Real-time production scheduling method

Abstract

According to the real-time production scheduling method provided by the invention, the material demand is triggered based on the production plan, the MES and the WMS collaboratively complete the global scheduling of the raw materials, after the raw materials are delivered to the production line, the material batch segmentation or combination or rework path optimization is executed in real time according to the abnormal condition, and the real-time production scheduling is realized through dynamically triggering the material demand, collaboratively and globally scheduling the MES and the WMS and a real-time abnormal response mechanism. And efficient configuration of production resources and rapid closed loop of abnormal working conditions are realized, so that the flexibility and intelligence level of the manufacturing process is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of industrial Internet, and particularly relates to a real-time production scheduling method. Background Art

[0002] With the global energy structure transformation towards cleaner energy, the photovoltaic industry, as one of the core areas of renewable energy, the efficiency and intelligence level of its manufacturing technology directly affect the industry competitiveness. As a key component of photovoltaic modules, the production process of photovoltaic cells involves multiple complex processes such as silicon material preparation, texturing, diffusion, coating, printing, etc., and needs to cope with frequent process adjustments, equipment status fluctuations, and diverse order requirements. In this context, the deep collaboration between the Manufacturing Execution System (MES) and the Warehouse Management System (WMS) has become a key technical means to improve production efficiency and resource utilization.

[0003] In traditional photovoltaic cell production scheduling methods, the MES system usually generates static material requirements based on fixed production plans, lacking the ability to dynamically respond to silicon material supply, resulting in the disconnection between raw material scheduling and production rhythm. For example, when the production plan needs to be adjusted due to order changes or equipment failures, the existing technology is difficult to trigger the recalculation of silicon material requirements in a timely manner, easily causing material shortages or inventory backlogs on the production line. In addition, the information interaction between the MES and WMS systems lags behind, and raw material distribution relies on manual experience judgment, unable to accurately match the real-time needs of the production line, further exacerbating the waste of production resources.

[0004] In the processing link after the silicon material is delivered to the production line, the existing technology mostly adopts preset process parameters and batch management strategies, which are difficult to cope with real-time process fluctuations, such as abnormal equipment processing rates, temperature and humidity changes, or quality abnormalities, such as hidden cracks and excessive fragmentation rates. When an abnormality occurs, it usually relies on manual intervention for batch adjustment or rework path planning, not only with slow response speed, but also easily leading to problems such as decreased equipment utilization and increased rework costs due to decision-making deviations. For example, when a silicon wafer batch needs to be split due to equipment failure, the traditional method cannot quickly allocate it to alternative equipment, resulting in production line stagnation; and the rework path planning lacks a cost optimization model, and may choose high-energy consumption or low-yield processing methods, affecting the overall production efficiency.

[0005] In existing patents and literature, although there are technical solutions for the integration of MES and WMS, most of them focus on data interconnection or inventory visualization, and do not solve the collaborative problems of dynamic demand triggering and real-time exception response. Some solutions attempt to optimize production scheduling through algorithms, but do not perform closed-loop linkage of silicon material requirements with process fluctuations and quality abnormalities, resulting in the disconnection between scheduling strategies and actual production status. Therefore, there is an urgent need for a real-time production scheduling method that can deeply integrate production plans, raw material scheduling, and exception response to cope with the complex scenarios of multi-process coupling and high real-time requirements in photovoltaic cell manufacturing. Summary of the Invention

[0006] In view of the deficiencies of the prior art, a real-time production scheduling method is proposed. Through dynamically triggering the demand for silicon materials, the collaborative global scheduling of MES and WMS, and the real-time process fluctuation and quality anomaly response mechanism, the efficient allocation of production resources and the rapid closed-loop of abnormal working conditions are realized, thereby improving the flexibility and intelligence level of photovoltaic cell manufacturing.

[0007] On the one hand, a real-time production scheduling method is provided, including the following steps:

[0008] Trigger the demand for silicon materials based on the production plan to generate dynamic raw material scheduling instructions;

[0009] The MES system and the WMS system cooperate to execute the global scheduling of raw materials and deliver the silicon materials to the designated production line;

[0010] After the silicon materials are delivered to the production line, the process fluctuation or quality anomaly data in the production process is monitored in real time;

[0011] Dynamically adjust the splitting and merging operations of the silicon wafer batches according to the monitoring results, or optimize the rework path to realize the real-time reallocation of production resources.

[0012] Preferably, the triggering of the demand for silicon materials based on the production plan includes: decomposing the production plan into process-level tasks, and matching the silicon material type and usage according to the bill of materials (BOM); combining the production line equipment status and historical yield data to generate a dynamic silicon material demand priority queue.

[0013] Preferably, the collaborative execution of the global scheduling of raw materials by MES and WMS includes:

[0014] MES sends a silicon material demand instruction to WMS, including the material code, demand quantity, and delivery time window; WMS generates a delivery plan according to the inventory status and the availability of logistics equipment, and feeds back the silicon material batch number and real-time location to MES; MES dynamically adjusts the production line feeding plan according to the feedback information and triggers a stock-out warning mechanism.

[0015] Preferably, after the silicon materials are delivered to the production line, it further includes: binding the silicon material batch information with the production work order and updating the batch status in MES in real time; collecting the process parameters in the silicon wafer processing process through sensors and comparing the deviations with the preset process specifications.

[0016] Preferably, the adjustment of the splitting or merging operation of the silicon wafer batches according to the process fluctuation includes: when it is detected that the deviation of the equipment processing rate exceeds the threshold, triggering a batch dynamic splitting instruction to split the original batch into multiple sub-batches and allocate them to parallel equipment; when it is detected that the equipment idle rate is higher than the preset value, triggering a batch merging instruction to merge adjacent batches to improve the equipment utilization rate.

[0017] Preferably, the optimized rework path includes: matching the rework process rule library according to the quality exception type to generate a set of optional rework paths; calculating the time consumption and resource consumption of each rework path based on the path cost model, and selecting the path with the lowest comprehensive cost and updating the production work order.

[0018] Preferably, the method further includes: real-time processing of process fluctuation data by an edge computing device, and dynamically adjusting the process parameter setting values in the MES; if the adjusted parameters still exceed the tolerance range, triggering an equipment maintenance work order and reassigning the in-process product batches.

[0019] Preferably, the calculation of the rework path cost model includes: using the equipment switching time, the energy consumption of the rework process, and the yield improvement rate after rework as weight factors; traversing the rework path topology graph using the Dijkstra algorithm to output the shortest weighted path.

[0020] Preferably, the triggering of the silicon material demand based on the production plan includes: the MES dynamically adjusts the silicon material demand through the formula where Q 理论 is the standard usage in the BOM, and the yield loss rate is dynamically corrected according to the equipment historical data.

[0021] Preferably, the MES dynamically adjusts the production line feeding plan according to the feedback information and triggers the stock shortage warning mechanism, including: when the silicon material ETA exceeds the lower limit of the delivery time window, triggering a yellow warning and notifying the production planner to intervene manually; when the silicon material does not arrive before the start of the process, triggering a red warning, automatically pausing the relevant work orders, and triggering the call for safety stock.

[0022] On the other hand, a real-time production scheduling system is provided, which includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements a real-time production scheduling method as described above.

[0023] A real-time production scheduling method provided by the present invention triggers the silicon material demand based on the production plan, and the MES and WMS cooperate to complete the global scheduling of raw materials. After the raw materials are delivered to the production line, according to the process fluctuations or quality anomalies, the silicon wafer batch splitting / merging or rework path optimization is executed in real time. Through the dynamic triggering of silicon material demand, the cooperation between the MES and WMS for global scheduling, and the real-time process fluctuation and quality anomaly response mechanism, the efficient allocation of production resources and the rapid closed-loop of abnormal working conditions are realized, thereby improving the flexibility and intelligence level of photovoltaic cell manufacturing. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 It is a production scheduling system architecture of a photovoltaic cell MES system provided by the present invention;

[0025] Figure 2 A production scheduling process of a photovoltaic cell MES system provided by the present invention;

[0026] Figure 3 A schematic diagram of a production scheduling system of a photovoltaic cell MES system provided by the present invention. Specific embodiments

[0027] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0028] It should be noted that the terms used here are only for describing specific embodiments, rather than intending to limit the exemplary embodiments according to the present application. As used here, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should be understood that when the terms "comprising" and / or "including" are used in this specification, they indicate the presence of features, steps, operations, devices, components, and / or combinations thereof.

[0029] Embodiment 1

[0030] Figure 1 It is a production scheduling system architecture of a photovoltaic cell MES system, mainly including modules such as production reports, process reports, and quality reports. The ERP module includes plan and work order management, front and middle section production execution, and back section production execution. Plan and work order management includes functions such as work order acceptance, work order material requisition, work order material return, work order daily plan query, and work order summary. Front and middle section production execution involves operations such as work order material requisition replacement, slurry material requisition replacement, and feeding in the unpacking room. Back section production execution includes small box code generation, large box code generation, label reprinting, OBA inspection, and OBA rework operations.

[0031] WMS includes warehouse management, product BOM and process modeling, factory calendar and system configuration. Warehouse management covers processes such as wafer guiding, wafer making, diffusion, chlorination, BSG, PEPOLY, expensive PSG, RCA, and test sorting. Product BOM and process modeling are used to manage the bill of materials and process information of products. Factory calendar and system configuration provide factory calendar and system configuration management functions. MES includes process management, quality management and traceability. Process management includes process registration, process exception handling, and process completion statistics. Quality management and traceability cover first piece inspection, process control, quality exception handling, etc.

[0032] The Internet of Things system realizes the real-time collection and monitoring of production data through devices such as AGV, barcodes, RFID, data acquisition and interaction terminals, PLC, sensors, and visualization terminals. The industrial bus protocol supports protocols such as OPC and Modbus to ensure data communication between devices and the system.

[0033] The system also includes report generation, regularly generating various production, process, and quality reports to facilitate decision-making analysis by management. System integration and interface management: Ensure data circulation and interface management between subsystems to achieve seamless operation of the overall system. By integrating multiple management systems such as ERP, MES, and WMS, combining Internet of Things technology and industrial bus protocol, this system realizes intelligent and visual management of the fully automated production line, improving production efficiency and product quality.

[0034] As Figure 2 shown, this embodiment provides a real-time production scheduling process of a photovoltaic cell MES system, which includes: Step S1, triggering the silicon material demand based on the production plan and generating a dynamic raw material scheduling instruction; Step S2, the MES system and the WMS system cooperate to execute the global raw material scheduling and deliver the silicon material to the designated production line; Step S3, after the silicon material is delivered to the production line, real-time monitor the process fluctuations or quality abnormal data during the production process; Step S4, dynamically adjust the splitting and merging operations of the silicon wafer batches according to the monitoring results, or optimize the rework path to achieve real-time reallocation of production resources.

[0035] In step S1, the triggering of silicon material requirements based on the production plan includes: decomposing the production plan into process-level tasks, and matching the silicon material type and quantity according to the bill of materials (BOM); combining the production line equipment status and historical yield data to generate a dynamic priority queue for silicon material requirements. Specifically, the MES disassembles the production order into process-level tasks, such as texturing, diffusion, coating, etc., and associates the silicon material type, unit quantity, and process time constraints required for each process based on the BOM. The silicon material types include single-crystalline P-type / N-type and silicon wafer thickness. Taking the diffusion process as an example, the diffusion process needs to match silicon materials with specific doping concentrations. The MES dynamically calls the corresponding data in the BOM according to the process parameters to achieve accurate mapping of silicon material requirements, and real-time corrects the silicon material type and quantity in the BOM by detecting the silicon wafer thickness and doping concentration to support adaptive adjustment during process changes. The generation of the dynamic priority queue for silicon material requirements by combining the production line equipment status and historical yield data specifically includes fusing multi-dimensional parameters such as the real-time overall equipment effectiveness (OEE) of the equipment, historical yield, and order urgency to generate a dynamic queue for silicon material requirements, ensuring that high-value silicon materials are preferentially allocated to high-yield equipment. The historical yield includes the fragmentation rate and conversion efficiency. In this process, dynamic demand calculation is required, combining the real-time equipment status and historical yield data. The equipment status includes the equipment OEE and the current load rate, and the historical yield data includes the fragmentation rate and the hidden crack rate. The MES dynamically adjusts the silicon material demand through the following formula: where Q 理论 is the standard quantity in the BOM, and the yield loss rate is dynamically corrected according to the historical data of the equipment.

[0036] In step S2, the collaborative execution of the global raw material scheduling by the MES and the WMS includes: the MES sends a silicon material requirement instruction to the WMS, and the instruction includes the material code, the required quantity, and the delivery time window; the WMS generates a delivery plan based on the inventory status and the availability of the logistics equipment, and feeds back the silicon material batch number and the real-time location to the MES; the MES dynamically adjusts the production line feeding plan according to the feedback information and triggers a stock-out warning mechanism.

[0037] Before the MES sends a silicon material requirement instruction to the WMS, instruction encapsulation is required. After the instruction is encapsulated, the instruction is sent. The MES encapsulates the requirement instruction into a structured message in JSON / XML format. The structured message includes the material code, the required quantity, the delivery time window, and the priority label. The material code is the unique identifier of the silicon material that conforms to the international standard SEMI. The required quantity is the actual required quantity based on dynamic calculation. The delivery time window is the delivery time interval inversely deduced according to the process scheduling time. The priority label is the priority level comprehensively generated by the order urgency and the equipment yield weight. The MES system pushes the instruction to the WMS system through the RESTful API or the message queue.

[0038] The WMS generates a distribution plan based on the inventory status and the availability of logistics equipment, and feeds back the silicon material batch number and real-time location to the MES, including: After receiving the demand instruction, the WMS retrieves the real-time inventory data (including available inventory, locked inventory, and in-transit inventory) based on the material code, and matches the demand through the inventory optimization algorithm, preferentially selecting the batch with the nearest expiration date and matching quality grade. If the inventory is insufficient, it triggers an automatic replenishment request or returns a shortage warning to the MES. The WMS combines the real-time status of logistics equipment such as automatic guided vehicle (AGV) and forklift, including their positions, battery levels, and task queues, and uses a path planning algorithm to generate a distribution plan. The WMS returns the silicon material batch number, real-time location, and estimated arrival time to the MES. The real-time location is determined by tracking the silicon material between the warehouse and the production line through RFID or UWB positioning technology, and the estimated arrival time (ETA) is dynamically updated based on the speed and path length of the logistics equipment.

[0039] The MES dynamically adjusts the production line feeding plan according to the feedback information and triggers a shortage warning mechanism. Based on the silicon material batch information (such as batch number, ETA) fed back by the WMS, the MES executes dynamic adjustment of the feeding plan, associates the silicon material batch with the production work order, and updates the status of the MES work order to "pending feeding". If the silicon material delivery is delayed, the MES reallocates the equipment tasks through a scheduling algorithm. For example, the task originally planned to be fed into equipment A is dynamically switched to equipment B, and the process parameter settings of equipment B are adjusted. The MES monitors the following conditions in real time and triggers a hierarchical warning. When the ETA of the silicon material exceeds the lower limit of the delivery time window, a yellow warning is triggered to notify the production planner to intervene manually. When the silicon material does not arrive before the start of the process, a red warning is triggered to automatically suspend the relevant work order and trigger the call of safety inventory. The MES pushes the adjusted feeding plan to the production line equipment controller and synchronizes the change information to the WMS at the same time to ensure that the subsequent distribution tasks match the latest production rhythm.

[0040] After the silicon material is delivered to the production line, it further includes: binding the silicon material batch information with the production work order and updating the batch status in real time in the MES; collecting the process parameters during the silicon wafer processing through sensors and comparing the deviations with the preset process specifications. Specifically, RFID tags or two-dimensional codes are deployed on the silicon material packaging or carriers to store information such as batch numbers, material codes, production dates, and quality grades. When the silicon material is transported to the production line by AGV or manually, the batch information is automatically identified by a fixed RFID reader or a barcode scanner and pushed to the MES system. The MES binds the silicon material batch with the corresponding production work order according to the current production line task queue. The production work order includes product models, process versions, and order numbers. The binding trigger mechanism includes automatic triggering or manual intervention. If the batch information is completely matched, the MES automatically completes the binding and generates a feeding instruction. If there are parameter deviations, such as the silicon wafer thickness exceeding the tolerance, the MES triggers an exception work order, and a process engineer needs to confirm and manually bind it. The MES establishes a "work order - batch" association table in the relational database MySQL to record batch numbers, binding times, operators, and work order statuses (such as "bound", "fed"), and stores the real-time status change logs through the time series database InfluxDB. Internet of Things devices are deployed at key nodes on the production line to update the batch status in real time: the status includes the feeding point, in processing, and completed / abnormal. The feeding point means that the RFID reader confirms that the silicon material enters the equipment feed port, and the status is updated to "feeding". In processing means that the programmable logic controller PLC sends a heartbeat signal to the MES, and the status is marked as "in processing". Completed / abnormal means that after the processing is completed, the equipment feeds back a completion signal, or when an abnormality is detected, the status of "awaiting quality inspection" or "rework" is triggered. The MES provides a visual interface. The batch map shows the current location of the silicon material batch based on the production line layout diagram. The current location is located in diffusion furnaces, PECVD equipment, etc. The status color blocks mark the batch status with different colors. For example, green indicates normal processing, and red indicates abnormal stagnation. When the batch residence time exceeds the threshold, a warning prompt and handling suggestions are automatically popped up.

[0041] Collect process parameters by deploying sensors, including physical parameters, environmental parameters, and optical parameters. Physical parameters include temperature sensors (diffusion furnace temperature), pressure sensors (vacuum chamber), and thickness gauges (wafer coating thickness). Optical parameters include CCD vision inspection equipment (hidden cracks, scratches), and spectrometers (doping concentration). Environmental parameters include humidity sensors (cleanroom humidity) and particle counters (dust concentration). Sensor data is filtered, denoised, and normalized through an edge gateway (such as Huawei Atlas500) to reduce network transmission load. Use the MQTT or OPCUA protocol to upload data to the MES. Preset a process specification library in the MES, which contains the standard values and tolerance ranges of process parameters for each process (such as diffusion temperature: 850°C ± 5°C), and directly determine whether the parameters exceed the upper and lower limits. For example, if the temperature > 855°C, an alarm is triggered. If a key parameter (such as the hidden crack rate) exceeds the standard, the MES immediately suspends processing and locks the current batch.

[0042] The operation of splitting or merging wafer batches according to process fluctuations includes: when it is detected that the processing rate deviation of the equipment exceeds the threshold, a batch dynamic splitting instruction is triggered, and the original batch is split into multiple sub-batches and assigned to parallel equipment; when it is detected that the equipment idle rate is higher than the preset value, a batch merging instruction is triggered to merge adjacent batches to improve equipment utilization.

[0043] Specifically, obtain the processing rate in real time through the device PLC or SCADA system, such as the number of wafers processed per hour, or deploy a laser speedometer or encoder to directly measure the wafer transfer speed. The system calculates the real-time rate deviation rate according to the preset standard rate V standard. When δ≥δ 阈值 , such as 15%, a splitting instruction is triggered. The device sends status signals every 5 seconds, and the status includes running, idle, and faulty. Statistically calculate the proportion of the equipment idle time in 1 hour. When η 空闲 ≥η 阈值 , a merging instruction is triggered, and the threshold can be set to 20%.

[0044] When the processing rate deviation exceeds the limit, or the equipment failure causes the original rate to be unable to be restored, a batch dynamic splitting instruction is triggered. The number of sub-batches is dynamically calculated according to the deviation degree, and the sub-batch capacity is evenly distributed according to the equipment processing capacity. Select equipment that meets process compatibility, status, and parameter matching from the equipment pool. Process compatibility means supporting the current process (such as coating, diffusion), being in an idle or low-load state η 空闲 ≥10%, and the process parameters such as temperature and pressure can cover the requirements of the original batch. The MES pushes the sub-batch processing tasks to the target device controller, and synchronously updates the process parameters and traceability identification information. At the same time, the WMS schedules the AGV to transport the split wafer carriers to the target device feed port to ensure seamless connection.

[0045] When the equipment idle rate exceeds the limit, or there are multiple small batches resulting in frequent equipment changeovers, a batch dynamic splitting instruction is triggered. Adjacent batches with the same process, the same silicon material type (such as N-type monocrystalline silicon), and consistent process versions are selected. After merging, the total number of silicon wafers does not exceed the maximum capacity of the equipment. If there are differences in the merged batch parameters (such as coating thickness ±2nm), the mean or maximum value within the tolerance range is taken. If the parameter difference exceeds the tolerance (such as thickness difference > 5nm), an artificial review process is triggered. MES binds the merged batch to a new work order and inherits the original batch traceability relationship (such as "merged batch B1 + B2"), assigns it to the equipment with the highest idle rate, and pre-loads the merged process parameter configuration file. The WMS schedules the robotic arm to integrate the silicon wafers in multiple carriers into a unified carrier, reducing the number of AGV transports.

[0046] The optimized rework path includes: matching the rework process rule library according to the quality exception type to generate a set of optional rework paths; calculating the time consumption and resource consumption of each rework path in the path cost model, and selecting the path with the lowest comprehensive cost and updating the production work order.

[0047] The calculation of the rework path cost model includes: using the equipment switching time, the energy consumption of the rework process, and the yield improvement rate after rework as weight factors, for each rework path P i , calculate the comprehensive cost C i , C i = αT i + βR i + γ(1 - Y i ), where T i represents the total time consumption, R i is the resource consumption, Y i is the predicted yield value after rework, and α, β, γ are weight values; use the Dijkstra algorithm to traverse the rework path topology graph and output the shortest weighted path.

[0048] The method also includes: real-time processing of process fluctuation data through an edge computing device and dynamically adjusting the process parameter setting values in MES; if the adjusted parameters still exceed the tolerance range, trigger an equipment maintenance work order and reassign the in-process batch. Deploy an edge computing gateway on the production line equipment side to support access to the device PLC through Modbus / TCP and OPCUA protocols. The edge node has a built-in rule engine that stores process expert experience rules. The process expert experience rules are as follows:

[0049] if temperature > set value + 5°C:

[0050] Adjustment plan = {"heating power": current value - 10%, "gas flow": current value + 5%}

[0051] elif film thickness standard deviation > 3nm:

[0052] Adjustment plan = {"Deposition rate": current value × 0.9}.

[0053] A real-time production scheduling method provided by the present invention triggers the demand for polysilicon materials based on the production plan. MES and WMS cooperate to complete the global scheduling of raw materials. After the raw materials have been delivered to the production line, according to process fluctuations or quality anomalies, real-time wafer batch splitting / merging or rework path optimization is performed. Through the dynamic triggering of silicon material requirements, the collaborative global scheduling of MES and WMS, and the real-time process fluctuation and quality anomaly response mechanism, the efficient allocation of production resources and the rapid closed-loop of abnormal working conditions are realized, thereby improving the flexibility and intelligence level of photovoltaic cell manufacturing.

[0054] Embodiment 2

[0055] In summary, a real-time production scheduling system provided by the present invention includes a processor 301, a memory 302, a communication module 303, and a computer program stored on the memory. When the processor 301 executes the computer program, the real-time production scheduling method described above is implemented.

[0056] Those skilled in the art should understand that the embodiments of the present application can be provided as methods, systems, or computer program products. Therefore, the present application can adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0057] Those skilled in the art should understand that the embodiments of the present application can be provided as methods, systems, or computer program products. Therefore, the present application can adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0058] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner described in the specification, such that the instructions stored in the computer-readable memory produce a manufactured article including an instruction device that implements the functions specified in one process Figure 1 one process or multiple processes and / or blocks Figure 1 in one block or multiple blocks.

[0059] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to operate in a particular manner, such that the instructions stored in the computer-readable memory produce a manufacture including an instruction means that implements the functions specified in one or more of the processes Figure 1 or blocks Figure 1 specified in one or more of the blocks.

[0060] It should be noted that the technical features in the above embodiments can be combined arbitrarily, and the combined technical solutions all fall within the protection scope of this application. And in this text, terms such as "including", "comprising" or any other variant thereof are intended to cover non-exclusive inclusion, so that a process, method, article or apparatus including a series of elements not only includes those elements, but also includes other elements not expressly listed, or further includes elements inherent to such process, method, article or apparatus. Without further limitation, an element defined by the statement "including one..." does not exclude the existence of another identical element in the process, method, article or apparatus including the said element.

[0061] Although the embodiments of the present invention have been shown and described, it will be understood by those of ordinary skill in the art that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principles and spirit of the present invention, and the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A real-time production scheduling method, characterized in that, It includes the following steps: Trigger the silicon material demand based on the production plan and generate a dynamic raw material scheduling instruction; The MES system and the WMS system cooperate to execute the global raw material scheduling and deliver the silicon material to the designated production line; After the silicon material is delivered to the production line, real-time monitor the process fluctuations or quality abnormal data during the production process; Dynamically adjust the splitting, merging operations of the silicon wafer batches, or optimize the rework path according to the monitoring results to achieve real-time reallocation of production resources.

2. The method according to claim 1, characterized in that, The triggering of the silicon material demand based on the production plan includes: Decompose the production plan into process-level tasks and match the silicon material type and quantity according to the bill of materials; Combine the production line equipment status and historical yield data to generate a dynamic silicon material demand priority queue.

3. The method according to claim 1, characterized in that, The cooperation between the MES and the WMS to execute the global raw material scheduling includes: The MES sends a silicon material demand instruction to the WMS, including the material code, the required quantity, and the delivery time window; The WMS generates a delivery plan according to the inventory status and the availability of logistics equipment, and feeds back the silicon material batch number and the real-time location to the MES; The MES dynamically adjusts the production line feeding plan according to the feedback information and triggers a stock-out warning mechanism.

4. The method according to claim 1, wherein After the silicon material is delivered to the production line, it further includes: Bind the silicon material batch information to the production work order and update the batch status in real time in the MES; Collect the process parameters during the silicon wafer processing through sensors and compare the deviations with the preset process specifications.

5. The method according to claim 1, wherein The adjustment of the splitting or merging of the silicon wafer batches according to the process fluctuations includes: When it is detected that the deviation of the equipment processing rate exceeds the threshold, trigger a batch dynamic splitting instruction, split the original batch into multiple sub-batches and allocate them to parallel equipment; When it is detected that the equipment idle rate is higher than the preset value, trigger a batch merging instruction to merge adjacent batches to improve the equipment utilization rate.

6. The method according to claim 1, characterized in that, The optimization of the rework path includes: matching the rework process rule library according to the quality abnormal type to generate a set of optional rework paths; Calculate the time consumption and resource consumption of each rework path based on the path cost model, select the path with the lowest comprehensive cost and update the production work order.

7. The method according to claim 1, wherein The method further includes: Real-time process the process fluctuation data through the edge computing device and dynamically adjust the process parameter setting value in the MES; If the adjusted parameter still exceeds the tolerance range, trigger an equipment maintenance work order and reallocate the in-process batch.

8. The method according to claim 6, characterized in that, The calculation of the rework path cost model includes: Use the equipment switching time, the energy consumption of the rework process, and the yield improvement rate after rework as weight factors; Adopt the Dijkstra algorithm to traverse the rework path topology graph and output the shortest weighted path.

9. The method according to claim 1 or 2, characterized in that, The silicon material demand triggered based on the production plan includes: MES dynamically adjusts the silicon material demand through the formula where Q 理论 is the standard usage in the BOM, and the yield loss rate is dynamically corrected according to the historical data of the equipment.

10. The method according to claim 3, wherein The MES dynamically adjusts the production line feeding plan according to the feedback information and triggers a stock-out warning mechanism includes: When the ETA of the silicon material exceeds the lower limit of the delivery time window, trigger a yellow warning and notify the production planner to intervene manually. When the silicon material does not arrive before the start of the process, trigger a red warning, automatically suspend the relevant work order, and trigger the call for safety inventory.

Images