for each voxel sub-region based on preset resource optimization and delivery urgency rules, the technical problem of traditional scheduling lacking comprehensive consideration of the degree of order clustering, real-time resource utilization, and material matching, resulting in scheduling adjustments without quantitative basis and difficulty in balancing resource optimization and order delivery urgency, is overcome, making scheduling decisions more in line with the actual production resource status and delivery needs.

[0088] In a preferred embodiment of the present invention, step 4 above may include:

[0089] Step 4.1: Based on the initial adjustment values ​​of each voxel sub-region, and combining the order attributes in the order dataset with the equipment and fixture configurations in the resource dataset, calculate the base value of the process switching cost between any two order types. Specifically, this includes: First, classifying all orders in the order dataset according to key attributes such as product size, resolution, core process requirements, and compatible equipment and fixture models to divide them into different order types, ensuring that orders of the same type have high similarity in production resource requirements and process flows; Next, determining the voxel sub-region to which each order type belongs, using the initial adjustment value of the voxel sub-region as the basic reference for calculating the switching cost. The higher the initial adjustment value, the better the resource matching degree and the higher the concentration of the corresponding order type, and the lower the potential switching cost; Then, for any two different order types, extracting their order attributes and corresponding... The equipment and fixture configuration information is analyzed in detail to determine the resource changes that will occur during the production process when switching from the former to the latter. This includes determining whether it is necessary to replace special production tools such as COG fixtures and bonding equipment, whether it is necessary to adjust the process parameters of the equipment, and whether it is necessary to change the sequence of key processes in the production flow. For cases requiring equipment or fixture replacement, the number of equipment to be replaced, the fixture model, and the standard time consumption for the replacement operation are calculated. For cases requiring adjustment of process parameters, the estimated time and labor cost for parameter debugging are calculated. For cases involving changes in the sequence of processes, the additional time consumption for process connection is assessed. The time and labor costs incurred in replacement, debugging, and connection are summarized, and then weighted and calibrated in combination with the initial adjustment values ​​of the voxel sub-regions of the two order types. Finally, the basic value of the process switching cost between the two order types is calculated.

[0090] Step 4.2: Based on the real-time status data of equipment and fixtures in the resource dataset, dynamically correct the calculated basic value of process changeover cost to obtain the corrected process changeover cost reflecting changes in the real-time production environment. Specifically, this includes: retrieving the current status data of all equipment and fixtures from the resource dataset in real time, and meticulously analyzing information such as real-time operating parameters, cumulative runtime, current load rate, whether under maintenance, and remaining available operating time for the equipment; simultaneously collecting status data such as the real-time occupancy status of fixtures, the number of idle fixtures, whether cleaning or calibration is required, and the next maintenance time; for each calculated basic value of process changeover cost, analyze the impact of the real-time status on the changeover cost by considering the real-time status of the equipment and fixtures involved in the corresponding changeover process. For example, if the changeover process requires the use of… If the equipment's current cumulative runtime is approaching the maintenance threshold, its stability is declining, and the expected commissioning time will be longer than the standard time, then a corresponding correction coefficient is set based on the degree of equipment wear, and the base value of the switching cost is increased. If the fixture required for switching is currently idle and does not require additional cleaning and calibration, it can be put into use directly, and the switching time will be shortened, then a downward correction coefficient is set, reducing the corresponding base value. If the equipment's current load rate is too high, and the switching process requires waiting for the equipment to become idle, then the corresponding time cost needs to be supplemented based on the waiting time, and the base value is corrected. Following this logic, a corresponding real-time status correction coefficient is matched for each base value of switching cost, and the base value is dynamically adjusted through coefficient calculation, ultimately obtaining the corrected process switching cost that reflects the status of the equipment and fixtures in the current production environment.

[0091] Step 4.3 involves constructing a complete process switching cost matrix from the corrected process switching costs between all pairs of order types. This includes: first, uniformly numbering and organizing all the divided order types to form a complete order type list, specifying the total number of order types in the list; then, constructing a two-dimensional matrix framework using the order types in the list as row and column dimensions, where the rows represent the order types currently in production and the columns represent the planned next order type to switch to, with each cell storing the corrected process switching cost between the corresponding two rows and two columns of order types; next, filling the calculated corrected switching costs between all pairs of order types into the matrix cells one by one according to their corresponding row and column positions. For example, if the currently producing order type is A and the planned switch to order type is B, then the corrected switching cost from A to B is filled into the cell in row A and column B of the matrix. During the filling process, each cell's corresponding order type combination and switching cost value are checked one by one to ensure no errors or omissions occur. Once all the corrected switching costs between all pairs of order types have been filled, a complete process switching cost matrix is ​​formed.

[0092] In this embodiment of the invention, by employing a combination of techniques—including calculating the basic value of process switching costs between any two types of orders based on the initial adjustment values ​​of each element region and combining order attributes with equipment and fixture configurations, dynamically correcting the basic value based on real-time status data of equipment and fixtures, and constructing a complete process switching cost matrix from the corrected switching costs of all pairs of order types—this overcomes the technical problems of traditional static scheduling, such as the lack of quantification of process switching costs, the lack of order and resource correlation in switching cost estimation, and the failure to consider changes in the real-time production environment, which lead to inaccurate switching cost calculations and consequently, production line switching time exceeding expectations and production efficiency decline. This achieves the technical effect of accurately quantifying and dynamically updating process switching costs, thereby improving the continuity of the production process.

[0093] In a preferred embodiment of the present invention, step 5 above may include:

[0094] Step 5.1: Based on the order delivery date and batch information in the order dataset, calculate the delivery urgency and production load coefficient of each order to obtain the initial priority base value. Specifically, this includes: firstly, extracting the delivery deadline and production batch information of each order from the order dataset, and simultaneously retrieving historical production data from the production line to determine the standard production cycle of different types of LCD display modules as a reference benchmark for calculating the delivery urgency. For each order, the remaining time from the current time to the delivery deadline is calculated and compared with the corresponding standard production cycle. If the remaining time is less than or equal to the standard production cycle, it is classified as high urgency; if the remaining time is 1.5 times or less of the standard production cycle, it is classified as medium urgency; and if the remaining time exceeds 1.5 times the standard production cycle, it is classified as low urgency. This classification method quantifies the delivery urgency of each order. Next, the production load coefficient is calculated. Combining the current equipment capacity, personnel configuration, and maximum processing capacity of each process, the impact of each order's production batch on the overall production line load is analyzed. The larger the batch and the more core process resources required, the higher the production load coefficient; the smaller the batch and the fewer core resources required, the lower the production load coefficient. This completes the quantification of the production load coefficient. Finally, according to the preset scheduling strategy, reasonable weights are assigned to the delivery urgency and production load coefficient. The delivery urgency weight focuses on ensuring on-time delivery, while the production load coefficient weight focuses on balancing the overall production line load. The two quantified indicators are multiplied by their corresponding weights and then summed to obtain the initial priority base value for each order.

[0095] Step 5.2: Based on the process switching cost matrix and the current order status being executed on the production line, calculate the switching cost impact factor of each pending order relative to the current production status. Specifically, this includes: first, clarifying the real-time status information such as the order type being executed on the production line, its current production process stage, and the configuration of the equipment and fixtures being used, to determine the core resource occupancy under the current production status; then, retrieving the constructed complete process switching cost matrix, using the currently executed order type as a benchmark, locating the row data corresponding to that order type in the matrix, finding the column cell corresponding to each pending order type in the row data, extracting the corrected process switching cost value stored in the cell, and for each pending order, analyzing the impact of its switching cost value relative to the currently executed order on production. The impact of efficiency is assessed by the following factors: a lower switching cost means less time is spent on equipment and fixture replacement, parameter adjustment, and process coordination when switching from the current order to the scheduled order, resulting in a smaller impact on the continuity of the production process and a higher switching cost impact factor. Conversely, a higher switching cost means more time and resources are wasted during the switching process, resulting in a lower impact factor. By setting a quantification range for the impact factor, the switching cost value is converted into a corresponding coefficient value. For example, when the switching cost is below a preset low threshold, the impact factor is set to a high range value; when the switching cost is within a preset middle threshold range, the impact factor is set to a middle range value; and when the switching cost is above a preset high threshold, the impact factor is set to a low range value. This process ultimately completes the calculation of the switching cost impact factor for each scheduled order relative to the current production status.

[0096] Step 5.3: Based on the initial adjustment values ​​of each voxel sub-region, obtain the resource optimization adjustment coefficients of the voxel sub-regions associated with each order. This specifically includes: constructing a set of order and resource-related data points; identifying the associated data points corresponding to each pending order; determining the voxel sub-region to which the data point falls; establishing a unique correspondence between orders and voxel sub-regions; and retrieving the comprehensive initial adjustment value for each voxel sub-region from the calculated results. The initial adjustment value comprehensively reflects key characteristics such as data point density, real-time load status of associated resources, and material demand matching degree. Different coefficient levels are assigned based on the magnitude of the initial adjustment value; the higher the initial adjustment value, the better the voxel sub-region's performance. The better the resource matching degree of internal orders, the higher the centralized production efficiency, and the stronger the material support capability, the higher the corresponding resource optimization adjustment coefficient. The lower the initial adjustment value, the worse the resource adaptability and the higher the production risk, and the lower the corresponding resource optimization adjustment coefficient. For example, when the initial adjustment value is in the high range, the resource optimization adjustment coefficient is set to a high-level value, which means that orders in this region will receive priority resource allocation; when the initial adjustment value is in the middle range, the coefficient is set to a medium-level value; when the initial adjustment value is in the low range, the coefficient is set to a low-level value. Through this level mapping, the corresponding resource optimization adjustment coefficient can be accurately obtained from the initial adjustment values ​​of the voxel sub-regions associated with each order.

[0097] Step 5.4 involves weighting and integrating the initial priority base value, switching cost impact factor, and resource optimization adjustment coefficient to dynamically calculate the real-time comprehensive priority of each order. Specifically, this includes: pre-setting the weights of the initial priority base value, switching cost impact factor, and resource optimization adjustment coefficient based on the core production goals of the LCD module production line. The weight of the initial priority base value prioritizes ensuring order delivery deadlines and production line load balance; the weight of the switching cost impact factor prioritizes reducing process changeover losses; and the weight of the resource optimization adjustment coefficient prioritizes improving resource utilization efficiency. The weight allocation can be dynamically adjusted according to actual production needs. For example, when the production line faces delivery pressure, the weight of the initial priority base value can be increased. When production line changeover losses are too high, the weight of the changeover cost impact factor is increased. Subsequently, the three parameters are uniformly quantified and calibrated to ensure that they are in the same numerical range, avoiding the influence of fusion results due to differences in units. The initial priority base value of each order is multiplied by the corresponding weight, the changeover cost impact factor is multiplied by the corresponding weight, and the resource optimization adjustment coefficient is multiplied by the corresponding weight. The three products are then summed. All pending orders are sorted according to the size of the sum. The larger the sum, the better the overall performance of the order in terms of delivery urgency, changeover cost economy, and resource adaptation optimization, and the higher the corresponding real-time comprehensive priority. Finally, the real-time comprehensive priority ranking of all pending orders is dynamically calculated.

[0098] In this embodiment of the invention, by employing the technical means of calculating the initial priority base value based on order delivery date and batch information, calculating the impact factor of the switching cost of pending orders based on the process switching cost matrix combined with the current production line order execution status, and obtaining the resource optimization adjustment coefficient from the initial adjustment value of the voxel sub-region, and weighting and fusing these three parameters, the technical problems of order priority in traditional static scheduling, which rely solely on delivery date settings and fail to consider process switching costs and real-time resource optimization needs, leading to inaccurate priority determination and thus causing scheduling imbalance, order delivery delays, or low resource utilization, are overcome. This achieves the dynamic generation of real-time comprehensive order priorities that fit the actual production scenario, enabling scheduling decisions to both ensure the urgency of delivery dates and reduce switching losses and optimize resource allocation.

[0099] In a preferred embodiment of the present invention, step 6 above may include:

[0100] Step 6.1: Based on the obtained real-time comprehensive priority scores of orders, an initial order processing sequence is generated as the basis for scheduling optimization. Specifically, this includes: First, organizing the real-time comprehensive priority scores of all orders to be scheduled, and then initially sorting all orders according to their scores from highest to lowest. The order with the highest score corresponds to the highest comprehensive priority and is placed at the beginning of the production sequence, followed by the order with the second highest score, and so on, forming a preliminary order ranking result. For orders with the same score, the delivery deadline in the order dataset is further considered, prioritizing orders with tighter delivery deadlines. If the delivery deadlines are also the same, the production batch size is considered, prioritizing orders with smaller batch sizes that can be completed quickly to improve production line efficiency. Through the sorting rules, all orders to be scheduled are arranged in an orderly manner to generate an initial order processing sequence. This sequence fully reflects the comprehensive advantages of each order in terms of delivery urgency, switching cost economy, and resource adaptability optimization, serving as the basis for optimization.

[0101] Step 6.2: Based on the process switching cost matrix, with the goal of minimizing the total switching cost, perform sequence optimization adjustments on the basic scheme to obtain the sequence optimization results. Specifically, this includes: retrieving the constructed complete process switching cost matrix to determine the corrected process switching cost between any two adjacent orders in the initial order processing sequence; calculating the total switching cost corresponding to the initial processing sequence, i.e., sequentially accumulating the switching cost values ​​of all adjacent order pairs in the sequence to obtain the basic total switching cost; and performing local adjustments and global optimizations on the initial processing sequence with the core objective of minimizing the total switching cost. Specific optimization methods include adjacent order exchange adjustments, i.e., detecting adjacent orders in the sequence with high switching costs. Yes, the process involves querying the cost matrix to find order combinations with lower switching costs. If swapping two adjacent orders reduces the switching cost of the region without significantly affecting the overall priority order, then the swap operation is performed. This also includes clustering and adjusting similar orders: finding orders belonging to the same sub-region in the sequence with extremely low switching costs, and moving them to adjacent positions for concentrated production to reduce losses from frequent switching. During the adjustment process, the total switching cost after each adjustment is calculated in real time, and the changes before and after the adjustment are compared to ensure that each adjustment reduces or maintains the total switching cost. After multiple rounds of local and global optimization, the optimal sequence result with the best total switching cost is obtained.

[0102] Step 6.3: Based on the initial adjustment values ​​of each voxel sub-region, resource conflict detection and resolution are performed on the obtained preliminary optimization sequence to obtain resource conflict resolution results. Specifically, this includes: retrieving the calculated initial adjustment values ​​of each voxel sub-region, and combining them with the real-time status information of equipment and fixtures in the resource dataset to perform order-by-order resource occupancy analysis on the obtained preliminary optimization sequence; firstly, clarifying the core equipment, matching fixtures, and usage time periods required for each order during the production process; simulating the production timeline of each order based on the production line process flow and equipment capacity, marking the occupancy of various resources in different time periods, and based on the resource load associated in the initial adjustment values ​​of the voxel sub-regions. The system detects resource conflicts within the sequence, such as the same core equipment being occupied by two or more orders simultaneously within the same time period, the demand for a certain type of specialized fixture exceeding the currently available quantity, or the resource requirements of one order overlapping with those of other orders and being uncoordinated. For detected resource conflicts, resolution is based on the initial adjustment value of the voxel sub-region. Orders with higher initial adjustment values ​​have stronger resource adaptability and optimization priority, and their resource requirements are prioritized. For orders with lower initial adjustment values, their production order is adjusted, postponing them to a time when the conflicting resources are available, or allocating them with other suitable and available alternative resources. If no alternative resources exist, the orders are reordered to ensure that the resource requirements of all orders in the resolved sequence are met, with no conflicts in the occupation of core resources such as equipment and fixtures, thus obtaining the resource conflict resolution result.

[0103] Step 6.4: Combining the sequence optimization results and resource conflict resolution results, with maximizing the order delivery rate as the final constraint, an executable initial scheduling sequence is obtained. Specifically, this includes: First, comparing the sequence optimization results and resource conflict resolution results to analyze the differences in order order ranking and clarify the reasons for these differences. For example, some orders might be arranged adjacently in the optimized sequence to reduce switching costs, but resource conflicts exist, leading to position adjustments in the resolution results. Combining this with the core objective of maximizing the order delivery rate for the LCD module production line as the final constraint, a comprehensive balance is made between the two results. For orders adjusted due to resource conflict resolution, it is checked whether the adjusted production cycle still meets the delivery deadline. If the estimated completion time of the adjusted order is within the delivery deadline and the total switching cost has not increased, the adjusted ranking is retained. If the adjustment causes order delivery delays, the resource conflict situation needs to be reassessed, and resolution methods explored, such as adjusting the order of other lower-priority orders or optimizing equipment utilization efficiency, to ensure that the order can be completed within the delivery period. For the parts of the sequence optimization with low total switching costs but no resource conflicts, the order order remains unchanged to maintain the advantage of minimizing switching losses. During the synthesis process, the production cycle, resource usage, and switching cost of each order are checked one by one to ensure that the final sequence not only meets the optimization objective of minimizing total switching costs but also does not have resource conflicts. Furthermore, the expected completion time of all orders is within the delivery deadline, maximizing the probability of on-time order delivery. Finally, an initial sequence that can be directly used for production line production is obtained.

[0104] In this embodiment of the invention, the combined technical means of generating an initial processing sequence based on real-time comprehensive priority scoring of orders, optimizing the sequence with the goal of minimizing the total switching cost based on the process switching cost matrix, detecting and resolving resource conflicts based on the initial adjustment value of voxel sub-regions, and combining the sequence optimization results and conflict resolution results with maximizing the order delivery rate as the final constraint, overcomes the technical problems of traditional static scheduling, which only sorts by priority, does not consider the optimization of total switching cost and resource conflict, and lacks feasibility of scheduling schemes, resulting in large production line switching losses, idle resources, or delayed order delivery. Thus, it achieves the generation of an executable initial scheduling sequence that takes into account minimizing total switching cost, avoiding resource allocation conflicts, and maximizing the order delivery rate, thereby improving the scientificity and practicality of the scheduling scheme.

[0105] In a preferred embodiment of the present invention, step 7 above may include:

[0106] Step 7.1: Based on the executable initial scheduling sequence, establish a real-time monitoring mechanism to continuously monitor order change events and material supply status updates. Specifically, this includes: using the executable initial scheduling sequence as the core reference, building a comprehensive real-time monitoring mechanism framework, clarifying the core monitoring objects, data sources, collection frequency, and event identification rules. The monitoring objects are divided into two categories: the first category is order change events, covering various order-related changes that may affect scheduling execution, such as new order insertion, existing order cancellation, order delivery deadline adjustment, order production batch change, and order product specification modification; the second category is material supply status update data, including real-time inventory quantity changes of key materials, deviations between actual and planned material arrival times, material quality inspection results, adjustments to supply plans for subsequent batches of materials, and material shortage warnings, etc., which are directly related to production assurance.

[0107] In terms of data sources, a real-time data interaction channel has been established with order management, the IoT platform, warehouse management, and supplier information to ensure that various change data are obtained as soon as possible. The monitoring frequency is set to combine real-time collection and timed verification. For the order management system and supplier information, event-triggered collection is adopted. Once there is an order change or material supply status update in the system, it is immediately synchronized to the monitoring mechanism. For the IoT platform and warehouse management system, millisecond-level real-time collection is adopted to continuously obtain material inventory dynamics and supply status data. At the same time, a full data verification is performed every hour to ensure that the data transmission is complete and without deviation.

[0108] In terms of event identification rules, various judgment criteria for change events are preset. For example, if the adjustment of the order delivery deadline exceeds the preset threshold, it is judged as a valid change event. If the material arrival delay exceeds 24 hours, it is judged as a supply status update that needs to be closely monitored. The monitoring mechanism has built-in data comparison and event identification algorithms, which continuously compare the collected data with the basic data corresponding to the initial scheduling program. Once an order change event or material supply status update that meets the judgment criteria is found, it will be detected.

[0109] Step 7.2: When the monitoring mechanism identifies an order change event and a material supply status update, it triggers the scheduling recalculation process and re-executes the dynamic priority calculation of orders based on the latest resource dataset. Specifically, this includes: when the real-time monitoring mechanism identifies and marks a valid order change event or a material supply status update, it automatically triggers the scheduling recalculation process. After the process starts, it first performs a data synchronization update operation, retrieves the latest complete data from various related systems, including all updated order information, the latest production line resource status data, updated material inventory data, estimated material delivery time, the latest supplier supply commitment, etc. After integrating these data, it replaces the original resource dataset to form the latest and complete resource dataset, ensuring that the recalculation process is based on the basic data that best matches actual production.

[0110] Subsequently, based on the updated resource dataset, the complete calculation process for dynamic order priority is re-executed. First, the delivery deadline and production batch information for each order are extracted. Combined with the current remaining production line capacity, the urgency of delivery and production load coefficient are recalculated to obtain a new initial priority base value. Next, the process switching cost matrix constructed in step 4 is retrieved. Combined with the progress and status of orders currently being executed on the production line, the switching cost impact factor of each pending order relative to the current production status is recalculated. Then, based on the voxel sub-region associated with each order, the corresponding initial adjustment value is retrieved or recalculated to obtain the latest resource optimization adjustment coefficient. Finally, according to the preset weighting weights, the new initial priority base value, switching cost impact factor, and resource optimization adjustment coefficient are weighted and fused to dynamically calculate the real-time comprehensive priority of each order, forming the updated order priority ranking result.

[0111] Step 7.3: Based on the recalculated dynamic priority of orders, the scheduling optimization algorithm is invoked to dynamically adjust and update the original initial scheduling sequence, resulting in an executable scheduling sequence with the latest production status. Specifically, this includes: based on the recalculated real-time comprehensive priority ranking result of orders, invoking a consistent scheduling optimization algorithm, and simultaneously combining the latest resource dataset and the actual production progress of the current production line to dynamically adjust and update the original initial scheduling sequence; firstly, comparing the differences between the old and new priority ranking results, identifying orders with increased or decreased priority, as well as newly added pending orders, clarifying the position of each order in the original sequence and its adjusted target position. For orders with increased priority, analyze the matching between production demand and current production line resources. If current resources allow, adjust them to a higher sequence position to ensure priority production. At the same time, check the process changeover costs of the adjusted orders and the orders before and after them to avoid a significant increase in total changeover costs due to priority adjustment. For orders with decreased priority, adjust them to the corresponding sequence position according to the new priority level. If there are resource conflicts with other orders after adjustment, re-coordinate according to resource optimization principles. For new orders, find the insertion position with the lowest changeover cost and no resource conflicts in the sequence based on their comprehensive priority, and include them in the scheduling process.

[0112] During the adjustment process, resource conflict detection and resolution operations are performed simultaneously. Based on the latest equipment, fixture status, and material supply situation, the production process of the adjusted sequence is simulated, and problems such as resource occupation conflicts and production process interruptions are marked. Conflicts are resolved by adjusting the order sequence, allocating alternative resources, and optimizing production time connections. At the same time, the total switching cost of the adjusted sequence is continuously calculated to ensure that the total switching cost remains within a reasonable range under the premise of meeting priority requirements and avoiding resource conflicts. After multiple rounds of local adjustments and global optimization, a scheduling sequence adapted to the latest production status is formed. The sequence fully reflects the impact of order changes and material supply status updates, while taking into account the goals of minimizing the total switching cost, maximizing the order delivery rate, and balanced resource utilization. The final output is an executable scheduling sequence under the latest production status.

[0113] In this embodiment of the invention, by employing a combination of techniques—including establishing a real-time monitoring mechanism based on an executable initial scheduling sequence to continuously monitor order change events and material supply status updates, triggering a scheduling recalculation process upon detecting changes and recalculating the dynamic priority of orders based on the latest resource dataset, and calling a scheduling optimization algorithm to dynamically adjust and update the original scheduling sequence—the technical problems of traditional static scheduling schemes being fixed and unable to respond in real time to sudden changes such as order insertion, cancellation, delivery date adjustments, and material supply delays during the production process, leading to a disconnect between scheduling and actual production status, increased order delivery risks, or low resource utilization, are overcome. This achieves the technical effect of enabling scheduling schemes to dynamically adapt to real-time production environment changes, ensuring that the scheduling sequence always possesses feasibility and optimality, and continuously guaranteeing order delivery rate and production line operating efficiency.

[0114] In a preferred embodiment of the present invention, step 8 above may include:

[0115] Step 8.1: During the generation of the initial scheduling sequence and the dynamic adjustment of the sequence, the Material Requirements Planning (MRP) is linked in real time to obtain the latest MRP data and estimated arrival times. Specifically, this includes: establishing a real-time data interaction channel with the MRP system throughout the entire process of generating the initial scheduling sequence and dynamically adjusting the scheduling sequence, ensuring that scheduling operations and material requirement information remain synchronized. This channel continuously acquires the latest MRP data, covering the details of various materials required for each pending order and in-production order, including the name, specifications, and required quantity of key materials such as panels, backlight modules, driver ICs, and flexible circuit boards. Simultaneously, the estimated arrival times for each type of material are accurately obtained, including the supplier's promised delivery time for ordered materials, the estimated arrival time for materials in transit, and the availability time for inventory materials. Data acquisition employs a combination of real-time triggering and timed refresh. When data in the MRP is updated, it is immediately synchronized to the scheduler. Furthermore, a full refresh of all material data is performed at fixed intervals to ensure that the acquired MRP data and estimated arrival times are always the latest and most accurate.

[0116] Step 8.2: Based on the latest material requirements planning (MRP) data, determine the material availability status of each order in the scheduling sequence and identify production conflicts caused by material delays. Specifically, this includes: based on the latest MRP data and the material requirements list for each order in the order dataset, determine the material availability status of each order in the scheduling sequence one by one; first, sort out all the types of materials required for each order's production, compare the current available inventory quantity, the confirmed expected arrival quantity, and the order requirement quantity for each material. If the sum of the available inventory quantity and the expected arrival quantity for each material required for an order is greater than or equal to the order requirement quantity, and the expected arrival time of at least one material is not later than the planned start time of the order, then the material availability status of the order is determined to be satisfactory; if the inventory and expected arrival quantity of some materials cannot meet the order requirement, and the expected arrival time of all materials is later than the planned start time of the order, then it is determined to be a material shortage; if the expected arrival time of some critical materials is later than the planned start time of the order, but non-critical materials can be met, then it is determined to be partially available and there is a production risk.

[0117] After determining the availability of materials, further identify production conflicts caused by material delays. For orders with material shortages or partial availability, compare the planned start time with the expected arrival time of key materials. If the expected arrival time of key materials is later than the planned start time of the order, and the delay exceeds the buffer time reserved in the order's production cycle, the order will not be able to start as planned, and this is considered a production conflict. If the expected arrival time of key materials is later than the planned start time, but the delay is within the buffer time, it can be compensated for by compressing the time of subsequent processes, and is not considered a serious production conflict at this time, but should be marked as a key concern. If the expected arrival time of materials can meet the order's start-up requirements, it will cause production to stop midway, which is also considered a production conflict. All production conflicts caused by material delays should be thoroughly investigated and recorded.

[0118] Step 8.3: Based on the expected arrival time of materials, adaptively adjust the order of orders with material shortage risk in the scheduling sequence to obtain a final executable scheduling sequence that matches the material supply plan, ensuring the feasibility of the production plan. Specifically, this includes: for orders identified with material shortage risk and production conflicts, using the expected arrival time of various materials as the core basis, combined with the real-time comprehensive priority of the order, the process changeover cost matrix, and the load status of the production line equipment, adaptively adjust the order in the scheduling sequence. For orders with a later expected arrival time of key materials, but which can still meet the order delivery deadline after the delay, adjust the scheduling order to be arranged in the time period when the materials are expected to arrive and production conditions are available. At the same time, ensure that the process changeover cost of the orders before and after the adjustment is within a reasonable range to avoid a significant increase in total changeover losses due to the adjustment.

[0119] For orders where the expected arrival time of critical materials is delayed but the priority is extremely high and the delivery deadline cannot be postponed, resources should be coordinated to ensure their production. If there are similar materials that can be substituted and whose specifications are compatible, the material requirements plan can be adjusted and the substitute materials should be called first. If there are no substitute materials, the scheduling order should be exchanged with subsequent orders that have complete sets of materials and lower priority, so that the high-priority orders can start work immediately after the materials arrive, while ensuring that the delivery deadline of the adjusted orders is not affected.

[0120] For multiple orders with material shortages and varying expected arrival times, the production sequence of these orders is rearranged according to the order of material arrival times, prioritizing orders with materials arriving first to avoid prolonged equipment idleness. During the adjustment process, the revised schedule is continuously verified for new resource conflicts or material mismatches. If any are found, further local optimization is performed until the production plans for all orders are precisely matched with the material supply plans. The final executable schedule is then formed, eliminating the risk of material shortages, preventing production conflicts, and balancing order priority with production efficiency. This ensures the smooth execution of the production plan and avoids production line shutdowns or order delivery delays due to material issues.

[0121] In this embodiment of the invention, by employing a combination of techniques—real-time linkage with the material demand plan throughout the initial scheduling sequence generation and dynamic adjustment process to obtain the latest material demand plan data and estimated arrival time, judging the material availability status of each order based on this data and identifying production conflicts caused by material delays, and adaptively adjusting the scheduling order of orders with shortage risks according to the estimated arrival time of materials—the technical problems of material supply and production plan disconnection in traditional static scheduling, inability to predict material shortage risks in advance, production line shutdowns and order delivery delays caused by material delays are overcome. This achieves the technical effect of accurately matching the scheduling sequence with the material supply plan, avoiding production interruptions caused by material shortages, ensuring the feasibility and stability of the production plan, and further improving the on-time delivery rate of orders.

[0122] like Figure 2 As shown, embodiments of the present invention also provide a dynamic scheduling management system for multi-variety orders in a liquid crystal display module production line, including:

[0123] The acquisition module is used to collect data on the size, resolution, interface type, batch size, and delivery date of all orders. It also acquires data on production line process equipment, fixture configuration, material inventory status, and material supply plan to obtain order datasets and resource datasets.

[0124] The construction module is used to build a set of order and resource associated data points based on the order dataset and resource dataset, and define a three-dimensional virtual voxel space in the virtual structure. The set of order and resource associated data points is mapped one by one to the corresponding voxels, and the number of data points falling into the three-dimensional virtual voxel space, the corresponding order type, resource number and material requirement matching information of each voxel are recorded to obtain voxel sub-regions with data attribute labels.

[0125] The calculation module is used to calculate the adjustment value based on the characteristics of the voxel sub-region;

[0126] The quantification module is used to quantify process changeover costs based on adjustment values, order datasets, and resource datasets, and obtain a process changeover cost matrix.

[0127] The priority module is used to dynamically calculate the priority of each order based on the order dataset, the process changeover cost matrix, and adjustment values.

[0128] The optimization module is used to allocate orders to each production line process based on the order dynamic priority, process changeover cost matrix and adjustment value, with the optimization objectives of minimizing total changeover cost and maximizing order delivery rate, and to obtain the initial schedule.

[0129] The monitoring module is used to monitor order change events and material supply status in real time based on the initial scheduling sequence, recalculate the dynamic priority of orders according to the current resource dataset, and update the scheduling sequence.

[0130] The processing module is used to link material demand planning data in real time during the generation and adjustment of scheduling sequences, and adjust the scheduling order according to the material arrival time.

[0131] The above description represents the preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for dynamic scheduling management of multi-variety orders in a liquid crystal display module production line, characterized in that, The method includes: Collect all order data including size, resolution, interface type, batch size, and delivery date. Simultaneously, acquire production line process equipment, fixture configuration, material inventory status, and material supply plan data to obtain order datasets and resource datasets. Based on the order dataset and resource dataset, a set of data points associated with orders and resources is constructed, and a three-dimensional virtual voxel space is defined in the virtual structure. The set of data points associated with orders and resources is mapped one by one to the corresponding voxels, and the number of data points falling into the three-dimensional virtual voxel space, the corresponding order type, resource number and material requirement matching information of each voxel are recorded to obtain voxel sub-regions with data attribute labels. Based on the characteristics of the voxel sub-regions, adjustment values ​​are calculated, including: Based on the formed voxel sub-regions with data attribute labels, the key feature parameters of each voxel sub-region are extracted. The key feature parameters include at least the data point density, the real-time load status of the associated resources, and the material demand matching degree. Based on the extracted key feature parameters, and according to the preset resource optimization and delivery urgency rules, a comprehensive initial adjustment value is calculated for each voxel sub-region. Based on the adjustment values, order datasets, and resource datasets, the process changeover cost is quantified to obtain a process changeover cost matrix, including: Based on the initial adjustment values ​​of each element sub-region, and combined with the order attributes in the order dataset and the equipment and fixture configuration in the resource dataset, calculate the basic value of the process switching cost between any two order types. Based on the real-time status data of centralized equipment and fixtures in the resource dataset, the calculated basic value of process changeover cost is dynamically corrected to obtain the corrected process changeover cost that reflects changes in the real-time production environment. The corrected process switching costs between all pairs of order types are used to construct a complete process switching cost matrix. Based on the order dataset, process changeover cost matrix, and adjustment values, the priority of each order is dynamically calculated, including: Based on the order delivery date and batch information in the order dataset, the delivery urgency and production load coefficient of each order are calculated to obtain the initial priority base value; Based on the process changeover cost matrix and the current order status being executed on the production line, calculate the changeover cost impact factor of each pending order relative to the current production status. Based on the initial adjustment values ​​of each voxel sub-region, obtain the resource optimization adjustment coefficients of the voxel sub-regions associated with each order; The initial priority base value, switching cost impact factor and resource optimization adjustment coefficient are weighted and integrated to dynamically calculate the real-time comprehensive priority of each order; Based on the dynamic priority of orders, the process changeover cost matrix, and adjustment values, with the optimization objectives of minimizing total changeover cost and maximizing order delivery rate, orders are allocated to each production line process to obtain an initial scheduling sequence. Based on the initial scheduling column, the system monitors order change events and material supply status in real time, recalculates the dynamic priority of orders based on the current resource dataset, and updates the scheduling column. During the generation and adjustment of the scheduling sequence, the material demand planning data is linked in real time, and the scheduling order is adjusted according to the material arrival time.

2. The method for dynamic scheduling management of multi-variety orders in a liquid crystal display module production line according to claim 1, characterized in that, Collect all order data including size, resolution, interface type, batch size, and delivery date. Simultaneously, acquire production line process equipment, fixture configuration, material inventory status, and material supply plan data to obtain order datasets and resource datasets, including: Through the first data interface, the raw data of all orders to be scheduled is received synchronously; Based on the production resource requirements corresponding to the received raw data, the system collects the associated production line resource status data in real time from the IoT platform and warehouse management via the second data interface. The received raw data is integrated with the collected production line resource status data to obtain integrated data. The integrated data is then cleaned and standardized to eliminate outliers and format differences, resulting in order datasets and resource datasets.

3. The method for dynamic scheduling management of multi-variety orders in a liquid crystal display module production line according to claim 2, characterized in that, Based on the order dataset and resource dataset, a set of order and resource-related data points is constructed. A three-dimensional virtual voxel space is defined within the virtual structure. Each set of order and resource-related data points is mapped to its corresponding voxel. The number of data points falling into the three-dimensional virtual voxel space, the corresponding order type, resource number, and material requirement matching information for each voxel are recorded, resulting in voxel sub-regions with data attribute labels, including: Based on the order dataset and resource dataset, we perform association matching to match the attribute features of each order with the required process equipment, fixtures and material resources, and construct a set of order and resource association data points. Based on the core feature dimensions of orders and resources contained in the set of associated data points, a three-dimensional virtual voxel space is defined in the virtual structure; Based on the defined three-dimensional virtual voxel space, the voxelization algorithm is invoked to map and assign each data point in the constructed set of order and resource-related data points to the corresponding voxel according to the attribute coordinates of each data point. For each voxel containing data points after mapping, record the internal features of each voxel. The features include at least the number of data points falling into the voxel, the set of order types corresponding to the data points, the list of associated resource numbers, and the material demand matching degree calculated based on the material inventory status, thereby forming multiple voxel sub-regions with data attribute labels.

4. The method for dynamic scheduling management of multi-variety orders in a liquid crystal display module production line according to claim 3, characterized in that, Based on dynamic order priority, process changeover cost matrix, and adjustment values, with the optimization objectives of minimizing total changeover cost and maximizing order delivery rate, orders are allocated to each production line process to obtain an initial scheduling sequence, including: Based on the obtained real-time comprehensive priority score of the orders, an initial order processing sequence is generated as the basis for scheduling optimization. Based on the process switching cost matrix, with the goal of minimizing the total switching cost, the basic scheme is sequentially optimized and adjusted to obtain the sequence optimization result. Based on the initial adjustment values ​​of each element region, resource conflict detection and resolution are performed on the obtained preliminary optimization sequence to obtain the resource conflict resolution results. By combining the sequence optimization results and resource conflict resolution results, and taking maximizing the order delivery rate as the final constraint, an executable initial scheduling sequence is obtained.

5. The method for dynamic scheduling management of multi-variety orders in a liquid crystal display module production line according to claim 4, characterized in that, Based on the initial scheduling column, order change events and material supply status are monitored in real time. The dynamic priority of orders is recalculated according to the current resource dataset, and the scheduling column is updated accordingly, including: Based on an executable initial scheduling sequence, a real-time monitoring mechanism is established to continuously monitor updated data on order change events and material supply status. When the monitoring mechanism identifies order change events and material supply status updates, it triggers the scheduling recalculation process and re-executes the dynamic priority calculation of orders based on the latest resource dataset. Based on the recalculated dynamic priority of orders, the scheduling optimization algorithm is invoked to dynamically adjust and update the original initial scheduling sequence, resulting in the latest executable scheduling sequence for production status.

6. The method for dynamic scheduling management of multi-variety orders in a liquid crystal display module production line according to claim 5, characterized in that, During the generation and adjustment of scheduling sequences, material requirements planning data is linked in real time, and the scheduling order is adjusted according to the material arrival time, including: During the generation of the initial scheduling sequence and the dynamic adjustment of the sequence, the material requirements planning is linked in real time to obtain the latest material requirements planning data and the estimated arrival time; Based on the latest material requirements planning data, determine the material availability status of each order in the scheduling sequence and identify order production conflicts caused by material delays. Based on the expected arrival time of materials, the order of orders with material shortage risk in the scheduling sequence is adaptively adjusted to obtain a final executable scheduling sequence that matches the material supply plan, thus ensuring the feasibility of the production plan.

7. A dynamic scheduling management system for multi-variety orders in a liquid crystal display module production line, wherein the system implements the method as described in any one of claims 1 to 6, characterized in that, include: The acquisition module is used to collect data on the size, resolution, interface type, batch size, and delivery date of all orders. It also acquires data on production line process equipment, fixture configuration, material inventory status, and material supply plan to obtain order datasets and resource datasets. The construction module is used to build a set of order and resource associated data points based on the order dataset and resource dataset, and define a three-dimensional virtual voxel space in the virtual structure. The set of order and resource associated data points is mapped one by one to the corresponding voxels, and the number of data points falling into the three-dimensional virtual voxel space, the corresponding order type, resource number and material requirement matching information of each voxel are recorded to obtain voxel sub-regions with data attribute labels. The calculation module is used to calculate the adjustment value based on the characteristics of the voxel sub-region; The quantification module is used to quantify process changeover costs based on adjustment values, order datasets, and resource datasets, and obtain a process changeover cost matrix. The priority module is used to dynamically calculate the priority of each order based on the order dataset, the process changeover cost matrix, and adjustment values. The optimization module is used to allocate orders to each production line process based on the order dynamic priority, process changeover cost matrix and adjustment value, with the optimization objectives of minimizing total changeover cost and maximizing order delivery rate, and to obtain the initial schedule. The monitoring module is used to monitor order change events and material supply status in real time based on the initial scheduling sequence, recalculate the dynamic priority of orders according to the current resource dataset, and update the scheduling sequence. The processing module is used to link material demand planning data in real time during the generation and adjustment of scheduling sequences, and adjust the scheduling order according to the material arrival time.

Images