FCBGA carrier plate production scheduling optimization method, system, equipment and medium

Abstract

The invention provides an FCBGA carrier plate production scheduling optimization method, system and device and a medium, and belongs to the technical field of semiconductor manufacturing. According to the method, technological parameter data of key procedures such as laser drilling, electroplating and laminating are collected in real time, and a quality risk digital twin dynamically evolved along with the production progress is constructed; calculating the quality risk entropy of each production batch based on the quality feature data in the digital twin; calculating a dynamic priority score in combination with the quality risk entropy, the delivery time data and the equipment state data; executing a shunting scheduling decision according to the priority score and the quality risk entropy, and generating a differentiated scheduling instruction; and finally, the manufacturing execution system drives the production equipment to execute corresponding operation. According to the method, accurate prediction and active prevention and control of the quality risk in the production process of the FCBGA carrier plate are realized, the process loss of high-value materials is effectively reduced, and the equipment utilization rate and the production yield are improved.

Description

Technical Field

[0001] This invention belongs to the field of semiconductor manufacturing technology, and more specifically relates to an FCBGA carrier board production scheduling optimization method, system, equipment and medium. Background Technology

[0002] FCBGA (Flip Chip Ball Grid Array) substrates are core components of high-end chip packaging. Their production process involves dozens of precision steps, including laser drilling, micro-line imaging, electroplating, and lamination, resulting in significant characteristics such as complex processes, long production cycles, and high manufacturing costs. In existing technologies, FCBGA substrate production scheduling primarily relies on traditional Manufacturing Execution Systems (MES), which typically schedule production based on fixed rules such as first-come, first-served, shortest processing time, or delivery urgency. However, such methods have significant limitations: they mainly focus on equipment utilization and order delivery timeliness, completely ignoring the impact of constantly changing quality conditions during production on scheduling decisions, leading to an inability to effectively prevent and control quality risks beforehand.

[0003] A more prominent problem is that quality inspection in existing production systems is often independent of the scheduling system, employing a passive "post-production interception" model. This means that quality assessment of work-in-process is only conducted after key processes are completed, using methods such as automated optical inspection (AOI) or X-ray inspection. Under this model, once a quality defect is detected, the batch of products has often already been transferred to multiple subsequent processes, resulting not only in costly waste of materials and production capacity but also severely disrupting the overall production plan, leading to delivery delays and increased production costs. For high-value products like FCBGA substrates, the cost of this quality control model is particularly high.

[0004] Furthermore, existing scheduling systems and quality management systems generally suffer from "information silos." Scheduling decisions lack the ability to perceive and predict potential quality risks implied by fluctuations in process parameters of upstream processes. For example, minute drifts in key process parameters such as the power stability of laser drilling and the uniformity of electroplating current density pose potential threats to the yield of subsequent processes, but traditional scheduling systems completely ignore such signals. This disconnect between scheduling and quality prevents the production process from achieving dynamic optimization and resource allocation based on real-time quality risks, severely restricting further improvements in production yield and efficiency. Summary of the Invention

[0005] To address the above problems, the present invention aims to provide an FCBGA carrier production scheduling optimization method, system, equipment, and medium. By constructing a digital twin of quality risk to quantify process characteristics in real time, and driving dynamic scheduling decisions based on quality risk entropy, the invention achieves collaborative optimization of proactive prediction of quality risk and adaptive allocation of production resources in FCBGA carrier production. Ultimately, it significantly improves production yield and overall equipment utilization while reducing the loss of high-value materials in the process.

[0006] To achieve the above objectives, the present invention employs the following technical solution: In a first aspect, embodiments of this application provide an FCBGA carrier board production scheduling optimization method, including: Real-time acquisition of process parameter data for key processes during FCBGA substrate production; Based on the real-time collected process parameter data, a quality risk digital twin is created for each production batch, which dynamically evolves with the production progress. Quality characteristic data for quality risk prediction is generated based on the process parameter data and stored in the quality risk digital twin. The quality feature data stored in the quality risk digital twin is used to calculate the quality risk entropy of the corresponding production batch. Based on the quality risk entropy, and combined with delivery date data and equipment status data, a dynamic priority score for each production batch is calculated. Based on dynamic priority scores and quality risk entropy, traffic diversion scheduling decisions are made, and scheduling instruction data is generated. The scheduling instruction data is sent to the manufacturing execution system, which then parses the scheduling instruction data into control commands that can be executed by the production equipment and drives the corresponding production equipment to perform physical operations.

[0007] In one optional implementation, the real-time acquisition of process parameter data for key processes during FCBGA carrier board production includes: Real-time laser power fluctuation data, which characterizes drilling quality, is collected by photoelectric sensors deployed in the laser drilling process. By deploying Hall effect sensors in the electroplating process, time-series data of electroplating current density characterizing the uniformity of the plating layer are collected. Infrared temperature sensors and piezoelectric pressure sensors deployed in the lamination process are used to collect lamination temperature-pressure time-series data characterizing the lamination quality. Based on the laser power fluctuation data, electroplating current density time series data, and lamination temperature-pressure time series data, process parameter data are generated.

[0008] In one optional implementation, based on the real-time collected process parameter data, a quality risk digital twin is created for each production batch, dynamically evolving with the production progress. Quality characteristic data for quality risk prediction is generated based on the process parameter data and stored in the quality risk digital twin, including: The process parameter data collected in real time are associated and stored with the corresponding batch unique identifier according to the production time sequence, and an initial framework for the corresponding quality risk digital twin is created. Based on the laser power fluctuation data, a laser power stability coefficient is generated by calculating its standard deviation. Based on the electroplating current density time series data, an electroplating uniformity index is generated by calculating the difference between its maximum and minimum values ​​within a set time window. Based on the lamination temperature-pressure time series data, it is compared with the pre-stored standard lamination temperature-pressure time series data of this model of FCBGA carrier board. The similarity between the two time series data curves is calculated by the dynamic time warping algorithm to generate the lamination curve fit. The process parameter data, along with the calculated laser power stability coefficient, electroplating uniformity index, and lamination curve fit, are used as quality feature data. These data are then stored in the initial framework of the quality risk digital twin according to the process sequence and the acquisition timestamp, thus constructing a quality risk digital twin containing a complete process history.

[0009] In an optional implementation, the step of calling the quality feature data stored in the quality risk digital twin to calculate the quality risk entropy of the corresponding production batch includes: The quality feature data is extracted from the quality risk digital twin to form a time-series feature vector; The temporal feature vector is input into a pre-trained quality risk prediction model, which is a machine learning model based on a gradient boosting decision tree architecture. Its input layer dimension matches the dimension of the temporal feature vector, and its output layer outputs a quality failure probability F between 0 and 1 through the Sigmoid activation function. Substituting the quality failure probability F into the following formula, the quality risk entropy Q of the corresponding batch is generated:

[0010] Where C is a cost-weighted coefficient that is positively correlated with the value of the FCBGA carrier board and the costs already invested in this production batch.

[0011] In an optional implementation, the step of calculating the dynamic priority score for each production batch based on the quality risk entropy, combined with delivery date data and equipment status data, includes: Obtain the delivery date and the current time of the FCBGA carrier board for this production batch from the production planning system. By calculating the difference in working days between the delivery date and the current date, and subtracting the standard working hours required to complete the remaining processes, the delivery date slack ST in working days is obtained. For the next pending process of the FCBGA carrier board of this production batch, obtain the status and historical performance data of all available devices corresponding to this process type from the equipment management system. By comprehensively calculating the real-time available status score and the historical overall equipment efficiency value of each device, and performing weighted fusion according to the preset weights, select the optimal value as the equipment efficiency factor Ee. Based on the quality risk entropy Q, the delivery date slack ST, and the equipment efficiency factor Ee, calculate the dynamic priority score S through the following formula:

[0012] Where, 、 、 Are preset weight coefficients.

[0013] In an optional embodiment, the shunt scheduling decision is executed based on the dynamic priority score and the quality risk entropy, and scheduling instruction data is generated, including: Compare the quality risk entropy Q with the preset high-risk threshold Th and low-risk threshold Tl respectively; If Q > Th, generate the first scheduling instruction data for guiding the FCBGA carrier board of this production batch to the inspection station with an online three-dimensional measurement function; If Tl < Q ≤ Th, generate the second scheduling instruction data for preferentially allocating the subsequent processing tasks of the FCBGA carrier board of this production batch to the equipment group ranked in the top 10% in terms of historical yield; If Q ≤ Tl, generate the third scheduling instruction data for planning and locking the material handling path and equipment sequence with the shortest processing cycle for the FCBGA carrier board of this production batch.

[0014] In an optional embodiment, the scheduling instruction data is sent to the manufacturing execution system, and the manufacturing execution system analyzes the scheduling instruction data into control commands executable by the production equipment and drives the corresponding production equipment to perform physical operations, including: Standardize and encapsulate the scheduling instruction data to form a standardized instruction data packet including the target equipment number, processing process code, and execution time window; Distribute the standardized instruction data packet to the corresponding workshop operation terminal and material handling system through the manufacturing execution system interface; The system receives real-time feedback data on the execution status of instructions from the manufacturing execution system and verifies whether the executed actions are consistent with the requirements of the scheduling instruction data.

[0015] Secondly, embodiments of this application also provide an FCBGA carrier board production scheduling optimization system, including: The parameter acquisition module is used to collect process parameter data of key processes during the production of FCBGA carrier boards in real time. The twin construction module is used to create a quality risk digital twin that dynamically evolves with the production progress for each production batch based on the real-time collected process parameter data, and to generate quality feature data for quality risk prediction based on the process parameter data, which is stored in the quality risk digital twin. The risk quantification module is used to call the quality feature data stored in the quality risk digital twin to calculate the quality risk entropy of the corresponding production batch. The priority calculation module is used to calculate the dynamic priority score of each production batch based on the quality risk entropy, combined with delivery date data and equipment status data. The scheduling decision module is used to execute traffic allocation scheduling decisions and generate scheduling instruction data based on dynamic priority scores and quality risk entropy. The instruction execution module is used to send scheduling instruction data to the manufacturing execution system, which then parses the scheduling instruction data into control commands that can be executed by the production equipment and drives the corresponding production equipment to perform physical operations.

[0016] Thirdly, embodiments of this application also provide an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the steps of the FCBGA carrier board production scheduling optimization method as described in any of the above.

[0017] Fourthly, embodiments of this application also provide a storage medium storing a computer program thereon, which, when executed by a processor, implements the steps of the FCBGA carrier board production scheduling optimization method as described in any of the above claims.

[0018] As can be seen from the above technical solutions, the present invention has the following advantages: The FCBGA carrier production scheduling optimization method provided in this application addresses the high-value, complex, and multi-stage production characteristics of FCBGA carriers. By constructing a digital twin of quality risk that integrates multi-source process parameters such as laser drilling, electroplating, and lamination, it achieves real-time perception and risk quantification of micron-level process deviations. Based on a collaborative decision-making mechanism of quality risk entropy and dynamic priority, it establishes early intervention paths for high-risk batches, process assurance paths for medium-risk batches, and efficiency-priority paths for low-risk batches. This forms an adaptive scheduling capability strongly correlated with the production quality of FCBGA carriers, ultimately reducing the process scrap rate of high-end carriers while achieving overall optimization of quality control costs and production timeliness.

[0019] This application achieves real-time quality monitoring of key processes in FCBGA substrate manufacturing by constructing a digital twin that integrates laser drilling power fluctuations, electroplating current density curves, and lamination temperature-pressure curves. Based on a dynamic time warping algorithm and a gradient boosting decision tree model for quality risk prediction, it can accurately identify potential defects caused by micron-level process deviations, thus elevating quality control precision to the process level.

[0020] This application addresses the high-cost material characteristics of FCBGA substrates by employing an early sorting mechanism driven by quality risk entropy to isolate high-risk batches before a cascading quality loss occurs. This predictive, proactive intervention strategy effectively avoids ineffective processing of precious metal materials and precision substrates, significantly reducing material scrap rates during production.

[0021] This application implements a differentiated equipment scheduling strategy based on quality risk levels. High-risk batches are directed to testing equipment for quality confirmation, medium-risk batches are assigned to high-stability equipment groups for processing, and low-risk batches are processed using an efficiency-first path. This hierarchical scheduling model achieves rational allocation of high-end packaging equipment resources while ensuring process quality.

[0022] This application establishes a time-series quality characteristic dataset covering multiple processes, enabling correlation analysis and coordinated control of the entire process parameters of FCBGA carrier boards. This global perspective based on digital twins effectively solves the coordination problem caused by the independent control of each process in traditional production, significantly improving the overall stability of complex process systems.

[0023] This application incorporates quality risk entropy and delivery time factors into a dynamic priority algorithm, achieving an organic balance between quality control and delivery efficiency in scheduling decisions. This mechanism can both prevent batch quality incidents through risk warnings and ensure the timely delivery of low-risk orders, providing optimal overall benefits for FCBGA carrier board production. Attached Figure Description

[0024] To more clearly illustrate the technical solution of the present invention, the accompanying drawings used in the description will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0025] Figure 1 This is a flowchart illustrating the FCBGA carrier board production scheduling optimization method provided in this application.

[0026] Figure 2 This is a schematic diagram of the FCBGA carrier board production scheduling optimization system provided in this application.

[0027] Figure 3 A schematic diagram of the structure of the electronic device provided in this application. Detailed Implementation

[0028] The various embodiments of this disclosure will be described more fully in the detailed steps of the FCBGA carrier board production scheduling optimization method described below. This disclosure may have various embodiments, and adjustments and changes may be made therein. However, it should be understood that there is no intention to limit the various embodiments of this disclosure to the specific embodiments disclosed herein, but rather this disclosure should be understood to cover all adjustments, equivalents, and / or alternatives falling within the spirit and scope of the various embodiments of this disclosure.

[0029] In the following, the terms “comprising” or “may include”, which may be used in various embodiments of this disclosure, indicate the presence of the disclosed functions, operations, or elements, and do not limit the addition of one or more functions, operations, or elements. Furthermore, as used in various embodiments of this disclosure, the terms “comprising,” “having,” and their cognates are intended only to indicate a particular feature, number, step, operation, element, component, or combination of the foregoing, and should not be construed as primarily excluding the presence of one or more other features, numbers, steps, operations, elements, components, or combinations of the foregoing, or the possibility of adding one or more combinations of the foregoing.

[0030] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0031] Please see Figure 1 The diagram shown is a flowchart of a method for optimizing FCBGA carrier board production scheduling in a specific embodiment. The method includes: S1: Real-time acquisition of process parameter data for key processes during the production of FCBGA carrier boards.

[0032] In the production of FCBGA carrier boards, real-time acquisition of process parameter data is fundamental to achieving intelligent scheduling. This step establishes a data acquisition system covering the three core processes of laser drilling, electroplating, and lamination by deploying a dedicated sensor network in key operations.

[0033] In a specific implementation, a high-precision photoelectric sensor is configured during the laser drilling process to continuously monitor the laser's power output with a sampling period of 100ms. When a power fluctuation value exceeding a threshold of ±2% is detected, the abnormal fluctuation range and its duration are recorded, forming a laser power fluctuation dataset. This dataset contains three key indicators: peak deviation, fluctuation frequency, and duration, providing the initial basis for subsequent quality assessment.

[0034] Data acquisition for the electroplating process is achieved using Hall effect sensors. The sensors record the current density value of the electroplating bath at 1-second intervals, forming continuous time-series data of the electroplating current density. During implementation, three characteristic phases of the current density are identified: the rising phase, the stable phase, and the falling phase, with particular attention paid to the fluctuation range of the stable phase. When the current density deviates from the set value by more than ±5% and persists for more than 30 seconds, it is marked as a process anomaly.

[0035] Data acquisition for the lamination process employs a combination of infrared temperature sensors and piezoelectric pressure sensors. The temperature sensor monitors the lamination temperature at 2-second intervals with an accuracy controlled within ±1℃; the pressure sensor acquires pressure data at a 10Hz frequency with an accuracy of ±0.1MPa. The two data sets are precisely synchronized using timestamps, forming lamination temperature-pressure time-series data pairs. During implementation, the matching degree of the temperature-pressure curves is verified to ensure that the process parameters comply with the process specifications for this type of FCBGA carrier.

[0036] All collected data is transmitted as process parameter data to a central database via industrial Ethernet. The data is then cleaned and formatted to eliminate sensor noise and transmission errors, providing a reliable foundation for subsequent analysis.

[0037] S2: Based on the real-time collected process parameter data, create a quality risk digital twin for each production batch that dynamically evolves with the production progress, and generate quality characteristic data for quality risk prediction based on the process parameter data, which is stored in the quality risk digital twin.

[0038] In a specific implementation, a dynamically evolving quality profile is created for each production batch by establishing a data association framework and a feature extraction mechanism.

[0039] First, a unique QR code identifier is assigned to each new production batch, and a corresponding digital twin framework for quality risk is initialized. This framework adopts a hierarchical data structure, including a basic information layer, a process parameter layer, a feature indicator layer, and a risk assessment layer. When a batch enters a new process step, the collected process parameter data is automatically associated with the corresponding layer according to the process sequence and timestamp.

[0040] Then, in the feature extraction stage, on the one hand, the standard deviation of the laser power fluctuation data is calculated to generate a laser power stability coefficient. Specifically, the power data from the most recent 100 sampling points is taken, and their standard deviation is calculated. When this value exceeds 0.05, it is considered to have insufficient stability. On the other hand, for the electroplating current density time-series data, a 5-minute time window is set, and the maximum and minimum values ​​within the window are extracted. The difference between these values ​​is calculated as the electroplating uniformity index. In practice, a process warning is triggered when this index exceeds 15 A / m².

[0041] Thirdly, a dynamic time warping algorithm is used to calculate the lamination curve fit. During implementation, the real-time collected lamination temperature-pressure time-series data is aligned and compared with the standard process curve, and the minimum cumulative distance between the two curves is calculated. This distance value is normalized and converted into a similarity score, which is used as the lamination curve fit. In practical applications, when the fit is below 85%, the batch is considered to have a lamination quality risk.

[0042] Using the three calculation methods described above, three types of performance indicators are calculated and used as feature data.

[0043] Finally, the process parameter data and all feature data are organized according to the process flow to establish a complete process history. The data is then stored in the initial framework of the quality risk digital twin according to the process sequence and the collection timestamp, thus constructing a quality risk digital twin containing a complete process history.

[0044] The quality risk digital twin updates the process history once a process is completed, ensuring that it keeps pace with the physical production progress and providing complete data support for quality risk assessment.

[0045] S3: Call the quality feature data stored in the quality risk digital twin to calculate the quality risk entropy of the corresponding production batch.

[0046] In a specific implementation, quality feature data is first extracted from the digital twin to construct a time-series feature vector. This vector includes the laser power stability coefficient, electroplating uniformity index, and lamination curve fit of the three most recent processes, totaling nine feature dimensions. After standardization, the feature vector is input into a pre-trained quality risk prediction model.

[0047] This predictive model is based on a gradient boosting decision tree architecture and is trained using historical production data. The model contains 100 decision trees with a maximum depth of 6 layers and a learning rate of 0.1. During implementation, after receiving feature vectors, the predictive model performs ensemble calculations across multiple rounds of decision tree computation, ultimately outputting the quality failure probability F via a sigmoid activation function. For example, when the F value exceeds 0.7, it indicates a high probability that the batch will have quality defects in subsequent processes.

[0048] After obtaining the probability of quality failure F, according to the formula Calculate the quality risk entropy Q. The cost-weighted coefficient C is dynamically determined based on the material value of the FCBGA carrier and the processing costs already incurred. In practice, for carriers using high-end substrates, the C value is set to 1.5; for ordinary substrates, the C value is set to 1.0. This differentiated setting ensures the economic rationality of the risk assessment.

[0049] S4: Based on the quality risk entropy, combined with delivery date data and equipment status data, calculate the dynamic priority score for each production batch.

[0050] In a specific implementation, the order delivery date of the FCBGA carrier board for this production batch is obtained from the production planning system. The difference in working days between this date and the current date is calculated, and then the standard working hours for completing the remaining processes are deducted as the delivery leeway ST. For example, if a batch has 3 remaining processes, a total standard working time of 48 hours, and the delivery date is 5 working days away, then ST = 5 - 2 = 3 working days.

[0051] At the same time, all equipment that can be used for the next process of producing the FCBGA carrier board of this production batch is first screened out, and then the real-time availability status score and historical overall equipment efficiency value of each equipment are calculated together.

[0052] Specifically, the calculation of the real-time availability status score of the equipment adopts a hierarchical evaluation model. It includes three core indicators: equipment fault code, maintenance status, and current load rate.

[0053] First, equipment fault codes are categorized by severity: warning codes are weighted at 0.2, general fault codes at 0.5, and severe fault codes at 0.8. Fault information transmitted by the equipment PLC is analyzed in real time, and a fault impact coefficient is calculated based on the code category. Second, maintenance status is assessed based on equipment maintenance records. A score of 1.0 is given for maintenance within the standard cycle since the last preventative maintenance, 0.7 for maintenance within 30% of the cycle, and 0.4 for maintenance more than 30% of the cycle. Finally, the current load rate is obtained through the equipment monitoring system: a load rate below 80% scores 1.0, 80%-90% scores 0.8, and above 90% scores 0.5. The three indicators are weighted and summed in a 4:3:3 ratio to obtain the real-time availability status score. The calculation formula is as follows: Real-time availability score = Fault impact coefficient × 0.4 + Maintenance status score × 0.3 + Load rate score × 0.3.

[0054] Historical Overall Equipment Efficiency (OEE) values ​​are calculated based on equipment operating data. During implementation, data is collected across three dimensions: availability (uptime / planned production time), performance rate (actual output / standard output), and yield rate (number of qualified products / total output). The OEE value is obtained by multiplying these three metrics. OEE = Availability × Performance Rate × Yield Rate After the calculation is completed, take the OEE data of the most recent 30 production days, remove outliers, and calculate the moving average as the historical overall equipment efficiency value.

[0055] At this point, a dynamic weighting mechanism is used to synthesize the equipment efficiency factor Ee. For example, the real-time availability score and the historical overall equipment efficiency value are standardized to eliminate the influence of unit dimensions. Then, weights are allocated according to production strategy requirements: in quality-first mode, the real-time status weight is 0.6, and the historical efficiency weight is 0.4; in efficiency-first mode, the weights are adjusted to 0.4 and 0.6. The weighted calculation formula is: Ee = Real-time availability score × W1 + Historical overall device efficiency value × W2 Where W1 and W2 are dynamic weighting coefficients. After performing the above calculations on all available equipment, the equipment with the highest Ee value is selected as the target equipment, and its Ee value is the equipment efficiency factor for the current process.

[0056] Finally, the dynamic priority score S is calculated by substituting the quality risk entropy Q, delivery leniency ST, and equipment efficiency factor Ee into the following formula:

[0057] The weighting coefficients are dynamically adjusted based on the production strategy: set when quality is prioritized. =0.6, =0.2, =0.2; Set when delivery time is prioritized =0.6, =0.2, =0.2. This flexible configuration can adapt to different production needs.

[0058] S5: Based on dynamic priority scores and quality risk entropy, execute traffic splitting and scheduling decisions and generate scheduling instruction data.

[0059] In the specific implementation, first, the quality risk entropy Q is compared with a preset threshold. The high-risk threshold Th is set to 1.2, and the low-risk threshold Tl is set to 0.6. These thresholds are obtained based on historical data statistical analysis and can effectively distinguish production batches of different risk levels.

[0060] For high-risk batches with Q > Th, the first scheduling instruction data is generated and guided to the inspection station equipped with an online three-dimensional measuring instrument. During implementation, such batches will be preferentially arranged to enter the inspection queue to ensure quality confirmation before entering subsequent high-cost processes. The inspection station can complete a full scan of the carrier board within 5 minutes and accurately identify micron-level defects.

[0061] For medium-risk batches with Tl < Q ≤ Th, the second scheduling instruction data is generated and assigned to the high-stability equipment group ranked in the top 10% in terms of historical yield. These equipment are specially calibrated, and the process parameter control accuracy is 15% higher than that of ordinary equipment. During implementation, equipment capacity will be reserved for such batches to ensure the stability of the processing process.

[0062] For low-risk batches with Q ≤ Tl, the third scheduling instruction data is generated, and the shortest path strategy is used for scheduling. The optimal equipment and material paths for their processing cycle are planned to maximize production efficiency. During implementation, such batches can skip some non-critical inspection links and directly enter the next process.

[0063] S6: The scheduling instruction data is sent to the manufacturing execution system, and the manufacturing execution system analyzes the scheduling instruction data into control commands executable by production equipment and drives the corresponding production equipment to perform physical operations.

[0064] In the specific implementation, first, the scheduling instruction data is encapsulated into a standardized data packet. This data packet uses the JSON format and contains fields such as the target equipment number, processing process code, execution time window, process parameter requirements, etc. For example, the instruction assigned to the laser drilling process will clearly specify key parameters such as laser power and drilling speed.

[0065] Then, through the manufacturing execution system interface, the standardized instruction data packet is distributed to the corresponding workshop operation terminals and material handling systems. During implementation, the publish-subscribe mode is used to ensure that the instructions are accurately delivered to the target equipment. For high-priority instructions, a transmission retry mechanism is set to ensure the reliable transmission of the instructions.

[0066] During the execution of the instructions, the status feedback data from the manufacturing execution system is received in real time. These data include the equipment start processing time, completion time, actual process parameters, etc. By comparing the actual execution data with the scheduling instruction requirements, the consistency of the execution actions is verified. When it is found that the deviation exceeds the allowable range, a warning is immediately issued, and an exception handling process is started.

[0067] Simultaneously, this step also provides data support for subsequent scheduling optimization by establishing a complete execution log archive. These records include key indicators such as instruction issuance time, device response time, and execution quality, which are used to continuously improve the accuracy and reliability of the scheduling algorithm.

[0068] In this embodiment, a quality risk prediction model based on a gradient boosting decision tree was established by constructing a multi-source process parameter acquisition system and a quality risk digital twin, enabling accurate quantification and early warning of quality risks during FCBGA carrier board production. A scientific and comprehensive equipment performance evaluation system was formed by introducing a real-time equipment availability status scoring mechanism that includes fault code parsing, maintenance status assessment, and load rate monitoring, combined with a dynamic weighted fusion algorithm based on historical overall equipment efficiency values. Finally, based on multi-dimensional parameters such as quality risk entropy and equipment efficiency factors, a diversion scheduling decision mechanism that balances quality control and production efficiency was constructed, achieving a comprehensive and beneficial effect of significantly improving the production yield of high-value FCBGA carrier boards, optimizing equipment resource utilization, reducing material flow losses, and ensuring timely order delivery.

[0069] like Figure 2 As shown, the following are embodiments of the FCBGA carrier board production scheduling optimization system provided in this disclosure. This system and the FCBGA carrier board production scheduling optimization methods in the above embodiments belong to the same inventive concept. For details not described in detail in the embodiments of the FCBGA carrier board production scheduling optimization system, please refer to the embodiments of the above FCBGA carrier board production scheduling optimization methods.

[0070] An FCBGA carrier board production scheduling optimization system includes: The parameter acquisition module is used to collect process parameter data of key processes during the production of FCBGA carrier boards in real time.

[0071] The digital twin construction module is used to create a quality risk digital twin that dynamically evolves with the production progress for each production batch based on the real-time collected process parameter data, and to generate quality feature data for quality risk prediction based on the process parameter data, which is stored in the quality risk digital twin.

[0072] The risk quantification module is used to call the quality characteristic data stored in the quality risk digital twin to calculate the quality risk entropy of the corresponding production batch.

[0073] The priority calculation module is used to calculate the dynamic priority score for each production batch based on the quality risk entropy, combined with delivery date data and equipment status data.

[0074] The scheduling decision module is used to execute traffic allocation scheduling decisions and generate scheduling instruction data based on dynamic priority scores and quality risk entropy.

[0075] The instruction execution module is used to send scheduling instruction data to the manufacturing execution system, which then parses the scheduling instruction data into control commands that can be executed by the production equipment and drives the corresponding production equipment to perform physical operations.

[0076] The FCBGA carrier production scheduling optimization system provided in this embodiment achieves micron-level process monitoring of key processes such as laser drilling, electroplating, and lamination of FCBGA carriers by constructing a quality risk digital twin that integrates multi-source sensor data. Based on a gradient-enhancing decision tree model for quality risk prediction and an information entropy quantification method, a precise quality risk early warning mechanism is established. Through collaborative decision-making based on quality risk entropy and dynamic priorities, a diversion scheduling strategy is formed that prioritizes early intervention for high-risk batches, equipment optimization for medium-risk batches, and efficiency priority for low-risk batches. This data-driven intelligent scheduling system effectively reduces process losses of high-value carriers, transforming traditional passive quality control into a new model combining proactive prediction and adaptive regulation. While ensuring order delivery, it significantly improves production line yield and overall equipment utilization efficiency, achieving overall optimization of quality costs and production benefits.

[0077] Figure 3 A schematic diagram of the hardware structure of an electronic device for implementing various embodiments of the present invention.

[0078] The FCBGA carrier board production scheduling optimization method provided in this application can be applied to electronic devices. Those skilled in the art will understand that the electronic device structure involved in the embodiments of this invention does not constitute a limitation on the electronic device. An electronic device may include more or fewer components than illustrated, or combine certain components, or have different component arrangements. In the embodiments of this invention, the electronic device includes, but is not limited to, laptop computers, desktop computers, workbenches, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device may also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the embodiments of this application described and / or claimed herein.

[0079] Electronic devices may include processors, external memory interfaces, internal memory, universal serial bus (USB) interfaces, charging management modules, power management modules, batteries, wireless communication modules, audio modules, speakers, microphones, sensor modules, buttons, cameras, displays, and SIM card interfaces, etc.

[0080] A processor may include one or more processing units, such as: a central processing unit (CPU), an application processor (AP), a modem processor, a graphics processing unit (GPU), an image signal processor (ISP), a controller, memory, a video codec, a digital signal processor (DSP), a baseband processor, and / or a neural network processing unit (NPU). Different processing units may be independent devices or integrated into one or more processors.

[0081] The processor can serve as the nerve center and command center of an electronic device. The controller can generate operation control signals based on the instruction opcode and timing signals to control the fetching and execution of instructions.

[0082] The processor may also include memory for storing instructions and data. In some embodiments, the memory in the processor is a cache memory. This memory can store instructions or data that the processor has just used or that are used repeatedly. If the processor needs to use the instruction or data again, it can retrieve it directly from this memory. This avoids repeated accesses, reduces processor latency, and thus improves system efficiency.

[0083] An external storage interface (ESI) can be used to connect external memory cards, such as microSD cards, to expand the storage capacity of electronic devices. The external memory card communicates with the processor through the ESI to perform data storage functions, such as saving music and video files on the external memory card.

[0084] Internal memory can be used to store computer executable program code, which includes instructions. The processor executes various functional applications and data processing of electronic devices by running the instructions stored in internal memory. Internal memory can include a program storage area and a data storage area. Internal memory can include high-speed random access memory, and can also include non-volatile memory, such as at least one disk storage device, flash memory device, universal flash storage (UFS), etc.

[0085] Wireless communication functionality in electronic devices can be achieved through antennas, wireless communication modules, modem processors, and baseband processors.

[0086] Wireless communication modules can provide solutions for wireless communication applications in electronic devices, including wireless local area networks (WLANs) (such as wireless fidelity (Wi-Fi) networks), Bluetooth (BT), global navigation satellite system (GNSS), frequency modulation (FM), near field communication (NFC), and infrared (IR) technologies.

[0087] Electronic devices can implement audio functions through audio modules, speakers, receivers, microphones, headphone jacks, and application processors.

[0088] Electronic devices can achieve shooting functions through ISPs, cameras, video codecs, GPUs, displays, and application processors.

[0089] Electronic devices can achieve display functions through GPUs, displays, and application processors.

[0090] A GPU is a microprocessor for image processing, connected to the display screen and application processor. GPUs are used to perform mathematical and geometric calculations for graphics rendering. A processor may include one or more GPUs, which execute program instructions to generate or modify display information.

[0091] A display screen is used to display images, videos, etc. A display screen includes a display panel.

[0092] The aforementioned electronic equipment realizes the comprehensive beneficial effects of the FCBGA carrier production scheduling optimization method of this application by constructing a quality risk digital twin based on multi-source process parameters and establishing a dynamic scheduling mechanism driven by quality risk entropy. This achieves accurate prediction and proactive prevention and control of quality risks in the FCBGA carrier production process, significant reduction of process losses of high-value materials, effective improvement of equipment resource utilization efficiency, steady improvement of production yield, and reliable guarantee of order delivery timeliness.

[0093] The storage medium provided in this application stores a program product capable of implementing the FCBGA carrier board production scheduling optimization method.

[0094] FCBGA carrier board production scheduling optimization methods include: Real-time acquisition of process parameter data for key processes during FCBGA substrate production; Based on the real-time collected process parameter data, a quality risk digital twin is created for each production batch, which dynamically evolves with the production progress. Quality characteristic data for quality risk prediction is generated based on the process parameter data and stored in the quality risk digital twin. The quality feature data stored in the quality risk digital twin is used to calculate the quality risk entropy of the corresponding production batch. Based on the quality risk entropy, and combined with delivery date data and equipment status data, a dynamic priority score for each production batch is calculated. Based on dynamic priority scores and quality risk entropy, traffic diversion scheduling decisions are made, and scheduling instruction data is generated. The scheduling instruction data is sent to the manufacturing execution system, which then parses the scheduling instruction data into control commands that can be executed by the production equipment and drives the corresponding production equipment to perform physical operations. In some possible implementations, the FCBGA carrier board production scheduling optimization method of this disclosure can be implemented as a program product including program code that, when the program product is run on a terminal device, causes the terminal device to perform the steps described in the "Exemplary Methods" section of this specification according to various exemplary embodiments of this disclosure.

[0095] The storage medium disclosed herein may be any combination of one or more readable media. A readable medium may be a readable signal medium or a readable storage medium. A readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples (a non-exhaustive list) of readable storage media include: an electrical connection having one or more wires, a portable disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof.

[0096] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for optimizing the production scheduling of FCBGA carrier boards, characterized in that, include: Real-time acquisition of process parameter data for key processes during FCBGA substrate production; Based on the real-time collected process parameter data, a quality risk digital twin is created for each production batch, which dynamically evolves with the production progress. Quality characteristic data for quality risk prediction is generated based on the process parameter data and stored in the quality risk digital twin. The quality feature data stored in the quality risk digital twin is used to calculate the quality risk entropy of the corresponding production batch. Based on the quality risk entropy, and combined with delivery date data and equipment status data, a dynamic priority score for each production batch is calculated. Based on dynamic priority scores and quality risk entropy, traffic diversion scheduling decisions are executed, and scheduling instruction data is generated. The scheduling instruction data is sent to the manufacturing execution system, which then parses the scheduling instruction data into control commands that can be executed by the production equipment and drives the corresponding production equipment to perform physical operations.

2. The FCBGA carrier board production scheduling optimization method according to claim 1, characterized in that, The real-time acquisition of process parameter data for key processes during FCBGA carrier board production includes: Real-time laser power fluctuation data, which characterizes drilling quality, is collected by photoelectric sensors deployed in the laser drilling process. By deploying Hall effect sensors in the electroplating process, time-series data of electroplating current density characterizing the uniformity of the plating layer are collected. Infrared temperature sensors and piezoelectric pressure sensors deployed in the lamination process are used to collect lamination temperature-pressure time-series data characterizing the lamination quality. Based on the laser power fluctuation data, electroplating current density time series data, and lamination temperature-pressure time series data, process parameter data are generated.

3. The FCBGA carrier board production scheduling optimization method according to claim 2, characterized in that, The process parameter data collected in real time is used to create a quality risk digital twin for each production batch that dynamically evolves with the production progress. Quality characteristic data for quality risk prediction is generated based on the process parameter data and stored in the quality risk digital twin, including: The process parameter data collected in real time are associated and stored with the corresponding batch unique identifier according to the production time sequence, and an initial framework for the corresponding quality risk digital twin is created. Based on the laser power fluctuation data, a laser power stability coefficient is generated by calculating its standard deviation. Based on the electroplating current density time series data, an electroplating uniformity index is generated by calculating the difference between its maximum and minimum values ​​within a set time window. Based on the lamination temperature-pressure time series data, it is compared with the pre-stored standard lamination temperature-pressure time series data of this model of FCBGA carrier board. The similarity between the two time series data curves is calculated by the dynamic time warping algorithm to generate the lamination curve fit. The process parameter data, along with the calculated laser power stability coefficient, electroplating uniformity index, and lamination curve fit, are used as quality feature data. These data are then stored in the initial framework of the quality risk digital twin according to the process sequence and the acquisition timestamp, thus constructing a quality risk digital twin containing a complete process history.

4. The FCBGA carrier board production scheduling optimization method according to claim 3, characterized in that, The step of calling the quality feature data stored in the quality risk digital twin to calculate the quality risk entropy of the corresponding production batch includes: Extract the quality feature data from the quality risk digital twin to form a time-series feature vector; Input the time-series feature vector into a pre-trained quality risk prediction model. The quality risk prediction model is a machine learning model based on the gradient boosting decision tree architecture. The dimension of its input layer matches the dimension of the time-series feature vector, and the output layer outputs a quality failure probability F between 0 and 1 through the Sigmoid activation function; Substitute the quality failure probability F into the following formula for calculation to generate the quality risk entropy Q for the corresponding batch: where C is a cost weighting factor that is positively correlated with the value and the invested cost of the FCBGA carrier board for this production batch.

5. The FCBGA carrier board production scheduling optimization method according to claim 4, characterized in that, Based on the quality risk entropy, combine the delivery date data and the equipment status data to calculate the dynamic priority score for each production batch, including: Obtain the delivery date of the FCBGA carrier board for this production batch and the current time from the production planning system. By calculating the number of working days difference between the delivery date and the current date, and subtracting the standard working hours required to complete the remaining processes, obtain the delivery date slack ST in working days; For the next pending process of the FCBGA carrier board for this production batch, obtain the status and historical performance data of all available devices corresponding to this process type from the equipment management system. By comprehensively calculating the real-time available status score and the historical overall equipment efficiency value of each device, and performing weighted fusion according to the preset weight, select the optimal value as the equipment efficiency factor Ee; Based on the quality risk entropy Q, the delivery date slack ST, and the equipment efficiency factor Ee, calculate the dynamic priority score S through the following formula: in, , , These are the preset weighting coefficients.

6. The FCBGA carrier board production scheduling optimization method according to claim 5, characterized in that, Based on the dynamic priority score and the quality risk entropy, execute the shunt scheduling decision to generate scheduling instruction data, including: Compare the quality risk entropy Q with the preset high-risk threshold Th and low-risk threshold Tl respectively; If Q > Th, generate the first scheduling instruction data for guiding the FCBGA carrier board of this production batch to the inspection station with an online three-dimensional measurement function; If Tl < Q ≤ Th, generate the second scheduling instruction data for preferentially allocating the subsequent processing tasks of the FCBGA carrier board of this production batch to the equipment group with the top 10% historical yield ranking; If Q ≤ Tl, generate the third scheduling instruction data for planning and locking the material handling path and equipment sequence with the shortest processing cycle for the FCBGA carrier board of this production batch.

7. The FCBGA carrier board production scheduling optimization method according to claim 6, characterized in that, Send the scheduling instruction data to the manufacturing execution system. The manufacturing execution system analyzes the scheduling instruction data into control commands executable by the production equipment and drives the corresponding production equipment to perform physical operations, including: Standardize and encapsulate the scheduling instruction data to form a standardized instruction data packet including the target equipment number, processing process code, and execution time window; Distribute the standardized instruction data packet to the corresponding workshop operation terminal and material handling system through the manufacturing execution system interface; ​ 8. An FCBGA carrier board production scheduling optimization system, characterized in that, The system employs the FCBGA carrier board production scheduling optimization method as described in any one of claims 1 to 7; The system includes: The parameter acquisition module is used to collect process parameter data of key processes during the production of FCBGA carrier boards in real time. The twin construction module is used to create a quality risk digital twin that dynamically evolves with the production progress for each production batch based on the real-time collected process parameter data, and to generate quality feature data for quality risk prediction based on the process parameter data, which is stored in the quality risk digital twin. The risk quantification module is used to call the quality feature data stored in the quality risk digital twin to calculate the quality risk entropy of the corresponding production batch. The priority calculation module is used to calculate the dynamic priority score of each production batch based on the quality risk entropy, combined with delivery date data and equipment status data. The scheduling decision module is used to execute traffic allocation scheduling decisions and generate scheduling instruction data based on dynamic priority scores and quality risk entropy. The instruction execution module is used to send scheduling instruction data to the manufacturing execution system, which then parses the scheduling instruction data into control commands that can be executed by the production equipment and drives the corresponding production equipment to perform physical operations.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the steps of the FCBGA carrier board production scheduling optimization method as described in any one of claims 1 to 7.

10. A storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the FCBGA carrier board production scheduling optimization method as described in any one of claims 1 to 7.

Images