Method and system for optimizing precise baking process of mini LED circuit board

Abstract

The invention provides a precise baking process optimization method and system for a mini LED circuit board, a baking oven is divided into a preheating area, a main baking area and a cooling area, a double-circulation hot air system and top vertical airflow are adopted, infrared and digital humidity sensors are combined, independent temperature and humidity regulation and control are realized, nitrogen-argon mixed gas is injected, and opening and closing of injection holes are adjusted based on the thickness of a substrate, so that the precise baking process of the mini LED circuit board is realized. Maintaining positive pressure and oxygen content in the box based on a three-dimensional pressure sensor network, prolonging heating time according to environment humidity in a preheating stage, performing gradient control on a heating rate, constructing a humidity-moisture absorption rate mapping table and a three-layer neural network model, adjusting power output in combination with Z-axis temperature difference, and adopting forced air cooling and nitrogen sealing procedures in a cooling area; and a plate explosion prediction model is established, a historical data triggering parameter adjustment mechanism is combined, baking parameters are written into the two-dimensional code of the substrate through a block chain technology, and traceability of the whole process is realized.

Description

TECHNICAL FIELD

[0001] The application relates to a miniLED circuit board precision baking process optimization method and system. BACKGROUND

[0002] Traditional ovens mostly use single temperature zone control, which is difficult to realize independent temperature and humidity adjustment in the preheating, main baking and cooling stages, resulting in large Z-axis temperature difference of the substrate, easy to cause uneven thermal stress and board explosion risk, and the humidity control relies on fixed parameters without adjustment in combination with the substrate thickness, storage period and the like, so that the moisture absorption rate is difficult to accurately control, the existing process mostly uses single nitrogen protection without introducing argon synergistic effect, the oxygen content control precision is low, and there is lack of gas flow adjustment mechanism based on pressure sensors, resulting in the risk of substrate oxidation or excessive drying.

[0003] Traditional baking parameters including temperature and time are mostly fixed values, which are not optimized according to the environmental humidity or Z-axis temperature difference, resulting in low water removal efficiency in high humidity environment, affecting the welding quality, lacking a board explosion prediction model based on historical data, being unable to identify potential defects caused by temperature and humidity fluctuations, moisture absorption rate anomalies and the like in advance, relying on manual experience for judgment, and being difficult to guarantee the yield, and the existing process does not realize the linkage of the blockchain record of the full-process data and the MES system, resulting in lagging parameter adjustment and being difficult to form a closed-loop optimization. SUMMARY

[0004] The application aims to provide a miniLED circuit board precision baking process optimization method and system.

[0005] The application solves the problems of insufficient temperature control precision, passive humidity management, unstable gas environment, static parameter defects, lagging risk prediction and missing data traceability in the existing MiniLED circuit board baking process.

[0006] A miniLED circuit board precision baking process optimization method adopts the technical scheme as follows: S1: dividing the miniLED circuit board oven into multiple temperature control regions, each region using an independent temperature and humidity control system, and configuring a double-circulation hot air system and a vertical air flow guiding device; S2: injecting an inert gas mixture into the oven, adjusting the gas injection amount according to the substrate characteristics, adjusting the gas flow through a pressure and humidity monitoring system, and maintaining a stable positive pressure environment in the oven; S3: setting a gradient temperature rising program in the preheating region, adaptively adjusting the preheating time and temperature rising rate according to the initial humidity condition of the environment, and controlling the smooth transition of the temperature rising process; S4: Establish a multi-parameter control mechanism in the main baking area. By monitoring the temperature difference inside the circuit board in real time, combining the mapping relationship between the substrate storage state and physical characteristics, and adjusting the power output of the heat source; S5: Implement a controlled cooling program in the cooling area, maintain an inert gas environment to reduce the risk of oxidation, and set up a humidity isolation device at the outlet to perform gas storage operation before the substrate is taken out of the oven; S6: Build a risk prediction model for the baking process. By comprehensively analyzing the temperature and humidity parameters, material storage state and structural characteristics, the risk of board explosion during the baking process is evaluated in real time; S7: When there is a risk of board explosion, optimize and adjust the baking conditions based on the historical process parameter database; S8: Write the key parameter information of the baking process into the substrate two-dimensional code, and realize traceable management of process parameters through blockchain technology.

[0007] Further, the miniLED circuit board oven in S1 is divided into three areas, using a double-circulation hot air system and a top-down vertical air flow, including: The internal space of the oven is divided into multiple functional areas, which are physically isolated by high-temperature-resistant partitions. Each functional area is equipped with an independent temperature and humidity control unit. The temperature and humidity monitoring uses a combination of non-contact sensing devices and digital humidity sensors. The monitoring data is transmitted in real time to the production management system through a wireless communication module; The oven is equipped with a double-circulation hot air structure, including an internal circulation system and an external circulation system. The internal circulation system uses a top-down vertical air flow design to cover the surface of the substrate to be processed. The external circulation system forms an external air circulation loop through the air flow channels set in the side wall. The hot air outlet is equipped with a flow uniformization device to reduce air flow disturbance and improve temperature field distribution uniformity through a microporous structure; The vertical air flow system uses a top-distributed air outlet design, with air flow direction strictly perpendicular to the substrate plane, effectively reducing the interference of horizontal air flow on the substrate surface.

[0008] Further, the S2 injects nitrogen and argon mixed gas into the oven, adjusts the opening and closing number of the injection hole based on the substrate thickness, and adjusts the gas flow combined with the pressure sensor and humidity, including: Multiple gas delivery channels are configured at the entrance of the oven, connected with different inert gas sources respectively. The proportion of each gas component is controlled by precise flow adjustment devices, and the oxygen content in the inert gas environment is maintained at a very low level. The opening state of the gas injection port is adjusted according to the substrate thickness. More injection ports are opened for thicker substrates, and the opening proportion of the injection port is reduced for thinner substrates; The multi-point pressure monitoring device is arranged inside the oven to form a spatially distributed pressure sensing network, and the gas injection parameters are adjusted according to a preset pressure threshold value; when the pressure of a local area deviates from the set range, the gas injection amount or the injection port switch state of the area is adjusted; meanwhile, a humidity monitoring feedback mechanism is configured, and when the humidity of a specific area is detected to be abnormal, the gas replacement rate of the area is correspondingly increased to realize the control of the internal environment parameters of the oven.

[0009] Further, the S3 is based on a humidity sensor, and the preheating area temperature rising time and rate are adjusted according to the initial humidity of the environment, including: A gradient temperature rising program is set for the preheating area, a variable frequency power adjusting device is used to control the heat source output, the temperature change rate is monitored, and the subsequent process preparation process is triggered after the preheating stage is completed; Humidity sensing devices are arranged at different height positions in the preheating area to build a vertical humidity distribution monitoring network, and the preheating time and temperature rising rate are adjusted according to the initial environmental humidity conditions; when the environmental humidity reaches a certain threshold value, the preheating time is proportionally prolonged and the temperature rising gradient is reduced, and at the same time, an upper limit of the total preheating time is set; When the environmental humidity exceeds the high threshold value and the humidity is still higher than the set standard after the prolonged preheating, a humidity abnormality early warning mechanism is triggered, the preheating process is suspended and manual intervention is prompted; A temperature abnormality protection mechanism is configured, when the temperature rising rate is detected to exceed a safe range, the heat source power output is reduced, and when the temperature exceeds the preset upper limit for multiple times, the power is further reduced and an alarm signal is triggered to prompt the operator to check the state of the heating system.

[0010] Further, the S4 is that the MES system establishes a humidity-hygroscopic rate mapping table according to the substrate storage period and thickness, establishes a power adjustment model based on the humidity-hygroscopic rate mapping table and the Z-axis temperature difference, and adjusts the power output of the main baking area, including: S41: Non-contact temperature monitoring devices are arranged at different height positions in the main baking area to build a temperature gradient monitoring network along the thickness direction of the substrate, temperature data are collected and transmitted to the production management system, and the temperature difference between different layers of the substrate is calculated; S42: Based on the historical storage data and physical characteristics of the substrate, the humidity absorption characteristics of the substrate under different storage periods and structure parameters are obtained, and a mapping relationship database between the humidity environment and the material hygroscopic rate is established, which covers multiple specifications of substrate physical parameters; S43: A multi-parameter power adjustment model is constructed, which takes the temperature gradient, material hygroscopic characteristics and substrate physical structure as input parameters, calculates the heat source power output value through a preset adjustment rule and a machine learning algorithm, wherein the adjustment rule sets different power compensation strategies according to the temperature gradient and the hygroscopic rate, and the machine learning model adopts a multi-layer network structure to train the historical process data to predict the power adjustment amount; S44: Based on the model calculation results, the output power of the heat source is adjusted by the frequency converter. When the power adjustment reaches the predetermined threshold, the airflow speed of the hot air circulation system is adjusted synchronously to ensure the uniformity of heat distribution. At the same time, the flow parameters of the protective gas are coordinated and adjusted to maintain the stability of the gas environment inside the oven. S45: Configure a temperature anomaly monitoring mechanism. When the temperature gradient is detected to continuously exceed the safe range, reduce the heat source power and trigger an early warning signal to prompt the operator to check the airflow distribution. After each batch is processed, execute the sampling and testing procedure and save the optimized process parameters to the system database based on the test results.

[0011] Furthermore, in S5, the cooling zone provides forced air cooling to the baked circuit board, with nitrogen continuously injected during the cooling process. A nitrogen sealing procedure is triggered before the substrate leaves the box, including: Multiple sets of industrial-grade airflow devices are configured at the top of the cooling zone, forming a vertically downward, full-area airflow channel with a special spatial layout. The airflow speed is precisely controlled by a frequency converter to monitor the temperature changes in each area of ​​the circuit board. When an abnormal cooling rate is detected in a local area, the airflow intensity in the corresponding area is adjusted. Configure a gas environment monitoring and regulation mechanism. When the concentration of protective gas is detected to be lower than the set threshold, increase the amount of protective gas injected and adjust the cooling airflow speed accordingly. If the gas environment continues to be abnormal, start the auxiliary gas replacement device to purify the environment, and then restore the normal protective gas injection process. A multi-layer isolation structure is designed at the oven outlet, including an inner gas sealing device and an outer mechanical locking device. A transition buffer space is formed between the two layers. Before the substrate leaves the oven, a gas protection sealing procedure is automatically performed, injecting protective gas into the buffer space to a preset pressure value and maintaining it for a specific time. Then, the substrate is sealed in a special protective bag, and a dual protection process of vacuuming and protective gas filling is performed to ensure a smooth transition of the substrate from the baking environment to the external environment.

[0012] Furthermore, in step S6, a baking-induced board explosion prediction model is established to predict whether the circuit board is at risk of exploding during the baking process, including: By collecting and analyzing the rate of change of temperature and humidity fields in the preheating zone and the main drying zone, characteristic parameters related to the integrity of the substrate structure are extracted. Based on the physical characteristics of substrates of different specifications, characteristic weight coefficients are set. Statistical anomaly detection methods are used to identify and remove outliers in the sensor data. Interpolation algorithms are used to reasonably complete missing values ​​in the production data. A risk prediction model based on artificial neural networks is constructed. The model adopts a multi-layer network structure design, takes process environment parameters, material storage state and physical structure parameters as input features, and structural failure probability as output index. The model accuracy is optimized by dividing the feature space, and the model is trained and validated using historical production datasets. Cross-validation technology is used to evaluate the model's generalization ability and control the prediction error range. A risk prediction mechanism is triggered at key temperature nodes during the baking process. An initial risk assessment is conducted based on the stability of the internal temperature gradient of the substrate. At specific temperature points during the cooling stage, a secondary risk verification is performed based on the protective gas environment parameters and humidity isolation status. The risk level is divided into multiple levels according to the predicted probability value, providing a basis for decision-making for subsequent process parameter adjustments.

[0013] Furthermore, if there is a risk of the baking plate bursting in S7, an adjustment mechanism is triggered to adjust the baking parameters based on historical baking parameter data in the MES system, including: Based on the output of the risk prediction model, combined with the temperature gradient monitoring data inside the substrate and the environmental humidity monitoring information, a multi-dimensional risk assessment mechanism is established. When any monitoring indicator exceeds the preset threshold, the risk warning process is automatically triggered. When multiple key indicators exceed the safe range at the same time, the emergency protection procedure is immediately activated to cut off the heat source power supply and shut down the airflow circulation system and the protective gas supply device. Retrieve relevant process records from the historical production database, including material physical properties, storage status, process environment parameters, and quality event markers. Based on the characteristic parameters of the currently processed substrate, select a set of process parameters with similar characteristics and stable quality as a reference for optimization and adjustment. Based on a comparative analysis of current process parameters and historical optimal parameter sets, parameter adjustment schemes are generated, including optimizing the setting of the main drying temperature. If there are differences between the current parameters and historical optimal parameters, the settings are adjusted to the range of historical optimal parameters. The preheating stage duration is adaptively extended to ensure that the material fully adapts to temperature changes. The supply of protective gas is adjusted to maintain the stability of the gas environment inside the oven, thereby achieving intelligent optimization of process parameters and proactive prevention and control of quality risks.

[0014] Furthermore, a precision baking process optimization system for miniLED circuit boards is provided to implement the precision baking process optimization method for miniLED circuit boards described in any of the above claims. The precision baking process optimization system for miniLED circuit boards includes: a temperature zone control module, an inert gas injection control module, a preheating parameter adjustment module, a main baking power adjustment module, a cooling and sealing control module, a board explosion prediction and risk warning module, a parameter adjustment and emergency response module, and a data recording and full-process traceability module. Temperature zone control module: used to realize physical isolation and environmental control of the internal space of the oven. Each functional area is separated by a high temperature resistant isolation structure. Each area is equipped with a non-contact temperature monitoring device and a digital humidity sensor. The monitoring data is transmitted to the production management system via wireless communication. The airflow system adopts a design that combines vertical air supply from top to bottom with dual circulation hot air to ensure uniform temperature field distribution. Inert gas injection control module: Configured with multiple gas delivery channels and flow regulation devices, constructing a spatially distributed pressure monitoring network, adjusting gas injection parameters according to changes in regional pressure and humidity, maintaining a stable positive pressure environment inside the chamber, and increasing the gas replacement rate when humidity is abnormal; Preheating parameter adjustment module: Configure a vertically distributed humidity monitoring network, adjust the preheating time and temperature gradient according to the initial humidity conditions of the environment, set a temperature anomaly protection mechanism, reduce the heat source power when the temperature rise rate or temperature value exceeds the safe range, and trigger an alarm and suspend the preheating process when the humidity continues to exceed the standard. Main drying power adjustment module: Constructs a temperature gradient monitoring network along the thickness of the substrate through a multi-point temperature monitoring device, establishes a moisture absorption characteristic database based on the physical properties of the material and the storage state, constructs a multi-parameter power adjustment model, adjusts the heat source output according to the temperature gradient and the moisture absorption state of the material, and synchronously coordinates the airflow speed and protective gas flow parameters. Cooling and sealing control module: A specially arranged airflow device is used to form a uniform cooling environment. It is equipped with a gas concentration monitoring and abnormal handling mechanism. When an abnormal protective gas concentration is detected, the gas flow rate is adjusted and the cooling intensity is reduced. If the abnormality continues, the auxiliary purification device is activated. The outlet area is equipped with a multi-layer isolation structure and a gas protection packaging system to achieve a smooth transition of the substrate from the baking environment to the external environment. Explosion prediction and risk warning module: By analyzing the temperature and humidity change rate in the process, and combining the physical properties of materials, a risk assessment model is constructed. Artificial intelligence algorithms are used to train and verify historical data to establish a multi-stage risk prediction mechanism, and risk assessment and graded warning are carried out at key process nodes. Parameter Adjustment and Emergency Response Module: Implements a graded response strategy based on the risk warning level. When a single indicator is abnormal, the warning process is triggered. When multiple indicators exceed the limit, the emergency protection procedure is initiated. Historical high-quality process parameter sets are retrieved, and after comparative analysis, parameter optimization schemes are generated. The process is restarted and the risk status is continuously monitored. Data recording and full-process traceability module: Key process parameter information is written into the substrate QR code, and an immutable process record is established through blockchain technology to realize full-process data traceability and quality control from raw materials to finished products.

[0015] The present invention also provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps in the above-described method for optimizing the precision baking process of a miniLED circuit board.

[0016] The beneficial effects of this invention are: by using three-dimensional temperature and humidity zone control and dual-circulation vertical airflow design, combined with an infrared non-contact sensor array, it achieves precise control of the substrate Z-axis temperature difference ≤5℃. The synergistic effect of nitrogen-argon mixed gas and pressure regulation keeps the oxygen content below 50ppm, effectively suppressing oxidation reaction. The temperature difference fluctuation between the substrate surface and interior is reduced, reducing the risk of board explosion caused by thermal stress. The power adjustment model built based on the multi-dimensional data fusion of the MES system realizes the adjustment of the main baking power. Combined with the linkage compensation of Z-axis temperature difference and moisture absorption rate, the baking parameters of different batches of substrates are adaptively optimized. Through the dual prediction model and the explosion probability classification, combined with the emergency response mechanism and the secondary prediction of the cooling stage, the probability of later failure is reduced. A preheating time compensation algorithm and gradient heating strategy for high humidity environments improve the moisture migration efficiency of humid substrates. Based on the synergy of a three-dimensional pressure sensor network and injection hole control, the gas permeation uniformity of substrates of different thicknesses is ensured. The combination of blockchain technology and QR code writing enables the solidification of data throughout the entire process, from substrate storage cycle to baking parameters. Attached Figure Description

[0017] Figure 1 A flowchart of an optimized precision baking process for miniLED circuit boards; Figure 2 This is a system module diagram for optimizing the precision baking process of miniLED circuit boards. Detailed Implementation

[0018] The present invention will be further described clearly and completely below, but the scope of protection of the present invention is not limited thereto.

[0019] Furthermore, the term "and / or" in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. Additionally, the character " / " in this article, unless otherwise specified, generally indicates that the preceding and following related objects have an "or" relationship.

[0020] Example 1 An optimized method for precision baking process of miniLED circuit boards, the technical solution of which is as follows: S1: The miniLED circuit board oven is divided into three zones: preheating, main drying, and cooling. The temperature and humidity of each zone are independently controlled, and a dual-circulation hot air system and a top-down vertical airflow are used. S2: Inject a mixture of nitrogen and argon into the oven, with a nitrogen to argon volume ratio of 4:1. Adjust the number of opening and closing injection holes based on the substrate thickness, and adjust the gas flow rate in conjunction with the pressure sensor and humidity to maintain a positive pressure of 0.2 kPa inside the oven. S3: Set the initial preheating temperature to 60±2℃, the initial heating time to 10min, the total continuous heating time to no more than 30min, the heating rate to ≤2℃ / min, and adjust the heating time and rate of the preheating area according to the initial ambient humidity based on the humidity sensor. S4: Set the main drying target temperature to 120±2℃, monitor the Z-axis temperature difference of the miniLED circuit board containing multi-layer HDI substrate, and the MES system establishes a humidity-moisture absorption rate mapping table based on the substrate storage cycle and thickness. Based on the humidity-moisture absorption rate mapping table and the Z-axis temperature difference, a power adjustment model is established to adjust the power output of the main drying area. S5: The cooling zone uses forced air cooling to cool the baked circuit boards at a rate of ≤2℃ / min. Nitrogen is continuously injected during the cooling process, and the oxygen content is maintained at ≤50ppm. A humidity locking device is installed at the oven outlet, and a nitrogen sealing procedure is triggered before the substrate leaves the oven. S6: Establish a baking explosion prediction model. Input parameters include temperature and humidity, storage period, and substrate thickness. Based on the model, predict whether there is a risk of the circuit board exploding during the baking process. S7: If there is a risk of the baking plate bursting, the adjustment mechanism will be triggered to adjust the baking parameters based on the historical baking parameter data in the MES system. S8: The final baking parameters, including temperature, time, and humidity, are written into the substrate's QR code, and the entire process is traceable through blockchain technology.

[0021] refer to Figure 1 The diagram shown is a flowchart of an optimized method for precision baking process of miniLED circuit boards.

[0022] Furthermore, in S1, the miniLED circuit board oven is divided into three areas, employing a dual-circuit hot air system and a top-down vertical airflow, including: The oven interior is divided into three physically isolated zones: a preheating zone (heating and dehumidification), a main drying zone (constant temperature curing), and a cooling zone (temperature gradient control). Each zone is spatially isolated by a high-temperature resistant partition with a thickness of ≥5mm. This thickness design ensures sufficient thermal insulation performance, preventing temperature crosstalk caused by heat conduction between zones and allowing each zone to maintain the set temperature and humidity environment. The edges of the partitions are treated with silicone sealing strips to further reduce gas exchange. Each temperature zone is equipped with an independent temperature and humidity control system, including heating elements, cooling devices, and a dehumidification module. The temperature sensor uses an infrared non-contact temperature sensor with an accuracy of ±0.5℃, monitoring the ambient temperature. It is installed 15cm above the substrate surface to avoid direct contact with the substrate and causing thermal interference. The humidity sensor integrates a digital humidity sensor with an accuracy of ±0.2%RH and is located at the airflow intersection. The temperature and humidity sensors transmit data to the MES system via a LoRa wireless module with a transmission interval of 1 second.

[0023] A dual-circulation hot air system, comprising an internal circulation system and an external circulation system, is installed at the top of the oven. The internal circulation system uses a top fan to drive a downward vertical airflow to cover the substrate surface. The airflow is distributed through a regular array of 2mm diameter outlets spaced 10cm apart, with an air velocity set between 0.5-1.2m / s to ensure uniform heat distribution. This parameter range is the optimal value determined through experiments. Air velocities below 0.5m / s will lead to uneven heat distribution, while velocities above 1.2m / s may cause substrate displacement or surface contamination. The 10cm outlet spacing ensures an airflow coverage rate of over 95%, avoiding... To avoid dead zones, the vertical airflow direction is perpendicular to the substrate plane, effectively reducing the disturbance of the substrate by the lateral airflow. The external circulation system has 4 sets of air inlets and 4 sets of exhaust outlets on the side wall of the oven, arranged diagonally to form a cross-shaped airflow with an air exchange rate of 15-20 times per hour, forming external air circulation and maintaining stable temperature and humidity inside the oven. A honeycomb-shaped flow equalization plate with a 1mm aperture is installed at the hot air outlet to reduce airflow turbulence through porous diffusion, and the temperature field uniformity is maintained at ±0.8℃. The two circulation systems work together, with the internal circulation ensuring uniform heat transfer on the substrate surface and the external circulation maintaining overall environmental stability.

[0024] Furthermore, in step S2, a mixture of nitrogen and argon gas is injected into the oven. The number of opening and closing injection holes is adjusted based on the substrate thickness, and the gas flow rate is adjusted in conjunction with pressure sensors and humidity. This includes: A dual-gas delivery pipeline is installed at the oven inlet, connecting to a nitrogen tank and an argon tank respectively. The flow rate is adjusted by a mass flow controller, the mixed gas volume ratio is controlled at 4:1, and the oxygen content is controlled to not exceed 50ppm. This threshold is a key indicator to prevent miniLED chip oxidation failure. Gas injection holes are distributed in a matrix on the four walls of the oven. Each wall has 8×8 micro solenoid valve control holes with a diameter of 1.5mm. The relationship between the substrate thickness and the number of injection holes open and closed is determined based on the heat transfer equation. The number of injection holes open and closed is adjusted proportionally according to the substrate thickness. The 1.6mm thick substrate has a large heat capacity and requires more airflow to penetrate for uniform dehumidification, so all injection holes are opened. The 0.8mm thick substrate has a small heat capacity and is easily deformed by airflow impact, so 50% of the injection holes are opened.

[0025] A high-precision pressure sensor with a range of 0-1 kPa and an accuracy of ±0.01 kPa is used, installed at the top, middle, and bottom of the oven to form a three-dimensional pressure monitoring network. The target pressure value is set at 0.2 kPa with an error of ±0.02 kPa. Sufficiently high positive pressure (>0.15 kPa) can effectively prevent the infiltration of external humid air, while excessive pressure (<0.25 kPa) avoids mechanical stress on the substrate structure. When the pressure in a certain area is detected to be below 0.18 kPa, the gas injection flow rate in that area is increased by no more than 10% to prevent sudden pressure changes. When the pressure exceeds 0.22 kPa, the injection holes far from the substrate surface are closed first, and then the flow rate is reduced to prevent damage to the substrate due to pressure fluctuations. When the humidity in a certain area exceeds 30%RH, the gas injection flow rate in that area is increased by no more than 15%; all gas parameter adjustments are achieved through a PID controller.

[0026] Furthermore, in step S3, based on a humidity sensor, the heating time and rate of the preheating zone are adjusted according to the initial ambient humidity, including: The initial temperature of the preheating zone is set at 60±2℃. Temperatures below 58℃ will result in insufficient moisture evaporation, while temperatures above 62℃ may cause premature curing of the substrate, affecting the subsequent main baking effect. The heating rate is ≤2℃ / min to ensure that the temperature difference between the inner and outer layers of the substrate does not exceed 5℃, preventing microcracks caused by thermal stress. The power is adjusted by the frequency conversion heating module, and the temperature data is sampled every 2 seconds. One digital humidity sensor with an accuracy of ±0.2%RH is installed at the top (15cm from the top plate) and the bottom (20cm from the substrate surface) of the preheating zone to form a vertical humidity monitoring network. When the initial ambient humidity is ≥40%RH, the preheating time is extended. For every 10%RH increase in humidity, the time is extended by 5 minutes. Based on the balance between production cycle and substrate heat resistance limit, the total continuous heating time is set to not exceed 30 minutes, while the heating rate is reduced to 1.2℃ / min.

[0027] When the initial ambient humidity is ≥50%RH and the preheating time is extended to 30 minutes, and the humidity is still ≥45%RH, a high humidity warning is triggered, preheating is suspended, and an inspection is carried out. This threshold setting is based on the analysis of 200 batches of historical data. If baking continues under these conditions, the board bursting rate will increase. When the heating rate is detected to exceed 2℃ / min, the heating power is reduced by 10%-20%. This protection threshold is calculated based on the heat conduction equation and the thermal capacity characteristics of the substrate to prevent local overheating. When the temperature of the preheating zone exceeds 62℃ three times consecutively, the heating power is reduced by 20%, and an alarm is triggered to prompt the operator to check the status of the heating module, preventing malfunctions caused by occasional sensor noise. When 30 minutes have elapsed, the main baking zone preparation program is triggered. The system sends a preparation signal to the main baking zone 5 minutes in advance, allowing it to gradually heat up to the target temperature.

[0028] Furthermore, in S4, the MES system establishes a humidity-hygroscopic rate mapping table based on the substrate storage cycle and thickness, and establishes a power adjustment model based on the humidity-hygroscopic rate mapping table and the Z-axis temperature difference to adjust the power output of the main drying area, including: S41: The target temperature for the main baking zone is set at 120±2℃. Temperatures below 118℃ will result in incomplete curing, while temperatures above 122℃ will cause premature aging of the material. Infrared non-contact temperature sensors with an accuracy of ±0.5℃ are installed at the top (5cm from the substrate surface), middle (center plane of the substrate), and bottom (5cm from the bottom of the substrate) of the main baking zone to form a vertical temperature difference monitoring network along the Z-axis. The sampling frequency is 2Hz, and the sensor data is uploaded to the MES system via a wireless module to calculate the temperature difference between the substrate surface and the bottom, i.e., the Z-axis temperature difference. This temperature difference directly reflects the thermal stress distribution in the thickness direction of the substrate. S42: Acquire 3000 sets of historical humidity data for the substrate, and combine them with the substrate storage time and thickness recorded by the MES system to conduct moisture absorption rate tests on substrates with different storage periods and thicknesses. The substrate samples are placed in a standard environment (25℃, 50%RH) for 24 hours, and the weight change is measured to calculate the moisture absorption rate. The standard is ≤0.01%, and exceeding this value will lead to an increase in the solder void rate. Establish a humidity-moisture absorption rate mapping table. The substrate thickness information includes 0.8mm, 1.0mm, and 1.6mm. In the mapping table, the moisture absorption rate of a 1.6mm substrate after 6 months of storage is ≥0.15%, reflecting the physical characteristics of thicker substrates, which have stronger moisture absorption due to their larger material volume and more internal pores. S43: Construct a power adjustment model with input parameters including Z-axis temperature difference, moisture absorption rate threshold (i.e., humidity-moisture absorption rate mapping table), and substrate thickness (0.8mm, 1.0mm, 1.6mm). Based on the Z-axis temperature difference and moisture absorption rate threshold, the output is the power adjustment range. Set power adjustment rules: when the Z-axis temperature difference is ≥5℃, the power decreases by 10%; when the moisture absorption rate is ≥0.02%, the power increases by 15%. Train a neural network model based on historical data using a three-layer feedforward network. The model input layer has 3 nodes corresponding to the Z-axis temperature difference, moisture absorption rate threshold, and substrate thickness encoding. The hidden layer has 4 nodes, which is determined to be the optimal structure after sensitivity analysis. Too few nodes will lead to underfitting, while too many nodes will cause overfitting. The output layer has 1 node to predict the power adjustment value. The learning rate is set to 0.01, and convergence is achieved after 5000 iterations. S44: Based on the output value of the power adjustment model, adjust the power of the heating module through the frequency converter. When the power adjustment range is ≥10%, adjust the wind speed of the dual-circulation hot air system and control it within ±0.2m / s to ensure uniform heat distribution. When adjusting the power, simultaneously adjust the injection flow rate of the nitrogen and argon mixed gas to maintain heat exchange balance and control it within ±10% to maintain the oxygen content ≤50ppm. S45: When the Z-axis temperature difference exceeds 8°C for three consecutive times, reduce the main baking power by 15% to control the temperature difference back to a safe range and trigger a temperature difference warning to prompt the operator to check the airflow distribution; after each batch of baking is completed, based on statistical sampling theory (confidence level 95%, error range ±0.005%), randomly select 3 substrates to test the surface moisture absorption rate. If the moisture absorption rate meets the standard, save the main baking parameters for this batch.

[0029] Furthermore, in S5, the cooling zone provides forced air cooling to the baked circuit board, with nitrogen continuously injected during the cooling process. A nitrogen sealing procedure is triggered before the substrate leaves the box, including: Multiple sets of axial-flow industrial fans with a diameter ≥400mm are installed at the top of the cooling area. The fans are arranged in a Z-shape, which reduces airflow interference compared to straight or grid arrangements. The spacing between each set of fans is 300mm, ensuring that the airflow overlap area between adjacent fans reaches 15-20cm, forming a vertically downward airflow channel that covers the entire area and ensures uniform airflow distribution. The wind speed is controlled by a frequency converter, with the initial wind speed set at 8m / s. A wind speed lower than 7m / s will lead to a longer cooling time, while a wind speed higher than 9m / s will cause the temperature gradient to exceed the limit. The corresponding temperature drop rate is ≤2℃ / min to prevent micro-cracks from forming between the miniLED chip and the substrate due to differences in cooling rates. If the temperature drop rate at the edge of the circuit board is detected to be too fast, the fan speed in the corresponding area is reduced to 5m / s based on feedback from the temperature sensor to maintain the overall temperature gradient consistency. 5m / s is the critical value for thermal balance in the edge area, which can maintain the overall cooling efficiency while reducing the phenomenon of overcooling at the edge.

[0030] When the oxygen content is >50ppm, increase the nitrogen injection flow rate to 120% of the design value, and at the same time reduce the fan speed to 60% to reduce the infiltration of outside air and reduce the gas exchange rate at the seal. If the oxygen content continues to be >50ppm for more than 30 seconds, start the vacuum pump for short-term vacuuming at a pumping speed ≥500L / min and a time ≤10 seconds. Then resume the nitrogen injection cycle. The 10-second upper limit prevents excessive negative pressure from causing substrate deformation. A double-layer isolation door is installed at the oven outlet. The inner layer is a nitrogen-sealed door with a leakage rate of <0.01L / min, and the outer layer is a mechanically locked door with a thickness of 3mm. A buffer chamber is located between the two doors. Before the substrate leaves the oven, a nitrogen sealing procedure is triggered, and nitrogen is injected into the buffer chamber to a positive pressure of 0.1kPa for 30 seconds. The substrate is then placed in an aluminum foil vacuum bag, and a vacuum pump is used to evacuate the bag to -0.08MPa. Nitrogen is then injected to a positive pressure of 0.05kPa inside the bag to form a protective layer.

[0031] Furthermore, in step S6, a baking-induced board explosion prediction model is established to predict whether the circuit board is at risk of exploding during the baking process, including: Temperature and humidity fluctuation rates in the preheating and main drying zones were calculated as key characteristics of board explosion risk. Weights were assigned based on substrate thickness. A 1.6mm substrate had a weight of 1.5 after 6 months of storage because thicker substrates have higher moisture absorption and longer storage time, leading to internal stress accumulation. A 0.8mm substrate had a weight of 1.0. Abnormal data from temperature and humidity sensors were removed using the Z-score method with a threshold of ±3 standard deviations. Linear interpolation was used to complete batches with missing storage cycle data.

[0032] A three-layer feedforward neural network was adopted. The input layer had three nodes: temperature and humidity (range 0-1), storage period (number of months), and thickness (0.8mm=1, 1.0mm=2, 1.6mm=3). The hidden layer had eight nodes, and the output layer had one node: the probability of board bursting. The Sigmoid activation function was used, and the feature space was divided by the substrate thickness and storage period. Historical baking records of the past 12 months were retrieved from the MES system, including labels for board bursting and non-blowout. Five-fold cross-validation was used, and the data was randomly divided into five groups. Four groups were used for training and one group for testing each time. The results were repeated five times and the average result was taken. The generalization error of the model was controlled to be ≤5%.

[0033] When the temperature in the main drying zone reaches 120±2℃ and the temperature difference along the Z-axis stabilizes, the internal stress of the material reaches its peak and its distribution becomes stable, triggering model prediction. When the temperature in the cooling zone drops to 80℃, a secondary prediction is made based on the oxygen content ≤50ppm and the humidity lock-in state during the cooling process. The probability of plate bursting is divided into three categories: low risk <1%, medium risk 1-5%, and high risk ≥5%. Low-risk batches can directly proceed to the next process, medium-risk batches have a 30% increased inspection ratio, and high-risk batches must trigger the adjustment mechanism.

[0034] Furthermore, if there is a risk of the baking plate bursting in S7, an adjustment mechanism is triggered to adjust the baking parameters based on historical baking parameter data in the MES system, including: Based on the bursting prediction model, the current prediction results are obtained, combined with the Z-axis temperature difference monitoring data mentioned in S4 and the humidity sensor data mentioned in S3; if any one of the following indicators reaches the threshold: bursting probability ≥5%, Z-axis temperature difference ≥8℃, or humidity fluctuation rate ≥15%RH / min, a bursting warning is triggered; if two or more indicators exceed the standard at the same time, such as bursting probability ≥5% and Z-axis temperature difference ≥8℃, an emergency stop procedure is immediately initiated, an emergency power-off command is sent to the main drying area to cut off the heating power supply, ensure that the temperature drops rapidly, and shut down the dual-circulation hot air system and nitrogen and argon injection valves to prevent the risk of high temperature or gas fluctuations from aggravating the problem.

[0035] Retrieve baking records for all batches within the past 6 months, including substrate thickness, storage period, temperature and humidity profiles, and board bursting event tags. Based on the current substrate thickness (0.8mm, 1.0mm, 1.6mm) and storage period (≤2 months, 2-6 months, 6-12 months), filter out low-risk batches by historical parameters including temperature, time, and humidity. If the current main baking temperature is 120±2℃, and the historical low-risk batch average temperature is 115±1℃, then reduce the main baking temperature to 115±1℃; if the current total preheating time does not exceed 20 minutes, but the historical low-risk batch preheating time is 30 minutes, then extend the preheating time to 30 minutes, which can reduce the internal moisture content of the 1.6mm thick substrate; if the current moisture absorption rate is 0.02%, and the historical low-risk batch moisture absorption rate is ≤0.01%, then increase the nitrogen injection amount by 10% to maintain the oxygen content ≤50ppm; After each parameter adjustment, the prediction model is rerun. If the risk drops to a low to medium level, the baking process is resumed. If the risk remains high, adjustments continue until it is safe or the process is completely terminated. All adjustment records include the original parameters, new parameters, reasons for adjustment, and effect evaluation, forming a closed-loop learning mechanism. Batches that trigger emergency stops are marked as special process batches and require subsequent electrical performance testing to ensure product quality.

[0036] Furthermore, a precision baking process optimization system for miniLED circuit boards is provided to implement the precision baking process optimization method for miniLED circuit boards described in any of the above claims. The precision baking process optimization system for miniLED circuit boards includes: a temperature zone control module, an inert gas injection control module, a preheating parameter adjustment module, a main baking power adjustment module, a cooling and sealing control module, a board explosion prediction and risk warning module, a parameter adjustment and emergency response module, and a data recording and full-process traceability module. Temperature zone control module: The preheating, main drying and cooling zones are spatially isolated by high-temperature resistant partitions with a thickness of ≥5mm. Each zone is equipped with an infrared non-contact temperature sensor and a digital humidity sensor, which wirelessly transmits data to the MES system. The top-down vertical airflow combined with the dual-circulation hot air system ensures temperature uniformity of ≤±0.8℃. Inert gas injection control module: Nitrogen to argon volume ratio 4:1, flow rate is regulated by mass flow controller, three-dimensional pressure sensor network monitors and adjusts the number of injection orifice openings and flow rate, maintaining positive pressure of 0.2kPa±0.02kPa. When the area humidity is >30%RH, increase gas flow rate ≤15%; when the pressure is abnormal, adjust injection orifice opening and flow rate ±10%. Preheating parameter adjustment module: The vertical humidity sensor monitors the initial humidity of the environment. When the humidity is ≥40%RH, the preheating time is extended and the heating rate is ≤2℃ / min. If the limit is exceeded, the power is reduced by 10%-20%. When the humidity is ≥50%RH and still ≥45%RH after 30 minutes of preheating, a high humidity warning is triggered and preheating is suspended. When the temperature in the preheating zone exceeds 62℃ for 3 consecutive times, the power is reduced by 20% and an alarm is triggered. Main drying power adjustment module: top, middle and bottom infrared sensors sample at 2Hz, calculate the Z-axis temperature difference, establish a moisture absorption rate threshold database based on substrate thickness and storage cycle, input Z-axis temperature difference, moisture absorption rate and substrate thickness, output power adjustment value, and simultaneously adjust hot air speed and gas flow rate during power adjustment to maintain oxygen content ≤50ppm; Cooling and sealing control module: Z-shaped axial flow fan layout, variable frequency speed regulation. When the oxygen content is >50ppm, the nitrogen flow rate is increased to 120%, the fan speed is reduced to 60%, and the vacuum pump is started to evacuate the vacuum if the oxygen content exceeds the limit for 30 seconds. The oven outlet has a double-layer isolation door. Before the substrate leaves the oven, nitrogen is injected into the buffer chamber to a positive pressure of 0.1kPa, and then vacuum-packed to -0.08MPa and nitrogen is injected to 0.05kPa. Module for predicting and warning of board explosion: Calculates the temperature and humidity fluctuation rate of the preheating and main drying areas, divides risk indicators based on substrate thickness weight, adopts a three-layer feedforward network, and has a 5-fold cross-validation error of ≤5%. The first prediction is made after the temperature of the main drying area stabilizes, and the second prediction is made when the temperature of the cooling area drops to 80℃. Risk is classified. Parameter adjustment and emergency response module: A single indicator triggers an early warning; if two or more indicators exceed the standard, an emergency stop procedure is initiated, low-risk batch data from the past 6 months are retrieved, the main baking temperature, preheating time, and nitrogen flow rate are adjusted, and baking is restarted after adjustment. The plate bursting prediction model continuously monitors the risk value. Data recording and full-process traceability module: The final baking parameters, including temperature, time and humidity, are written into the QR code of the substrate, and the full-process traceability is realized through blockchain technology.

[0037] refer to Figure 2The diagram shown is a module diagram of a precision baking process optimization system for miniLED circuit boards.

[0038] Example 2 A batch of miniLED circuit boards with a thickness of 0.8mm and a storage time of 2 months need to be baked.

[0039] Oven preparation and gas settings: For a 0.8mm thin substrate, only 50% of the gas injection holes are opened, with 32 holes in an 8×8 matrix on each wall; nitrogen and argon are mixed and injected at a volume ratio of 4:1, and precisely regulated by a mass flow controller; the three-dimensional pressure monitoring network (top, middle, bottom) shows that the pressure is stable at 0.2kPa, with the error controlled within ±0.02kPa.

[0040] Preheating operation: The initial temperature is set to 60℃, and the initial ambient humidity is measured to be 40%RH. When the initial ambient humidity is ≥40%RH, the preheating time needs to be extended. Since the humidity is exactly 40%RH, it is extended by 5 minutes (5 minutes for every 10%RH increase). The heating rate is controlled at 1.2℃ / min (lower than the specified upper limit of 2℃ / min) to ensure that the temperature difference between the inner and outer layers of the substrate does not exceed 5℃. The total preheating time is 15 minutes (10 minutes initially + 5 minutes extension), which does not reach the upper limit of 30 minutes, so there is no need to trigger the high humidity warning. Temperature monitoring shows that the temperature of the preheating zone is stable in the range of 59-61℃, which does not exceed 62℃, so there is no need to reduce the heating power.

[0041] Main drying stage parameter adjustment: The main drying zone temperature is set to 120℃, the Z-axis temperature difference monitoring shows 5.2℃, the top is 62℃, and the bottom is 56.8℃; when the Z-axis temperature difference is ≥5℃, the power needs to be reduced by 10%, and the heating module power needs to be adjusted; the humidity-hygroscopic rate mapping table shows that the moisture absorption rate of the 0.8mm substrate after 2 months of storage is 0.02%, which reaches the threshold, and the power needs to be increased by 15%; the power adjustment model combines the two parameters and finally determines a net increase of 5% (-10%+15%); since the power adjustment range (5%) <10%, it is not necessary to adjust the wind speed of the dual circulation hot air system; the gas flow rate is synchronously adjusted by +5% (within ±10%) to ensure that the oxygen content is maintained at 48ppm, which is below the 50ppm threshold.

[0042] Cooling phase control: After cooling begins, the initial fan speed of the axial fan is set to 8 m / s, and the temperature drop rate is 1.8℃ / min, which is below the upper limit of 2℃ / min; the edge temperature sensor shows that the temperature difference between the edge area and the center area is 2.5℃, which does not exceed the 3℃ threshold, so there is no need to reduce the fan speed in the edge area; the oxygen content monitoring remains stable in the range of 45-48 ppm, which does not exceed 50 ppm, so there is no need to start the nitrogen flow increase or vacuum pump program; the cooling process lasts for 25 minutes, from 120℃ to room temperature of 25℃.

[0043] Board explosion risk prediction and handling: When the temperature in the main drying zone reaches 120℃ and the Z-axis temperature difference stabilizes at 5.2℃, the board explosion prediction model is triggered; Model input parameters: temperature and humidity comprehensive index 0.35, storage period 2 months, substrate thickness 0.8mm; The prediction result shows that the probability of board explosion is 0.8%, which is lower than the 1% threshold and belongs to the low risk category; When the temperature in the cooling zone drops to 80℃, a second prediction is performed, and the result is 0.6%, which is confirmed as low risk; No adjustment mechanism needs to be triggered, and the process can proceed directly to the next step.

[0044] Final processing and data recording: Before the substrate leaves the box, a nitrogen sealing procedure is triggered: nitrogen is injected into the buffer chamber to a positive pressure of 0.1 kPa for 30 seconds; the substrate is placed in an aluminum foil vacuum bag, evacuated to -0.08 MPa, and then nitrogen is injected to a positive pressure of 0.05 kPa inside the bag; final baking parameters: preheating temperature 60℃, time 15 minutes, main baking temperature 120℃, time 25 minutes, humidity 40%RH, which are written into the substrate's QR code. The entire process data is uploaded through blockchain technology to generate a unique traceability ID: BATCH20231105-08MM-2M.

[0045] Quality verification: Three substrates were randomly selected from this batch for moisture absorption rate testing; the test results showed that the average moisture absorption rate was 0.008%, which is lower than the 0.01% standard and meets the quality requirements; all parameters were saved to the MES system parameter library as recommended settings for baking under similar conditions (0.8mm thickness, ≤2-month storage period).

[0046] Example 3 A batch of miniLED circuit boards with a thickness of 1.6mm and a storage time of 6 months needs to be baked.

[0047] Preheating stage: The initial temperature is set to 60℃. Since the initial ambient humidity is 50%RH, the preheating time needs to be extended to 30 minutes. The heating rate is limited to within 2℃ / min to prevent stress cracks from forming inside the substrate.

[0048] Main baking stage: The target temperature is 120℃. The temperature difference along the Z-axis is monitored by a sensor. If the temperature difference exceeds 8℃ three times, the power is reduced by 15%. According to the humidity-moisture absorption rate mapping table, the moisture absorption rate of this batch of substrates is 0.15%, which is higher than the threshold of 0.02%. Therefore, the power output is increased by 15%.

[0049] Cooling phase: During the cooling process, if the edge area is detected to be cooling too fast, reduce the fan speed in that area to 5m / s to maintain the consistency of the overall temperature gradient; when the oxygen content exceeds 50ppm, increase the nitrogen injection amount to 120% of the design value, and start the vacuum pump to evacuate if necessary.

[0050] Plate bursting prediction: After the temperature in the main drying zone stabilizes, the model predicts a plate bursting probability of 3%, which is a medium-risk level and requires increasing the inspection ratio; after confirming that the prediction is correct again during the cooling stage, the subsequent processes can continue.

[0051] Parameter adjustment: If the prediction results show a high risk, adjust the baking parameters based on historical data, such as lowering the main baking temperature to 115℃ or extending the preheating time to 30 minutes, until the risk is reduced to an acceptable level.

[0052] Example 4 This embodiment provides a computer device, including 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 the steps in the miniLED circuit board precision baking process optimization method as described in Embodiment 1 above.

[0053] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of hardware embodiments, software embodiments, or embodiments combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage and optical storage) containing computer-usable program code.

[0054] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0055] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

[0056] This invention provides an optimized method and system for the precision baking process of miniLED circuit boards. The oven is divided into three zones: preheating, main baking, and cooling. A dual-circulation hot air system and a top vertical airflow are used, combined with infrared and digital humidity sensors to achieve independent temperature and humidity control. A nitrogen-argon mixed gas is injected, and the opening and closing of the injection holes are adjusted based on the substrate thickness. A three-dimensional pressure sensor network is used to maintain positive pressure and oxygen content inside the oven. During the preheating stage, the heating time is extended according to the ambient humidity, and the heating rate is controlled by gradient. A humidity-hygroscopic rate mapping table and a three-layer neural network model are constructed. Power output is adjusted by combining the Z-axis temperature difference. The cooling zone adopts a forced air cooling and nitrogen sealing program, and a board explosion prediction model is established. A parameter adjustment mechanism is triggered by historical data. Baking parameters are written into the substrate QR code through blockchain technology to achieve full-process traceability.

Claims

1. A mini LED circuit board precision baking process optimization method, characterized in that, The method comprises the following steps: S1: The mini LED circuit board oven is divided into a preheating zone, a main baking zone and a cooling zone, and each zone is independently configured with a temperature and humidity control system, a double-circulation hot air device is arranged in the oven and a vertical air flow is formed from top to bottom; S2: Nitrogen and argon mixed gas is injected into the oven, the opening and closing state of the gas injection hole is adjusted according to the substrate thickness parameter, and the gas flow is adjusted according to the feedback of the pressure sensor and the humidity sensor; S3: The initial temperature parameter, the heating time parameter and the temperature rising rate parameter are set in the preheating zone, and the temperature rising time and the temperature rising rate of the preheating zone are adjusted based on the data collected by the environmental humidity sensor; S4: The target temperature parameter is set in the main baking zone, the temperature distribution difference of the multi-layer HDI substrate in the Z-axis direction is monitored through the temperature sensor network, the power control model is established based on the substrate humidity-hygroscopic rate mapping table and the Z-axis temperature difference data constructed by the MES system, and the heating power of the main baking zone is adjusted; S5: The circuit board after baking is subjected to forced air cooling treatment in the cooling zone, the temperature drop gradient is controlled, nitrogen is continuously introduced during the cooling process, a humidity isolation device is arranged at the oven outlet, and a nitrogen protection program is executed before the substrate leaves the oven; S6: A substrate explosion risk prediction model is constructed based on the temperature and humidity parameters, the substrate storage period parameters and the substrate thickness parameters, and the substrate explosion risk state in the baking process is judged according to the prediction result; S7: When there is a risk of substrate explosion, a parameter adjustment mechanism is triggered, and the current baking process parameters are adjusted based on the historical baking parameter database stored in the MES system; S8: The finally determined baking temperature parameter, time parameter and humidity parameter are written into the two-dimensional code of the substrate, and the whole process traceability of the baking process parameters is realized through the blockchain technology.

2. The mini LED circuit board precision baking process optimization method of claim 1, wherein, In the S1, the mini LED circuit board oven is divided into a preheating zone, a main baking zone and a cooling zone, and a double-circulation hot air device is arranged in the oven to form a vertical air flow from top to bottom, which comprises: The internal space of the oven is divided into a preheating zone, a main baking zone and a cooling zone by high-temperature-resistant partitions, and each zone is configured with an independent temperature and humidity control system, wherein the temperature and humidity control system comprises an infrared non-contact temperature sensor and a digital humidity sensor, and each sensor is connected with the MES system through a wireless communication module to establish data connection; A double-circulation hot air device is arranged at the top of the oven, which comprises an inner circulation system and an outer circulation system, the inner circulation system is configured to provide a vertical air flow from the top of the oven to the surface of the substrate, the outer circulation system is configured with an air inlet channel and an air outlet channel on the side wall of the oven to form an external air flow circulation path, and the hot air outlet is configured with a honeycomb-shaped flow uniform structure; the air outlets of the inner circulation system are uniformly distributed on the top of the oven to form a directional air flow field perpendicular to the surface of the substrate, and the air outlets are arranged at a predetermined interval.

3. The mini LED circuit board precision baking process optimization method of claim 1, wherein, The S2 further comprises: Nitrogen and argon delivery channels are arranged in the gas inlet area of the oven, the two channels are connected with corresponding gas storage devices respectively, the mass flow control device is used to adjust the flow ratio of each gas, the mixed gas with a preset volume ratio is formed, and the opening number and distribution state of the gas injection hole are configured according to the substrate thickness parameter; A pressure sensor array is arranged at different height positions in the interior space of the oven to construct a spatially distributed pressure monitoring system, preset target pressure parameters are set, and the opening and closing state of the gas injection hole and the gas flow parameter are adjusted based on the monitored pressure data, while the gas flow output of the corresponding area is adaptively adjusted according to the regional humidity monitoring data.

4. The mini LED circuit board precision baking process optimization method of claim 1, wherein, The S3 includes adjusting the preheating time and the preheating rate of the preheating area based on the data collected by the environmental humidity sensor, including: The initial temperature threshold and the preheating rate threshold of the preheating area are set, the output power of the heating module is adjusted by the variable frequency power control unit, the temperature change rate is monitored, and when the preset preheating time threshold is reached, the main baking area preparation instruction is triggered; Digital humidity sensor units are arranged at different height positions in the preheating area to construct a vertical humidity monitoring system, and the preheating time parameter and the preheating rate parameter are adjusted according to the initial humidity parameter and the humidity change rate parameter of the environment; When the initial humidity of the environment exceeds the first preset threshold, the preheating time is extended, and for every preset percentage increase in humidity, the preheating time is correspondingly extended, and the total preheating time does not exceed the preset upper limit, while the preheating rate is reduced to a preset safety value; When the initial humidity of the environment exceeds the second preset threshold and the preheating time has reached the upper limit, and the humidity still exceeds the third preset threshold, the high humidity early warning mechanism is triggered, the preheating program is suspended and the inspection process is started; When the monitored preheating rate exceeds the preset rate threshold, the heating power is reduced by a preset percentage, and when the preheating area temperature continuously exceeds the preset temperature threshold for a set number of times, the heating power is reduced by a fixed percentage and the alarm mechanism is triggered, prompting the operator to check the equipment status.

5. The mini LED circuit board precision baking process optimization method of claim 1, wherein, The S4 includes constructing a substrate humidity-hygroscopic rate mapping table and a Z-axis temperature difference data power control model based on the MES system, adjusting the heating power of the main baking area, including: S41: Set the target temperature parameter of the main baking area, configure infrared non-contact temperature sensor units at different height positions along the Z-axis direction in the main baking area to construct a vertical temperature difference monitoring network, collect temperature data at a preset sampling frequency and transmit it to the MES system, and calculate the Z-axis temperature difference between the substrate surface and the bottom; S42: Integrate the historical humidity data of the substrate, the storage period parameter and the thickness parameter, test the hygroscopic characteristics of substrates of different storage periods and thickness specifications, and establish a humidity-hygroscopic rate mapping relationship table; S43: Construct a power control model, the input parameters of the model include Z-axis temperature difference data, hygroscopic rate threshold determined based on the humidity-hygroscopic rate mapping relationship table, and substrate thickness parameter, and the output is power adjustment amplitude value, preset power adjustment rule, when the Z-axis temperature difference exceeds the first threshold, the power is reduced by the first percentage, when the hygroscopic rate exceeds the second threshold, the power is increased by the second percentage, the power control model adopts a neural network prediction structure trained based on historical data; S44: According to the adjustment value output by the power control model, adjust the heating module power output through the variable frequency control unit, when the power adjustment amplitude exceeds the preset percentage threshold, the air speed parameter of the double-circulation hot air system and the inert gas injection flow parameter are adjusted synchronously; S45: When the Z-axis temperature difference is continuously monitored to exceed the third threshold value for a preset number of times, the main baking power is reduced by a fixed ratio, and a temperature difference warning mechanism is triggered. After each batch of baking process is completed, the treated substrate is sampled for moisture absorption rate detection. When the detection result meets the preset standard, the current main baking parameters are saved to the process database.

6. The method of claim 1, wherein the mini LED circuit board precision baking process optimization method is characterized by, The S5 implements forced air cooling treatment on the baked circuit board in the cooling area. Nitrogen is continuously introduced during the cooling process. Nitrogen protection program is executed before the substrate is taken out of the oven, including: A plurality of axial flow fan devices are arranged at the top of the cooling area. A global airflow channel vertically downward is formed by using a spatial distribution layout. The fan speed parameters are adjusted by a frequency conversion control unit. The fan speeds of different regions are adjusted based on temperature distribution monitoring data; An oxygen content monitoring unit is provided. When it is detected that the oxygen content exceeds a preset safety threshold, the nitrogen injection flow is increased in proportion and the fan speed is adjusted. If the oxygen content continuously exceeds the safety threshold for a preset time, the vacuum treatment device is started for short-time vacuum operation, and then the nitrogen injection is restored; A double-layer isolation door structure is arranged at the outlet of the oven, including an inner gas sealing door and an outer mechanical locking door. A buffer cavity is formed between the two doors. Nitrogen sealing process is executed before the substrate is taken out of the oven. Nitrogen is injected into the buffer cavity to a preset positive pressure value and maintained for a preset time. Then the substrate is packaged in a gas-tight packaging material. Vacuum and nitrogen filling processes are performed to maintain the internal pressure of the package at a preset positive pressure state.

7. The mini LED circuit board precision baking process optimization method of claim 1, wherein, The S6 further includes: The temperature fluctuation parameters and humidity fluctuation parameters of the preheating area and the main baking area are calculated as key characteristic parameters for board explosion risk assessment. The characteristic weights are assigned according to the substrate thickness parameters. The abnormal values of the temperature and humidity sensor data are identified and processed. The missing data of the storage period is interpolated and completed; A multi-layer neural network prediction model is constructed. The input features of the model include environmental temperature and humidity parameters, substrate storage period parameters and substrate thickness parameters. The output is the board explosion probability value. The feature space is divided based on the substrate characteristic parameters. The model is trained using the MES system historical baking record data set. The data set contains labeled board explosion and non-explosion samples. The model prediction performance is verified by cross-validation method. When the temperature of the main baking area reaches the preset process temperature and the Z-axis temperature difference is stable, the first board explosion risk prediction is performed. When the temperature in the cooling area drops to a preset threshold, the second risk prediction is performed based on the cooling environment parameters and the humidity control state. According to the predicted board explosion probability value, the risk level is divided into low risk, medium risk and high risk according to the preset threshold interval.

8. The mini LED circuit board precision baking process optimization method of claim 1, wherein, The S7 further includes: Based on the output results of the board explosion prediction model, the comprehensive risk assessment is performed combining the Z-axis temperature difference monitoring data and the environmental humidity monitoring data. When any risk indicator exceeds the preset threshold, the board explosion warning mechanism is triggered. When two or more risk indicators exceed the corresponding threshold at the same time, the emergency stop program is immediately started. The main baking area heating power is cut off and the hot air circulation system and the inert gas injection device are closed. Retrieving historical baking record data within a preset time period from a process database, the data containing substrate thickness parameters, storage period parameters, temperature and humidity parameter curves, and board explosion event markers, screening a set of historical process parameters with low risk markers according to the thickness and storage period characteristics of the current processing substrate; Comparing the differences between the current process parameters and the screened low-risk historical parameter set, when the current main baking temperature is higher than the historical low-risk temperature threshold, adjusting the main baking temperature to the historical low-risk temperature range, when the current preheating time is shorter than the historical low-risk preheating time threshold, extending the preheating time to the historical low-risk time range, when the current moisture absorption rate is higher than the historical low-risk moisture absorption rate threshold, increasing the inert gas injection amount to maintain the environmental oxygen content within a preset safe range. 9.A mini LED circuit board precision baking process optimization system, characterized in that, A miniLED circuit board precision baking process optimization method as claimed in any one of claims 1-8, the miniLED circuit board precision baking process optimization system comprising: a temperature zone partition control module, an inert gas injection control module, a preheating parameter adjustment module, a main baking power adjustment module, a cooling and storage control module, a board explosion prediction and risk warning module, a parameter adjustment and emergency response module, and a data recording and full-process tracing module: The temperature zone partition control module: the baking oven space is divided into a preheating zone, a main baking zone and a cooling zone by high-temperature resistant partitions, each zone is equipped with non-contact temperature sensing units and digital humidity sensing units, and the monitoring data is transmitted to the MES system through a wireless communication network, and a top vertical air flow and a double-circulation hot air cooperative structure are used; The inert gas injection control module: nitrogen and argon are mixed according to a preset proportion, the gas flow is adjusted through a mass flow control device, the opening and closing state and flow parameters of the gas injection holes are adjusted based on a three-dimensional distributed pressure sensing network to maintain a positive pressure environment inside the baking oven, and the gas flow is adjusted adaptively according to the change in the humidity of the region; The preheating parameter adjustment module: the initial humidity parameters of the environment are monitored through vertically distributed humidity sensing units, the preheating time parameters and the temperature rise rate parameters are adjusted based on humidity thresholds, a temperature and power overrun protection mechanism is set, and the prewarning and suspension mechanism is triggered when the humidity and temperature parameters exceed the preset threshold combination; The main baking power adjustment module: temperature sensing units are arranged at different heights in the baking oven, the temperature difference in the Z-axis direction of the substrate is calculated, a moisture absorption rate threshold database is constructed based on the substrate thickness parameters and the storage period parameters, a power adjustment value is output through a prediction model with multiple parameter inputs, and the air flow speed parameters of the hot air system and the inert gas flow parameters are adjusted synchronously during the power adjustment process; The cooling and storage control module: uniform cooling is realized through a variable frequency controlled air cooling system, the nitrogen flow and the air cooling intensity are adjusted according to the oxygen content monitoring data, the vacuum treatment device is started when the oxygen content continues to exceed the standard, a double-layer isolation door structure is arranged at the outlet of the baking oven, and nitrogen injection and vacuum packaging processes are performed in the buffer cavity before the substrate exits the oven. The explosion plate prediction and risk early warning module calculates the temperature and humidity fluctuation parameters of the preheating zone and the main baking zone, combines with the substrate characteristic parameters to evaluate the risk index, adopts a machine learning model to predict the risk probability, sets a double risk evaluation mechanism for the main baking stage and the cooling stage, and divides the risk level according to the risk probability value. The parameter adjustment and emergency response module triggers a hierarchical response mechanism based on multi-source risk monitoring data, starts an early warning process when a single risk index exceeds the standard, executes an emergency stop program when multiple risk indexes exceed the standard at the same time, calls a historical low-risk process parameter database, adjusts the current process parameters, and restarts the baking process. The data recording and full-process tracing module writes the finally determined process parameters into the substrate two-dimensional code, and realizes the tamper-proof recording and full-process tracing of process data through the blockchain technology.

10. A computer device comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, The processor executes the computer program to realize the steps of the mini LED circuit board precision baking process optimization method in any one of claims 1-8.

Images