Humidity monitoring and moisture-proof storage management method and device for SMD reel material

Abstract

The application provides a kind of SMD reel material humidity monitoring and moisture-proof storage management method, comprising: deployment humidity sensor data acquisition is extracted by RFID material humidity sensitive grade and cumulative exposure time output attribute record;Filter data to construct three-dimensional humidity field by radial basis function interpolation Mark abnormality, according to local humidity to determine exposure coefficient to update cumulative time to calculate remaining effective time to execute material classification to generate early warning, according to the highest sensitive level in the area to determine target humidity based on door opening event to calculate humidity intrusion compensation to generate control instruction;PID controller is adjusted to regulate dehumidification or proportional valve to regulate inert gas flow to make humidity approach target value, material state and early warning are synchronized to warehouse management system and manufacturing execution system to trigger linkage, alarm event is classified and pushed. The application replaces discrete point monitoring with three-dimensional humidity field perception, replaces manual ledger with exposure coefficient dynamic tracking, and performs humidity intrusion compensation on door opening driven by sensitive level control.

Description

Technical Field

[0001] This application relates to the fields of electronic component warehousing environment control and Internet of Things technology, specifically to a method and device for humidity monitoring and moisture-proof storage management of SMD reel materials. Background Technology

[0002] Humidity-sensitive components in surface mount devices absorb moisture from the air when exposed to the workshop environment. During subsequent reflow soldering, the moisture vaporizes and expands due to heat, which may cause the package to crack, delaminate, or the internal interconnects to break. To address this, industry standards have established a humidity sensitivity rating system, requiring high-sensitivity components to be mounted and soldered within a specified time after opening or to be restored in a low-humidity environment; otherwise, they must be baked back into service before use.

[0003] Existing SMD material storage systems primarily use distributed thermometers and hygrometers for humidity monitoring. This sparse density of monitoring points fails to capture the micro-environmental humidity gradients created by factors such as air conditioning vent layout, shelving obstructions, and varying door opening frequencies. When humidity levels exceed limits in certain areas, these blind spots allow materials in those areas to be exposed undetected for extended periods. Current humidity-sensitive material timeliness management relies mainly on manual recording of opening times and paper or spreadsheet records, leading to significant errors in timeliness calculations and a high rate of uncontrolled tracking of highly sensitive materials. This results in materials that have already deteriorated being delivered to the production line, causing welding quality defects.

[0004] Existing moisture-proof storage equipment mostly uses on / off control with fixed humidity setpoints or simple proportional-integral-derivative control, failing to dynamically adjust the target humidity based on the humidity sensitivity distribution of the currently stored materials and the frequency of door openings. This results in insufficient moisture protection when highly sensitive materials are mixed in, and wasted energy when only low-sensitivity materials are stored. Furthermore, the humidity status of the materials operates independently from the production scheduling system. Materials exceeding their shelf life must be manually discovered, locked, and determined for drying or disposal by warehouse staff, leading to long response times and insufficient automation. Summary of the Invention

[0005] To address the problems of insufficient spatial resolution in humidity monitoring, large time-tracking errors in humidity sensitivity levels, and separation of moisture control and production scheduling in existing technologies, this application provides a method and device for humidity monitoring and moisture-proof storage management of SMD reel materials. This method utilizes a high-density distributed sensor array and three-dimensional humidity field modeling, dynamic exposure coefficient time-tracking based on micro-environment humidity, and material state-driven moisture control and production linkage. It replaces discrete point monitoring with continuous three-dimensional humidity field spatial sensing, replaces manual record-keeping with dynamic exposure coefficient accumulation, replaces fixed-value on / off control with adaptive target humidity setting and door opening compensation, and replaces manual handling with automatic material state linkage.

[0006] To solve at least one of the above problems, this application provides the following technical solution: In a first aspect, this application provides a method for humidity monitoring and moisture-proof storage management of SMD reel materials, comprising: deploying temperature and humidity sensor nodes in the storage area at a preset spatial density, collecting temperature and humidity data via a wireless network and aggregating it into a raw environmental data stream; receiving material tags read by an RFID reader to extract humidity sensitivity level and packaging status, and outputting material attribute records based on historical cumulative exposure time; constructing a three-dimensional humidity field based on the sensor spatial coordinates and radial basis function interpolation after applying an exponential weighted moving average filter to the raw environmental data stream, and marking anomalies through spatial gradient; and reading the humidity sensitivity level in the material attribute records and looking up a time threshold table to determine the upper limit of exposure time. After determining the exposure coefficient and updating the cumulative time based on the local humidity interpolation of the storage location, the remaining effective time is calculated. Based on the proportion of effective time, the material status classification is executed to generate an early warning. The target humidity is determined based on the highest sensitivity level of the materials in the area, and the humidity intrusion compensation is calculated based on the door opening event to generate a control command. After parsing the control command, the power of the dehumidifier unit is adjusted through the proportional integral derivative controller or the inert gas flow rate is adjusted through the proportional valve to make the humidity of the storage area approach the target humidity. The material status classification result and the early warning are synchronized to the warehouse management system and the manufacturing execution system through the communication interface to trigger priority outbound or material locking linkage. Alarm events are pushed in a graded manner and manual operations are recorded in the audit log.

[0007] Furthermore, it also includes: installing the temperature and humidity sensor nodes under the shelf panels with the probes facing the material space; each node synchronizes with the wireless network time base beacon and collects temperature and relative humidity values ​​according to a preset sampling period; the data includes node identifier, collection time, and temperature and humidity values, which are relayed hop by hop to the edge gateway via the wireless network according to a preset time delay limit; the RFID reader at the inbound workstation performs an inventory check on the material tags at a preset distance, extracts the material code and batch number from the tag user storage area, and queries the material database to obtain the historical cumulative time and the humidity sensitivity level; when the material is first put into storage and is sealed, the cumulative time is set to zero; when the material is returned to storage, the historical cumulative time is used as the starting value to continue accumulating the time.

[0008] Furthermore, it also includes: applying exponentially weighted moving average filtering to the temperature and humidity time series of each sensor node in the original environmental data stream, with a preset smoothing factor to suppress transient noise; using the spatial coordinates of each node and the filtered humidity value as scatter points and employing Gaussian kernel radial basis function interpolation to construct a continuous humidity field function covering the entire storage area; obtaining the humidity gradient vector field by taking the first-order partial derivatives of the continuous humidity field function along three spatial directions; marking the spatial locations where the gradient magnitude exceeds a preset threshold as humidity anomaly areas; using the coordinates of the material's location as query points to perform interpolation to obtain local humidity values ​​and thereby determine the exposure coefficient and update the cumulative exposure time.

[0009] Furthermore, it also includes: extracting the humidity sensitivity level and opening mark from the material attribute record; querying the timeliness threshold table based on the humidity sensitivity level to obtain the upper limit of workshop exposure time; setting the cumulative exposure time to zero when the material is first put into storage and is still sealed; reading the historical cumulative time from the database and continuing to count when the material is returned to storage; accumulating the cumulative exposure time by multiplying the exposure coefficient by the calculation cycle increment every second; subtracting the updated cumulative exposure time from the upper limit of workshop exposure time to obtain the remaining effective time; and comparing the ratio of the remaining effective time to the total timeliness with a preset ratio threshold condition to determine the current state of the material.

[0010] Furthermore, it also includes: calculating the ratio of the remaining effective time to the upper limit of the workshop exposure time to obtain the remaining effective time ratio, comparing it with the first and second preset ratio threshold conditions in sequence; when it is higher than the first threshold condition, it is normal; when it is between the two, it triggers a warning and a priority outbound reminder; when it is between the second threshold condition and zero, it triggers an emergency outbound or dry cabinet transfer instruction; statistically analyzes the distribution of humidity sensitivity levels of materials in each moisture-proof storage cabinet according to a preset grid and takes the highest value as the highest sensitivity level in the area; queries the corresponding target humidity setting value from the preset mapping table; calibrates the door opening time by detecting the level jump of the cabinet door status sensor and records the duration; and calculates the humidity intrusion compensation amount according to the exponential decay model and adds it to the target humidity setting value.

[0011] Furthermore, it also includes: parsing the target humidity value and control mode field in the control command; adjusting the output power of the semiconductor dehumidifier unit through the proportional-integral-derivative controller in the drying cabinet mode, with the proportional gain, integral time and derivative time taking preset set values; adjusting the gas flow rate through the proportional valve in the inert gas cabinet mode and controlling the oxygen content not to exceed the preset concentration threshold; reading the maximum modulus of the humidity field spatial gradient in the current cabinet and adjusting the internal circulation fan speed according to the preset mapping relationship to promote the uniformity of humidity in the cabinet; closing the inert gas inlet valve and increasing the fan speed to the preset ratio when the cabinet door is opened, thereby forming an airflow barrier at the door.

[0012] Furthermore, it also includes: encoding the material status classification results into status synchronization data packets according to a preset synchronization cycle and pushing them to the warehouse management system and the manufacturing execution system via a descriptive status transfer interface or a unified process control architecture interface; prioritizing the production of materials in the warning status by the manufacturing execution system and marking them with a priority outbound tag by the warehouse management system; classifying the alarm events generated in each stage into levels according to their severity, with material timeout lockout and equipment failure classified as the highest level and pushed to the responsible person's terminal in real time, requiring a response within a preset time limit; material nearing its expiration date and abnormal humidity classified as the middle level and pushed to the dashboard display; and material warnings and energy-saving reminders classified as the lowest level and only written into the log for summary in the periodic report.

[0013] Secondly, this application provides a humidity monitoring and moisture-proof storage management device for SMD reel materials, comprising: a multi-point sensing module, used to deploy temperature and humidity sensor nodes in the storage area at a preset spatial density, collect temperature and humidity data through a wireless network to aggregate into a raw environmental data stream, receive RFID readers to read material tags to extract material attributes, and output material attribute records in combination with historical cumulative exposure time; a humidity field modeling and anomaly detection module, used to construct a three-dimensional humidity field based on the sensor spatial coordinates and radial basis function interpolation after filtering the raw environmental data stream with an exponentially weighted moving average, and mark anomalies through spatial gradient; and an exposure time tracking module, used to read the humidity sensitivity level in the material attribute records, query the time threshold table to determine the upper limit of exposure time, and determine the upper limit of exposure time based on the location of the goods. The system calculates the remaining effective time after interpolating humidity to determine the exposure coefficient and updating the cumulative exposure time. The status grading and control decision module generates early warnings based on the proportion of effective time, determines the target humidity based on the highest sensitivity level of materials in each moisture-proof storage area, and calculates humidity intrusion compensation based on door opening events to generate control commands. The dehumidification and inert gas execution module parses the control commands and adjusts the dehumidifier power via a proportional-integral-derivative controller or adjusts the inert gas flow rate via a proportional valve to bring the humidity in the storage area closer to the target value. The system linkage and anomaly management module synchronizes the material status grading results and the early warnings to the warehouse management system and manufacturing execution system via a communication interface to trigger linkage actions, pushes alarm events in a graded manner, and records manual operations to the audit log.

[0014] Thirdly, this application provides 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 humidity monitoring and moisture-proof storage management method for SMD reel materials.

[0015] Fourthly, this application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the aforementioned method for humidity monitoring and moisture-proof storage management of SMD reel materials.

[0016] Fifthly, this application provides a computer program product, including a computer program / instruction, which, when executed by a processor, implements the steps of the aforementioned method for humidity monitoring and moisture-proof storage management of SMD reel materials.

[0017] As can be seen from the above technical solution, this application provides a method and device for humidity monitoring and moisture-proof storage management of SMD reel materials. It uses a high-density distributed sensor array and three-dimensional radial basis function interpolation humidity field modeling to continuously sense the humidity of the microenvironment and quickly locate anomalies. It uses dynamic exposure coefficient time tracking based on local humidity of the storage location to replace manual ledger recording for precise time-sensitive control of materials. It uses dynamic setting of target humidity driven by the distribution of material humidity sensitivity level and feedforward compensation for door opening events to balance the accuracy of moisture-proof control and energy consumption. It uses a warehouse management system and manufacturing execution system driven by material status classification to automatically link and quickly lock and automate the disposal process of expired materials. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below.

[0019] Figure 1 This is a flowchart illustrating the method for humidity monitoring and moisture-proof storage management of SMD reel materials in the embodiments of this application. Figure 2 This is a schematic diagram of distributed sensor deployment and data aggregation in an embodiment of this application; Figure 3 This is a schematic diagram illustrating the process of three-dimensional humidity field modeling and exposure coefficient tracking in the embodiments of this application; Figure 4 This is a flowchart illustrating the material state classification and moisture control decision-making process in the embodiments of this application. Figure 5 This is a schematic diagram illustrating the dehumidification and nitrogen filling control and system linkage in the embodiments of this application; Figure 6 This is a schematic diagram illustrating the alarm hierarchy push and manual intervention management process in the embodiments of this application; Figure 7 This is a schematic diagram of the humidity monitoring and moisture-proof storage management device for SMD reel materials in the embodiments of this application. Detailed Implementation

[0020] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are some embodiments of this application, but not all embodiments.

[0021] Considering the shortcomings of existing SMD material warehousing in humidity monitoring, timely tracking of humidity sensitivity levels, and moisture-proof control, this application provides a method for humidity monitoring and moisture-proof storage management of SMD reel materials. This method utilizes a high-density distributed sensor array and three-dimensional humidity field modeling, timely tracking of dynamic exposure coefficients based on microenvironment humidity, adaptive moisture-proof control driven by the distribution of material humidity sensitivity levels, and automatic linkage between material status and the production system. It replaces discrete point monitoring with continuous spatial sensing of the three-dimensional humidity field, replaces manual ledgers with dynamic accumulation of exposure coefficients, replaces fixed value control with adaptive target humidity and door opening compensation, and replaces manual handling with automatic status linkage.

[0022] This application provides an embodiment of a method for humidity monitoring and moisture-proof storage management of SMD reel materials, see [link to embodiment]. Figure 1 The method specifically includes the following: Step S101: Deploy temperature and humidity sensor nodes in the storage area according to the preset spatial density, collect temperature and humidity data through wireless network and aggregate them into raw environmental data stream, receive material tags read by RFID reader to extract humidity sensitivity level and packaging status, and output material attribute records based on historical cumulative exposure time.

[0023] This embodiment begins with the physical deployment of the storage area, distributing temperature and humidity sensor nodes under each shelf in the shelving array according to a preset spatial density. The spatial density of the nodes is determined based on the highest humidity sensitivity level of the materials stored in that area; the higher the sensitivity level, the smaller the node spacing, ensuring a denser humidity sampling resolution in highly sensitive areas. The sensor probes are all oriented towards the material storage space to avoid measurement deviations caused by airflow obstruction by the metal structural components of the shelving.

[0024] Each node integrates a low-power microcontroller and a wireless transceiver module, using the periodic time base beacon of the wireless network for network-wide clock synchronization. Each sensor node synchronously collects temperature and relative humidity values ​​according to a preset sampling period, and the collected data is encapsulated in data frame format. Each frame contains four fields: node identifier, collection time, temperature value, and humidity value. The data frames are transmitted to the edge gateway via the wireless network in a hop-by-hop relay manner, with each relay completed within a preset transmission delay limit.

[0025] The path selection strategy employs an adaptive routing protocol for low-power lossy networks, and the routing metric integrates hop count and link quality to balance transmission efficiency and link reliability. After receiving data frames from all active nodes, the edge gateway performs aggregation and sorting, arranging the temperature and humidity sequences of each node in ascending order of acquisition time, using the node identifier as the grouping key. After sorting, a time-continuous raw environmental data stream is formed and written to the edge gateway's shared buffer.

[0026] Validity checks are performed synchronously on each data frame. If the temperature or humidity value exceeds the preset valid range, the data at that node at that moment is marked as invalid and recorded in the data quality log to prevent abnormal data introduced by sensor failures or communication errors from contaminating the downstream modeling pipeline. UHF RFID readers are deployed at the receiving workstation. When a material pallet arrives at the predetermined position at the receiving workstation, the arrival sensor outputs a trigger signal to drive the reader to perform tag counting.

[0027] The system extracts four fields from the label user storage area: material code, batch number, humidity sensitivity level, and packaging status. The material code is a pre-defined unique identifier; the batch number is a concatenated string of date and serial number; the humidity sensitivity level is an integer value within a pre-defined range; and the packaging status is an enumerated value for "open" or "unopened." Based on the material code, the system queries the exposure tracking record table in the material database, determining whether it's an initial receipt or a return to inventory.

[0028] If the query returns an empty record and the packaging status is "unopened," it is considered an initial inbound transaction. At this point, an exposure tracking record is created, and the cumulative exposure time field is initialized to zero. If the query returns a valid record, it is considered a return to the database. The historical cumulative time field is read from the record as the starting value for timing, and it continues to accumulate in a background thread in seconds from the time of inbound, ensuring that the exposure time of multiple inbound and outbound transactions remains continuously traceable rather than being reset to zero.

[0029] The material attribute record consists of seven fields: material code, humidity sensitivity level, cumulative exposure time, packaging status, storage location column coordinates, storage location layer coordinates, and inbound timestamp. The material code is used as the unique primary key and written to the material status database. The writing process employs insert or update semantics to accommodate updates to existing records of returned materials. Upon completion of the writing, a material record ready signal is sent to the exposure timeliness tracking pipeline in step S102.

[0030] The material attribute record uses the material code as the primary key and includes fields for humidity sensitivity level and cumulative exposure time, which are read by step S102 at the exposure time tracking entry point for timeliness update calculation.

[0031] Step S102: After filtering the original environmental data stream by exponential weighted moving average, construct a three-dimensional humidity field based on the sensor spatial coordinates and interpolate using radial basis functions. Mark anomalies by spatial gradient. Read the humidity sensitivity level in the material attribute record, look up the time threshold table to determine the upper limit of exposure time, and determine the exposure coefficient based on the local humidity of the cargo location. Update the cumulative time and calculate the remaining effective time.

[0032] In this embodiment, temperature and humidity time series data from each sensor node in the original environmental data stream written to the shared cache in step S101 are read as the starting input for humidity field modeling. The reading process locates data according to node identifiers, and the continuity of the node's acquisition time is verified each time a data acquisition occurs. If the acquisition interval exceeds a preset multiple of the sampling period, it is determined that the node has missing data. The missing positions are filled with the most recent valid value, and the filled positions are marked to ensure the spatial integrity of the interpolation engine's input data.

[0033] The temperature and humidity time series data for each sensor node are applied using an exponentially weighted moving average filter. A preset smoothing factor is used to balance transient noise suppression and response speed to humidity changes. The current filtered value is equal to the smoothing factor multiplied by the current sampled value, plus the unit value minus the smoothing factor, multiplied by the previous filtered value. The computational cost of the recursive operation is linearly related to the number of sampling points, making it suitable for real-time operation in large-scale node deployments without constituting a computing bottleneck.

[0034] The spatial three-dimensional coordinates of each sensor node are read from the calibration data during the installation phase and paired with the filtered humidity values ​​to form a spatial scatter dataset. The coordinates are in metric rectangular coordinates with the lower left corner of the storage area as the origin. Using this scatter dataset as input, a continuous humidity field function covering the entire storage area is constructed using the Gaussian kernel radial basis function interpolation method. The kernel function bandwidth parameter is determined based on a preset ratio of the average spacing between adjacent sensor nodes, ensuring that the humidity contributions of adjacent nodes overlap sufficiently in space without losing local humidity variation characteristics.

[0035] The weighting coefficient vector is obtained by solving a system of symmetric positive definite linear equations constructed by substituting the Euclidean distances between the scattered points into the kernel function. The right-hand side of the equations represents the filtered humidity value of each scattered point. The interpolation matrix is ​​invertible due to the positive definiteness of the Gaussian kernel, and a pre-defined numerical method is used to ensure numerical stability even when the matrix is ​​ill-conditioned. For any spatial location within the storage area, the estimated humidity value at that location can be obtained by substituting its Euclidean distances to each sensor node into the radial basis function interpolation expression.

[0036] The first-order partial derivatives of the continuous humidity field function along three spatial orthogonal directions are used to obtain the three components of the humidity gradient vector. The gradient magnitude is taken as the square root of the sum of the squares of the three components. Spatial locations where the gradient magnitude exceeds a preset threshold condition are marked as humidity anomaly regions. This threshold condition is determined by multiplying the statistical quantile of the humidity spatial gradient under normal air conditioning conditions with a preset safety factor. The continuous spatial range of the anomaly region is defined by an isosurface extraction algorithm, and the diagonal coordinates of the circumscribed cuboid of the anomaly region and the maximum gradient magnitude within the region are output.

[0037] The exposure time tracking pipeline reads the humidity sensitivity level and cumulative exposure time fields from the material attribute record output in step S101. Using the humidity sensitivity level as an index, it queries a preset time threshold table to obtain the upper limit of workshop exposure time corresponding to that level; this table is set according to industry-standard settings. For each material with an exposure tracking record, it performs radial basis function interpolation from the continuous humidity field function, using the coordinates of the material's location as the query point, to obtain the local humidity value at that location.

[0038] The exposure coefficient is determined based on the preset humidity range to which the local humidity value belongs. When the humidity inside the drying cabinet is controlled, the exposure coefficient is set to zero to indicate that the exposure timer is paused. In low-humidity environments, a preset coefficient lower than the unit value is used; in normal workshop environments, a unit value is used; and in high-humidity environments, a preset coefficient higher than the unit value is used to indicate accelerated moisture absorption. In the background timing thread, all materials are traversed at a preset calculation cycle, and the exposure coefficient is multiplied by the increment of the calculation cycle duration and added to the current cumulative exposure time to obtain the updated cumulative time.

[0039] The remaining valid time is obtained by subtracting the updated cumulative time from the upper limit of workshop exposure time. The updated cumulative time and the remaining valid time are synchronously written back to the exposure tracking record table of the material status database. When the remaining valid time becomes a non-positive value, the material timeout flag is immediately set to bypass the periodic scan delay of step S103, ensuring that the timeout event is captured within the same calculation cycle. The remaining valid time and the updated cumulative time are read by step S103 at the status classification entry.

[0040] Step S103: Generate an early warning by classifying the material status according to the proportion of effective time, determine the target humidity based on the highest sensitivity level of the material in the area, and generate a control command by calculating the humidity intrusion compensation based on the door opening event.

[0041] This embodiment uses the remaining effective time and the upper limit of workshop exposure time written into the material status database in step S102 as inputs to perform multi-level classification of material status. The ratio of the remaining effective time to the upper limit of exposure time is calculated to obtain the remaining efficiency ratio, which is a floating-point number within the closed interval from zero to a unit value. The remaining efficiency ratio is sequentially input into a cascaded comparator composed of a first preset ratio threshold condition and a second preset ratio threshold condition, and each level of comparison independently outputs the status determination result.

[0042] When the remaining time-sensitive percentage is higher than the first threshold condition, the material status output is normal, and the material can participate in regular inbound / outbound and production scheduling processes. When the remaining time-sensitive percentage is between the first and second threshold conditions, the material status output is an early warning, and the material's outbound priority is increased in the outbound task queue according to a preset weighting rule, triggering a priority outbound reminder. When the remaining time-sensitive percentage is between the second threshold condition and zero, the material status output is near expiration, triggering an emergency outbound command or an automatic transfer command to the drying cabinet to pause the exposure timer and extend the usable time.

[0043] If the material status output is not higher than zero, it is considered timed out, triggering a material locking operation and a prohibition on outbound command to prevent timed-out materials from entering the production process. The environmental anomaly correction rule performs an inclusion determination on the spatial range of the humidity anomaly area output in step S102 and the current material location coordinates. If the location coordinates fall within the outer cuboid of the anomaly area and the current material status is not timed out, the status level is increased by one level based on the original criteria.

[0044] The adjusted linkage actions are updated synchronously. When the timeout level is reached after the adjustment, an additional immediate execution of material locking is triggered. When the status changes, a status change event is generated and written to the status change log. The log records include the event timestamp, material identifier, and the status fields before and after the change. The status hierarchy of each material is calculated independently and can be parallelized. It is executed concurrently with the material code as the group key to improve the throughput of hierarchical classification for large batches of materials.

[0045] The target humidity setting for the moisture-proof storage cabinet is dynamically determined by the distribution of humidity sensitivity levels of the currently stored materials within the cabinet. The distribution of humidity sensitivity levels of materials in each moisture-proof storage cabinet is statistically analyzed according to a preset spatial grid, and the highest value is taken as the highest sensitivity level for that area. The corresponding target humidity setting value is then retrieved from a preset level-to-target humidity mapping table, indexed by the highest sensitivity level. In this mapping table, the materials with the highest sensitivity level correspond to the most stringent target humidity setting value, with lower sensitivity levels corresponding to progressively more lenient setting values.

[0046] The detection of cabinet door opening events is triggered by a level transition of the cabinet door status sensor. The start time of the opening event is recorded when the cabinet door transitions from the closed state to the open state, and the end time is recorded when it transitions from the open state to the closed state. The difference between the two times is the duration of the opening. During the opening period, the ambient humidity outside the cabinet is obtained from the real-time humidity reading of the sensor node closest to the cabinet door, and the humidity inside the cabinet is obtained from the sampled value of the humidity sensor inside the cabinet at the moment the door is closed.

[0047] The humidity intrusion compensation is calculated using an exponential decay model. The compensation equals the intrusion coefficient multiplied by the difference between the external and internal humidity, multiplied by 1 minus the negative door opening duration of the natural constant, divided by the exponential value of the time constant. The intrusion coefficient and time constant are determined through fitting and calibration of humidity recovery curves within the cabinet under different door opening durations. The humidity intrusion compensation is then superimposed on the target humidity setpoint to obtain the corrected target humidity.

[0048] The control command is encoded in a structured data format, containing four fields: target humidity value, control mode, humidity intrusion compensation amount, and compensation duration. The control mode takes an enumerated value of dehumidification, nitrogen filling, standby, or energy saving. The generated control command is written to the control command queue for the execution control link in step S104 to read in a first-in-first-out order.

[0049] Step S104: After parsing the control command, adjust the power of the dehumidifier unit through a proportional-integral-derivative controller or adjust the flow rate of inert gas through a proportional valve to make the humidity of the storage area approach the target humidity.

[0050] In this embodiment, control commands are retrieved from the control command queue in a first-in, first-out (FIFO) order, and the target humidity value and control mode field are parsed as the initial input for execution control. When the control mode is set to dryer mode, the dehumidification control link is activated, and the output power of the semiconductor dehumidifier unit is adjusted using a proportional-integral-derivative (PID) controller. The error input is the difference between the target humidity value and the current feedback value from the humidity sensor inside the cabinet. The control cycle is set to a preset value in seconds to ensure a balance between the adjustment response speed and the actuator lifespan.

[0051] The proportional gain, integral time, and derivative time of the proportional-integral-derivative (PID) controller are tuned to preset values ​​using the step response method and executed according to a preset control cycle. The proportional term generates an instantaneous output proportional to the current error; the integral term accumulates historical errors to eliminate steady-state residual error and adds anti-integral saturation processing; and the derivative term provides lead correction based on the error change rate to prevent overshoot. The sum of the three terms is output-limited and written to the power regulation register to drive the pulse width modulation duty cycle of the dehumidifier unit.

[0052] When the inert gas cabinet mode is selected, the nitrogen purging control link is activated, and the inert gas flow rate is adjusted by an electronic proportional valve. The control adopts a cascade structure: the outer loop uses proportional-integral-derivative control with humidity error to output the target setpoint for gas flow rate, and the inner loop uses flow sensor feedback to form a flow regulation closed loop. The oxygen content sensor collects the oxygen volume fraction in the cabinet at a preset measurement cycle. When the oxygen content exceeds the preset concentration threshold, the target setpoint for gas flow rate is increased to ensure the maintenance of the inert atmosphere.

[0053] Humidity field spatial homogenization control is based on the humidity gradient vector field output in step S102. The gradient magnitude of the current sampling points within the cabinet is extracted, and the maximum magnitude is taken. The internal circulation fan speed is adjusted according to a preset linear mapping relationship with this maximum magnitude; the larger the gradient magnitude, the higher the fan speed, thus accelerating the uniform diffusion of moisture within the cabinet. The slope and intercept of the mapping function are determined through debugging and testing to ensure that the preset homogenization time index can be achieved under different gradient intensities.

[0054] The cabinet door opening event triggers an interlock protection action, which cuts off the power to the inert gas inlet solenoid valve within a preset response time via an isolation relay, causing the valve to close mechanically and preventing gas from continuously being injected into the external space when the cabinet door is open. Simultaneously, the internal circulation fan speed command is directly set to a preset ratio of the maximum speed, bypassing the gradient mapping path, to form a high-speed downward airflow barrier at the cabinet door opening to slow down the convection of air between the inside and outside of the cabinet.

[0055] When the deviation between the target humidity value and the current humidity value is lower than the preset in-position window width for multiple consecutive control cycles, it is determined that humidity regulation has reached a steady state, and the regulation mode is automatically switched to the standby mode to reduce system energy consumption. In the standby mode, humidity inspection is performed at a preset frequency reduction period, and when the deviation exceeds the standby trigger threshold condition, the corresponding dehumidification mode or nitrogen charging mode is automatically restored. The switching event of the regulation mode is written into the regulation status log for synchronous use in step S105.

[0056] Step S105: synchronizing the material status grading result and the early warning to a warehouse management system and a manufacturing execution system through a communication interface to trigger linkage for priority outbound or material locking.

[0057] In this embodiment, data to be synchronized is read from the material status grading result of step S103 and the regulation status of step S104, and used as the initial input for system linkage. Status synchronization data packets are generated by coding at a preset synchronization period, and the packet body includes fields of material identifier, humidity sensitivity level, current status, remaining effective time, storage position code and data generation timestamp. Each packet can carry multiple material records to reduce communication frequency overhead, and the packet body adopts compact binary coding to control network bandwidth occupation.

[0058] The communication interface adopts a dual-channel architecture to balance real-time performance and reliability. The first channel is a representational state transfer interface, which sends data packets to preset network endpoints of the warehouse management system and the manufacturing execution system by means of a push request method of hypertext transfer protocol. The second channel is a publish-subscribe mechanism based on unified architecture for process control, and the system, serving as a server, publishes material state variables to an address space for subscription and receiving by upper-level system clients. The channel selection preferentially uses the unified architecture for process control channel, and falls back to the representational state transfer channel when the former is unavailable.

[0059] After the status synchronization data packet is sent, an acknowledgment waiting session is started to wait for the receiver to return an acknowledgment receipt containing a reception status code and a reception time within a preset acknowledgment time limit. Retransmission is triggered when no acknowledgment receipt is received after timeout, the maximum number of retransmissions is a preset value, and the interval between adjacent retransmissions increases in an exponential backoff manner to avoid network congestion. After all retransmissions fail, a communication fault event is recorded in a fault log, and the linkage mode is degraded to local acousto-optic alarm and on-site large-screen display of a dashboard, so as to ensure that information is not lost under extreme communication conditions.

[0060] Differential disposal is performed for linkage actions according to material status grading. For materials in an early warning state, the manufacturing execution system preferentially selects them from the material demand list of to-be-scheduled work orders for scheduling matching, and the warehouse management system raises the priority of tasks related to the materials in the outbound task queue and marks them with a priority outbound mark. For materials in an expiry-imminent state, an urgent material calling notification is triggered and pushed to relevant responsible persons, and the warehouse management system automatically generates a warehouse moving task for transferring the materials from the current cargo location to the nearest available drying cabinet and sets the task as the highest priority.

[0061] Materials in an expired state are removed from the available bill of materials by the Manufacturing Execution System (MES) and prevented from being allocated and scheduled by subsequent production orders. The Warehouse Management System (WMS) triggers a material lock and automatically generates a baking assessment work order. The baking work order queries the baking parameter table to obtain the baking temperature and baking time parameters corresponding to the material's humidity sensitivity level and cumulative exposure time. After baking is completed, the operator performs a confirmation operation on the human-machine interface (HMI). The system automatically resets the material's cumulative exposure time field to zero, clears the lock flag, and updates the material status to normal.

[0062] The response delay of the linkage action is calculated from the timestamp of the material status change event to the moment the receiver returns the confirmation receipt. If this delay exceeds the preset upper limit threshold, a linkage timeout alarm event is registered. Timeout alarm events are categorized by alarm level and summarized in the periodic report in the form of a linkage response timeliness statistics table. The statistics table includes three statistical quantities for system operation and maintenance evaluation: the mean, the maximum, and the timeout ratio of linkage response delays.

[0063] Step S106: Push alarm events in a tiered manner and record manual operations to the audit log.

[0064] This embodiment uses a pre-defined alarm system, categorizing alarm events generated at each stage into preset levels based on severity and impact, as input for tiered alarm push notifications. Material timeout locking events, dehumidifier unit malfunction events, inert gas generator malfunction events, and large-scale sensor node offline events are classified as the highest alarm level. Alarms at this level are pushed to the responsible person's mobile terminal in real time and trigger the warehouse's on-site audible and visual alarm devices, requiring the responsible person to respond and handle the situation within a preset response time limit.

[0065] Events such as materials nearing their expiration date, abnormal humidity areas persisting for longer than a preset duration, single sensor offline events, and cabinet door opening times exceeding a preset duration are categorized as intermediate alarm levels. These alarms are pushed to the warehouse dashboard and displayed with color-coded visual markers distinct from normal information, requiring handling within a preset timeframe. Intermediate-level alarms that are not handled within the time limit are automatically escalated to the highest level and re-pushed to the responsible person's terminal according to the escalated rules.

[0066] Material warning events, periodic energy-saving reminders, and sensor calibration expiration reminders are classified as the lowest alarm level. Alarms at this level are only written to the alarm log and summarized in periodic reports; real-time push notifications are not implemented to reduce the frequency of interference with operators. This three-tiered push strategy balances alarm timeliness with the information load on maintenance personnel, ensuring that alarms of different severity receive a response speed commensurate with their risk.

[0067] Alarm lifecycle management covers four stages: generation, push notification, confirmation, and closure. When an alarm is generated, the system assigns a unique alarm identifier and creates a record in the alarm table. The record includes fields for alarm identifier, alarm type, alarm level, generation time, associated material identifier or equipment identifier, and alarm description. The alarm confirmation operation updates the alarm record status field to "confirmed" and registers the confirmer's identifier and confirmation time. Alarm closure is completed through manual operation to close or automatic triggering of the alarm clearance event.

[0068] A human-machine interface is provided for operators to view the list of currently active alarms, confirm alarms, perform manual review, and, when necessary, revoke automatic system decisions. The revocation operation, as a high-privilege intervention, requires the operator's employee ID to be collected and a reason to be selected from a preset set of revocation reasons. Reason options include sensor reading errors, tag information errors, system misjudgments, or production emergencies. After revocation, the system records the pre-operation and post-operation status values ​​to the audit log.

[0069] The audit log uses the operation timestamp as the primary key and records five fields: employee ID, operation type, operation summary, pre-operation status value, and post-operation status value. Logs are archived in separate tables daily, and the retention period meets preset audit compliance requirements. An immutable append-only writing policy is implemented for the audit logs; written records cannot be modified or deleted. The audit log provides a complete and traceable compliance record chain for material disposal operations.

[0070] After the material is locked, the baking process queries the baking parameter table to obtain the baking temperature and baking time parameters corresponding to the humidity sensitivity level and cumulative exposure time. The baking work order includes the material code, baking temperature, baking time, and baking equipment number fields, and is pushed to the baking equipment control system and the warehouse manager's terminal. After baking is completed, the operator confirms on the human-machine interface, and the system automatically resets the cumulative exposure time field of the material to zero, unlocks the material, and updates the material status to normal.

[0071] As can be seen from the above description, the humidity monitoring and moisture-proof storage management method for SMD reel materials provided in this application embodiment uses a high-density distributed sensor array and three-dimensional humidity field modeling to continuously sense the humidity of the microenvironment and quickly locate anomalies. It uses a dynamic exposure coefficient based on the local humidity of the storage location to replace the fixed-rate time accumulation to accurately track materials. It uses the dynamic setting of target humidity driven by the distribution of material humidity sensitivity level and the feedforward compensation of door opening events to balance the moisture-proof control accuracy and energy consumption. The warehouse management system driven by the material status and the manufacturing execution system automatically link and alarm hierarchical push constitute a closed-loop management of the entire process from humidity sensing to material disposal.

[0072] In one embodiment of the humidity monitoring and moisture-proof storage management method for SMD reel materials in this application, see [link to relevant documentation]. Figure 2It can also specifically include the following: Step S201: Install the temperature and humidity sensor nodes under the shelf panels with the probes facing the material space. After each node is synchronized with the wireless network time base beacon, it collects temperature and relative humidity values ​​according to a preset sampling period. The data includes node identifier, collection time, and temperature and humidity values, which are relayed hop by hop to the edge gateway via the wireless network according to a preset time delay limit.

[0073] This embodiment starts with the physical layout of each shelf in the storage area to install and calibrate the sensor nodes. Each temperature and humidity sensor node is installed at a preset spatial spacing at the center below the shelf, with the probes uniformly facing the material storage space to maximize the perception of the true temperature and humidity of the area where the material is located. After installation, a laser rangefinder is used to measure the three-dimensional offset of the node relative to the origin of the storage area. The calibration results are written to the non-volatile memory of each node and simultaneously uploaded to the sensor configuration table for use in three-dimensional humidity field modeling.

[0074] Time synchronization uses the edge gateway as the master time base node. The gateway broadcasts a synchronization beacon frame containing the gateway's system time to the entire network within a preset beacon period. Each slave node, upon receiving the beacon frame, calculates its own clock deviation and calibrates its time base using a digital phase-locked loop. After a preset convergence period, the clock synchronization accuracy of all network nodes is controlled within a preset millisecond-level error range. Synchronization acquisition commands are controlled by the acquisition trigger flag within the beacon frame. Upon receiving a beacon frame with the trigger flag set, all network nodes synchronously initiate sampling in the next time slot.

[0075] The sampled data is encapsulated in a data frame format. The frame structure includes four fields: node identifier, sampling time, temperature value, and humidity value. The temperature value is encoded using a preset bit width and a fixed number of points in Celsius. The humidity value is encoded using the same bit width and a fixed number of points in the percentage of relative humidity. The data frame is temporarily stored locally at the node in a circular buffer. The buffer depth is preset to provide buffer space and prevent data loss when the wireless channel is congested.

[0076] Data frame transmission is performed in the wireless network using a hop-by-hop relay method. Temporary frames are transmitted to the edge gateway via the relay path in a first-in-first-out order. The dwell time of each data frame at each hop node does not exceed the preset transmission delay limit. The routing metric combines two dimensions: hop count and link quality indicator. The hop count weight is a preset value, and the link quality indicator is the moving average of the success rates of the most recent transmissions to dynamically reflect the link status.

[0077] Nodes periodically broadcast route discovery messages to their neighbors to maintain their routing tables. In the event of a link failure, a local repair mechanism is triggered to search for an alternative path within a preset repair timeframe. If no alternative path is found, the node is marked as offline and the offline event is recorded in the fault log. After receiving data frames from all active nodes, the edge gateway uses the node identifier as the grouping key and sorts the temperature and humidity sequences of each node in ascending order of the time of data collection.

[0078] The fusion verification module performs validity checks on each frame of data. If the temperature or humidity value exceeds the preset valid range, the frame of data is marked as invalid and an invalid data event is registered in the data quality log. Simultaneously, the validity check performs rate-of-change detection on the temperature and humidity differences between adjacent sampling points. If the rate of change exceeds the preset physical upper limit, it is determined to be a sensor transient fault, and the sampling point is removed. The sorted valid data forms the raw environmental data stream and is written to the shared cache of the edge gateway.

[0079] The raw environmental data stream uses node identifiers as the primary key and acquisition time as the time sequence index, allowing step S301 to locate and read the data at the filter pipeline entry point based on node identifiers. Each node's buffer is organized in a circular queue, with the latest sampled frame at the head of the queue. Downstream modules move the read pointer after reading the data; a buffer overflow alarm is triggered when the distance between the read and write pointers exceeds the queue capacity.

[0080] Step S202: After the RFID reader at the warehousing station performs an inventory check on the material tags at a preset distance, it extracts the material code and batch number from the tag user storage area and queries the material database to obtain the historical cumulative time and the humidity sensitivity level. When the material is first stored in the warehouse and is still sealed, the cumulative time is set to zero. When the material is returned to the warehouse, the historical cumulative time is used as the starting value to continue accumulating the time.

[0081] This embodiment initiates the material attribute acquisition process upon the arrival signal at the receiving station. An UHF RFID reader / writer is deployed at the receiving station and connected to a circularly polarized antenna. The reader / writer forms an effective tag activation area within a preset reading distance at a preset transmission power. When the material pallet arrives at the predetermined position at the receiving station and the arrival sensor outputs a valid signal, the reader / writer initiates a tag inventory command within a preset response delay. It supports an anti-collision protocol that allows simultaneous reading of multiple tags within a single cycle to meet the needs of scenarios involving multiple reels of material being received together.

[0082] The material code and batch number are extracted from the label user storage area. The raw data from the label response undergoes cyclic redundancy check to ensure transmission integrity. If the check fails, a label reading anomaly event is registered, triggering a manual barcode scanning verification process. This involves scanning the material's outer packaging barcode using a handheld terminal for secondary confirmation. After manual confirmation, the verification result is entered into the system, and the correct material code and batch number are then backfilled into the material attribute acquisition process.

[0083] The material database's exposure tracking record table uses the material code as a unique primary key and includes fields for cumulative exposure time, timer start time, and most recent update time. When querying this table by material code, if the query returns empty and the current packaging status is unopened, it is considered an initial receipt. In this case, a new exposure tracking record is created, the cumulative exposure time field is set to zero, and the timer start time is set to the current system time. In the initial receipt scenario, the system simultaneously assigns a target storage location to the material, with the allocation strategy considering both the current distribution of empty storage locations and the material's sensitivity level zoning requirements.

[0084] When a query returns a valid record and the packaging status is "open," it is determined to be a return to the warehouse. The historical cumulative time field value is read from the exposure tracking record as the starting value of the current cumulative time, and the time continues to accumulate in seconds in the background timing thread starting from the time of warehouse entry. In the case of a return to the warehouse, the timing start time field is not updated, maintaining the initial timing starting point of the original record to maintain a complete exposure history chain, making the accumulation of exposure time from multiple warehouse entries and exits traceable.

[0085] Material attribute records are written using the material code as the unique primary key for insert or update operations. Each record contains eight fields: material code, batch number, humidity sensitivity level, cumulative exposure time, packaging status, location column coordinates, location layer coordinates, and inbound timestamp. The location column coordinates and layer coordinates are pre-allocated by the location allocation process in step S101 before the material attribute records are generated. The allocation results are synchronously updated to the location occupancy status table to prevent duplicate allocation.

[0086] After the writing is completed, a material record ready signal is sent to the exposure timeliness tracking pipeline in step S102. This signal is transmitted through an inter-process message queue with the material code as the payload. The material attribute record uses the material code as the primary key for step S302 to query and read at the exposure coefficient determination entry point. The read operation is accelerated by the database index to ensure query response time in scenarios with a large number of materials being added to the warehouse concurrently.

[0087] In one embodiment of the humidity monitoring and moisture-proof storage management method for SMD reel materials in this application, see [link to relevant documentation]. Figure 3 It can also specifically include the following: Step S301: The temperature and humidity time series of each sensor node in the original environmental data stream are filtered by exponential weighted moving average and the smoothing factor is set to a preset value to suppress transient noise. The spatial coordinates of each node and the filtered humidity value are used as scatter sources and Gaussian kernel radial basis function interpolation is used to construct a continuous humidity field function covering the entire storage area.

[0088] In this embodiment, the raw environmental data stream written to the shared buffer in step S201 is read sequentially according to the node identifier, and the time-series data of each sensor node is used as the starting input of the filtering pipeline. The reading location key is used as the node identifier, and each read retrieves all newly added sampling frames of that node since the last read. Exponentially weighted moving average filtering is performed on the temperature and humidity time series respectively, and the filter runs in a recursive manner to support streaming data processing.

[0089] The recursive formula for filtering is: the current filter value equals the smoothing factor multiplied by the current sampled value plus the unit value minus the smoothing factor, multiplied by the filter value from the previous time step. The smoothing factor, within a preset range, is determined by a trade-off between suppressing transient measurement noise and responding to real humidity changes. A smaller value results in smoother filtering but increased response lag, while a larger value leads to faster response but increased residual noise. Data loss is determined when the sampling interval exceeds a preset multiple of the sampling period.

[0090] Missing locations are filled with the most recent valid value, and a fill flag is set for downstream modules to identify, preventing the fill value from being misinterpreted as the actual measurement value due to the fill operation. The spatial three-dimensional coordinates of each sensor node are read from the sensor configuration table and paired with the filtered humidity value to form a spatial scatter dataset. Each scatter record contains four fields: X-axis coordinate, Y-axis coordinate, Z-axis coordinate, and humidity value, with the coordinates represented in a Cartesian coordinate system in meters.

[0091] Radial basis function interpolation expresses the target humidity field as a weighted linear combination of kernel functions centered at all sensor nodes. The weight coefficient vector is obtained by solving a system of linear equations, and the interpolation matrix is ​​calculated by substituting the Euclidean distances between the scattered points into the Gaussian kernel function. The Gaussian kernel function takes the form of a natural exponential function, and the bandwidth parameter is set to a preset value. The interpolation matrix is ​​guaranteed to be invertible due to the positive definiteness of the Gaussian kernel. The solution employs Choliski decomposition to improve numerical efficiency by utilizing the matrix symmetry and positive definiteness.

[0092] The Gaussian kernel bandwidth parameter is determined based on a preset ratio of the average sensor spacing. This ratio strikes a balance between excessively small bandwidth leading to overfitting at scatter points and excessive fluctuations between scatter points, and excessively large bandwidth leading to overly smooth interpolation results and loss of local humidity characteristics. For any spatial location within the storage area, the humidity estimate at that location can be obtained by substituting its Euclidean distance to each sensor node into the radial basis function interpolation expression. The computational cost is linearly related to the number of sensor nodes.

[0093] The calculation of the spatial gradient vector of the humidity field utilizes the property that the partial derivatives of the Gaussian kernel radial basis function have a closed analytical expression. The partial derivative of the humidity field with respect to the X-axis is equal to the sum of the weight coefficient multiplied by the X-coordinate of the query point minus the node's X-coordinate multiplied by the Gaussian kernel value at that point, divided by the square of the bandwidth parameter. The same applies to the Y-axis and Z-axis directions. The gradient magnitude is taken as the square root of the sum of the squares of the three components. Spatial locations where the gradient magnitude exceeds a preset threshold are marked as humidity anomaly regions.

[0094] The threshold condition is determined by multiplying the spatial gradient statistical quantile under normal environmental conditions by a preset safety factor. The spatial range of the abnormal region is defined by a three-dimensional isosurface extraction algorithm, outputting the diagonal coordinates of the circumscribed cuboid of the abnormal region, as well as the maximum gradient magnitude and maximum humidity value within the region. The continuous humidity field function and abnormal region information are called in step S302 at the local humidity interpolation entry point. The calling interface returns the humidity estimate and gradient vector with spatial coordinates as input parameters.

[0095] Step S302: Obtain the first-order partial derivative of the continuous humidity field function along the three spatial directions to obtain the humidity gradient vector field. Mark the spatial location where the gradient magnitude exceeds the preset threshold as the humidity abnormal area. Use the coordinates of the material location as the query point to perform interpolation to obtain the local humidity value and determine the exposure coefficient accordingly to update the cumulative exposure time.

[0096] In this embodiment, the local humidity value at the current material location coordinates is obtained by performing radial basis function interpolation from the continuous humidity field function constructed in step S301. The query process sequentially calculates the Euclidean distance between the point and each sensor node, substitutes it into the Gaussian kernel function to obtain the value, and then performs a linear combination and summation of each kernel function value and the corresponding weight coefficient. The interpolation calculation utilizes the weight coefficient vector and sensor coordinate set already solved in step S301, eliminating the need to resolve the equation system.

[0097] The exposure coefficient is determined by segmented mapping based on the preset humidity range to which the local humidity value belongs. When the humidity inside the drying cabinet does not exceed the preset upper limit percentage, the material is in a controlled state, and the exposure coefficient is set to zero to indicate that the moisture absorption rate is zero and the exposure timer is paused. In low-humidity environments, the exposure coefficient is set to a preset coefficient lower than the unit value to indicate that the moisture absorption rate is approximately half that of the standard environment. The boundary values ​​of each preset range and the corresponding coefficients of each range are determined by fitting data from material moisture absorption rate experiments under different humidity conditions.

[0098] In normal workshop conditions, the exposure coefficient is a unit value, representing the moisture absorption rate consistent with standard test conditions. In high humidity environments, the exposure coefficient is a preset coefficient higher than the unit value to represent accelerated moisture absorption. The acceleration coefficient is determined by a fitting function with the high humidity environment as the independent variable and the moisture absorption rate ratio curve as the dependent variable. The segmented mapping table of the exposure coefficient is stored using the upper and lower boundaries of the humidity range as index keys. During a query, the corresponding coefficient value is located by the range in which the local humidity value falls.

[0099] The incremental update of the cumulative exposure time is executed in a background thread in a preset calculation cycle. In each cycle, the material status database is traversed and the materials with exposure tracking records are updated in batches. For each material, the local humidity value of the current storage location is first obtained by humidity field interpolation to determine the exposure coefficient. Then, the exposure coefficient is multiplied by the calculation cycle duration to obtain the increment for the current cycle, which is added to the current cumulative time to obtain the updated cumulative time.

[0100] The atomicity of batch updates is guaranteed by database transaction mechanisms. Before committing the update operation, a row-level lock is applied to the material status database to avoid conflicts with concurrent writes from other threads. After the update is completed, the remaining valid time is calculated by subtracting the updated cumulative time from the upper limit of workshop exposure time corresponding to the humidity sensitivity level of the material. The remaining valid time is synchronously written back to the material status database. The write-back operation and the cumulative time update are committed within the same transaction to ensure data consistency.

[0101] When the remaining valid time becomes non-positive, an additional immediate setting of the material timeout flag is triggered to bypass the cycle scan delay, ensuring that the timeout event is captured within the same calculation cycle. The setting operation of the timeout flag is implemented through a database trigger mechanism, automatically executing condition judgment and flag field writing after the cumulative time field is updated. The updated cumulative time and remaining valid time are read by step S401 at the status classification entry point.

[0102] In one embodiment of the humidity monitoring and moisture-proof storage management method for SMD reel materials in this application, see [link to relevant documentation]. Figure 4 It can also specifically include the following: Step S401: Calculate the ratio of the remaining effective time to the upper limit of the workshop exposure time to obtain the remaining effective time ratio, and compare it with the first and second preset ratio threshold conditions in turn. If it is higher than the first threshold condition, it is normal; if it is between the two, it is a warning trigger for priority outbound reminder; if it is between the second threshold condition and zero, it is an emergency outbound or dryer transfer instruction triggered near expiration.

[0103] In this embodiment, the raw data required for status classification is read from the remaining effective time and the upper limit of workshop exposure time written into the material status database in step S302. The remaining effective time ratio is calculated by performing a floating-point division operation with the remaining effective time as the numerator and the upper limit of exposure time as the denominator. The ratio output is a floating-point number between zero and a unit value. In the special case where the denominator is zero, a preset minimum positive number is used to avoid division by zero anomalies, and at the same time, an anomaly event log is triggered.

[0104] The status classification uses a cascading comparison structure for step-by-step determination. First, it checks if the remaining aging period ratio is higher than a first preset threshold. If so, the material status is directly output as normal without further comparison. If not, it checks if the remaining aging period ratio is higher than a second preset threshold. If so, the material status is output as an early warning and a priority outbound reminder is triggered. The values ​​of the two threshold conditions are configured according to the material's humidity sensitivity level, with higher-sensitivity materials having stricter threshold conditions.

[0105] If the first two comparisons fail, check if the remaining time-limit ratio is higher than zero. If it is, the material status is output as "near expiration". If it still fails, the material status is output as "timeout". Each status output includes not only a status identifier but also a corresponding action command: a warning corresponds to a priority outbound reminder, near expiration corresponds to an emergency outbound or dryer transfer command, and timeout corresponds to a material locking and prohibition of outbound actions command.

[0106] The environmental anomaly correction process performs an inclusion check between the spatial range of the humidity anomaly area output in step S301 and the current material location coordinates. If the location coordinates fall within the bounded cuboid of the anomaly area and the current material status is not timed out, the status level is increased by one level based on the original criteria, and the corresponding linkage action is updated synchronously. After the increase, if the timeout level is reached, an additional immediate material locking is triggered to ensure that materials in the humidity anomaly area receive stricter protection even if the timeout period has not expired.

[0107] When a status change occurs, a status change event is generated and written to the status change log. The log record contains five fields: event timestamp, material identifier, status before change, status after change, and trigger code. The trigger code indicates the triggering source of the status change, and its values ​​include time-sensitive criteria triggering, anomaly correction triggering, or both, providing complete traceability for subsequent status auditing and anomaly analysis.

[0108] The status classification of each material is executed independently and in parallel, using the material code as the grouping key for concurrent calculation to improve the classification throughput of large batches of materials. The calculation results are summarized in a status classification result table, which uses the material code as the primary key and includes four fields: current status, remaining time limit, whether it is in an abnormal area, and the time of the most recent status change. The status classification result table is used by the control decision-making link in step S402 and the system linkage link in step S502.

[0109] Step S402: Statistically analyze the distribution of humidity sensitivity levels of materials in each moisture-proof storage cabinet according to the preset grid and take the highest value as the highest sensitivity level in the area. Query the corresponding target humidity setting value from the preset mapping table, detect the door status sensor level jump to calibrate the door opening time and record the duration, and calculate the humidity intrusion compensation amount according to the exponential decay model and add it to the target humidity setting value.

[0110] This embodiment reads the inputs required for control decisions from the material state classification result table in step S401 and the humidity field data in step S301. The statistical analysis of material distribution within the moisture-proof storage cabinet uses a preset spatial grid as the statistical unit. The boundaries of each grid unit are aligned with the physical compartments of the moisture-proof cabinet to avoid errors introduced by cross-compartment statistics. The humidity sensitivity level of each material stored in each grid is read one by one, and the maximum value is taken. When there is no material in the grid, the highest sensitivity level is taken as the default lowest value to reduce energy consumption in the material-free area.

[0111] The humidity sensitivity level mapping table stores the target humidity setting value corresponding to each level, indexed by the humidity sensitivity level. The highest sensitivity level corresponds to the strictest target humidity setting value, i.e., the minimum humidity value, while lower sensitivity levels correspond to progressively more lenient settings. The mapping table content is based on the storage humidity conditions recommended by the material manufacturer. When the composition of materials in the cabinet changes, the target humidity setting value is recalculated. The recalculation operation is performed asynchronously in the callback after the material inbound or outbound transaction is submitted.

[0112] The cabinet door status sensor's detection circuit continuously monitors the voltage level of the magnetic proximity switch installed on the cabinet door frame with a preset microsecond-level response rate. A high voltage level corresponds to the cabinet door being closed, and a low voltage level corresponds to the cabinet door being open. The system captures the falling edge of the voltage level to record the start time of the door opening and the rising edge to record the end time via an interrupt. The interrupt service routine is set to a higher priority than ordinary background threads to ensure the real-time capture of the door opening event.

[0113] During cabinet door opening, real-time humidity readings from temperature and humidity sensor nodes near the cabinet door area are read at preset sampling intervals to form a real-time sequence of external ambient humidity values. The sampling interval is shorter than a preset proportion of the typical cabinet door opening time to ensure sufficient data resolution. The humidity inside the cabinet is measured at the moment the door is closed to reflect the stable humidity state inside the cabinet before the door is opened.

[0114] The humidity intrusion compensation is calculated using an exponential decay model incorporating an intrusion coefficient and a time constant. The compensation is equal to the intrusion coefficient multiplied by the difference between the external ambient humidity and the internal humidity, multiplied by 1 minus the negative door opening duration of the natural constant, divided by the exponential value of the time constant. The intrusion coefficient and time constant are determined by fitting and calibrating the internal humidity recovery curve data under different door opening durations, and are configured separately for different cabinet types to adapt to the differences in physical characteristics of different cabinet capacities and sealing levels.

[0115] The humidity intrusion compensation amount is superimposed on the target humidity setting value to serve as the corrected target humidity. The superimposed value is then subjected to amplitude limiting processing with the lower limit of the target humidity setting value as the boundary. The control command includes four fields: target humidity value, control mode field, humidity intrusion compensation amount, and compensation duration. These fields are serialized and encoded in a structured format and written into the control command queue for step S104 to read in a first-in-first-out order.

[0116] In one embodiment of the humidity monitoring and moisture-proof storage management method for SMD reel materials in this application, see [link to relevant documentation]. Figure 5 It can also specifically include the following:

[0117] Step S501: Analyze the target humidity value and control mode field in the control command. In the drying cabinet mode, the output power of the semiconductor dehumidifier is adjusted by the proportional-integral-derivative controller and the proportional gain, integral time and derivative time are set to preset values. In the inert gas cabinet mode, the gas flow rate is adjusted by the proportional valve and the oxygen content is controlled to not exceed the preset concentration threshold.

[0118] In this embodiment, control commands are retrieved from the control command queue in a first-in, first-out (FIFO) order. The target humidity value and control mode field are parsed as the initial input for control execution. When the control mode is set to drying cabinet mode, the dehumidification control link is activated, and the error signal is the target humidity value minus the current feedback value of the humidity sensor inside the cabinet. The control cycle is set to a preset value in seconds to achieve a balance between the adjustment response speed and the thermal cycle life of the semiconductor dehumidifier unit.

[0119] The three outputs of the proportional-integral-derivative (PID) controller are constructed in discrete form. The proportional output equals the proportional gain multiplied by the current error, providing an instantaneous adjustment component proportional to the error amplitude. The integral output equals the integral gain multiplied by the sum of historical errors, with added anti-integral saturation processing. When the controller output reaches the actuator's physical upper or lower limit, integral accumulation is paused to prevent the integral term from continuously accumulating in the saturation region. The derivative output equals the derivative gain multiplied by the current error minus the previous cycle error, then divided by the control cycle.

[0120] The sum of the three parameters, after being output-limited, is written into the power regulation register to drive the pulse width modulation duty cycle of the semiconductor dehumidification module. The proportional gain, integral time, and derivative time are tuned to preset values ​​using the step response method. The tuning process uses a step change in the target humidity as the excitation signal, and the rise time and overshoot of the humidity response curve inside the cabinet as optimization indicators to determine the parameter combination.

[0121] In inert gas cabinet mode, an electronic proportional valve is used as the actuator, employing a cascade control structure to balance humidity regulation accuracy and flow control stability. The outer loop uses proportional-integral-derivative (PID) control for humidity error, and its output serves as the target setpoint for gas flow. The inner loop uses flow sensor feedback to form a closed-loop flow regulation. The proportional gain and integral time of the flow loop are preset values, and the control cycle of the inner loop is shorter than that of the outer loop to ensure rapid tracking of changes in the outer loop output.

[0122] Oxygen content monitoring uses an independent electrochemical oxygen sensor to collect oxygen volume fraction data once every preset measurement cycle. When the oxygen content exceeds the preset concentration threshold, the target set value of the gas flow rate is increased first. If the flow rate has reached the preset maximum limit but the oxygen content still does not meet the standard, gas source pressure and purity check alarms are triggered and recorded in the fault log. When the oxygen content is within the safe range and the humidity has reached the target value, the gas flow rate is gradually reduced to the maintenance flow rate to save nitrogen consumption.

[0123] Humidity field spatial homogenization control extracts the maximum gradient magnitude of the current sampling grid points within the cabinet from the humidity gradient vector field in step S301. The internal circulation fan speed is adjusted according to a preset linear mapping relationship with this maximum magnitude. The slope and intercept of the mapping relationship are determined through debugging tests by gradually increasing the gradient. The cabinet door opening event triggers an interlocking protection sequence. The first step is to cut off the power supply to the inert gas inlet solenoid valve within a preset response time via an isolation relay, causing the valve to close mechanically. The second step is to directly set the internal circulation fan speed command to a preset proportion of the maximum speed to form an airflow barrier at the doorway.

[0124] Step S502: Read the maximum modulus of the humidity field spatial gradient in the current cabinet and adjust the internal circulation fan speed according to the preset mapping relationship to promote the uniformity of humidity in the cabinet. When the cabinet door is opened, close the inert gas inlet valve and increase the fan speed to a preset ratio to form an airflow barrier at the door.

[0125] This embodiment continues the execution path of cabinet homogenization and interlocking protection from the internal circulation fan control output in step S501. The fan speed is adjusted based on the maximum value of the gradient magnitude at the sampling grid points in the cabinet, continuously updated according to a preset linear mapping relationship. The gradient data is reread and the speed setpoint is updated in each control cycle. The slope and intercept of the mapping function are determined during the debugging phase by homogenization time tests under different gradient disturbance conditions to ensure that the homogenization time of the gradient magnitude within the preset upper limit does not exceed the preset index.

[0126] When the cabinet door is opened, the fan speed directly switches to a preset ratio of maximum speed, bypassing the gradient mapping path to ensure response speed. This high-speed downward airflow forms an air curtain at the cabinet door opening. The physical principle is that the velocity field created by the high-speed air jet at the door opening generates a pressure difference pointing inwards and outwards from the cabinet, thus slowing the leakage of dry air from inside the cabinet and the intrusion of humid air from outside. The effectiveness of this airflow barrier is monitored and evaluated online by the rate of change of the humidity sensor readings inside the cabinet during door opening.

[0127] The status synchronization data packet is generated by extracting material records whose status has changed since the previous period from the status classification result table in step S401 at a preset synchronization period. The data packet contains four top-level fields: version number, message sequence number, generation timestamp, and material record list. Each material record contains subfields: material code, humidity sensitivity level, current status, remaining valid time, and storage location.

[0128] The selection of communication interface channels follows a priority strategy. When a channel in the unified process control architecture is available, its publish / subscribe mechanism is used first to reduce transmission latency. If the channel is unavailable, it falls back to the push method of the descriptive state transition channel. Data packets are pushed to the preset network endpoint using the Hypertext Transfer Protocol and include a message authentication code for data integrity verification. The receiver calculates the message authentication code using the same key and compares it with the received value to verify that the data has not been tampered with.

[0129] Upon receiving synchronized data, the Manufacturing Execution System (MES) implements differentiated scheduling based on material status. Materials in a warning state are prioritized for resource allocation in the shortage list of pending production orders. Multiple warning materials are sorted in ascending order of their remaining effective time to prioritize the consumption of materials with the shortest remaining time. The Warehouse Management System prioritizes the outbound tasks of warning materials in the outbound task queue and highlights them visually on the Kanban board.

[0130] Materials nearing their expiration date trigger two parallel processing paths: the notification path pushes an emergency material requisition notification containing the material code and remaining validity period to the production scheduler via instant messaging; the transfer path creates a transfer task from the current storage location to the nearest available drying cabinet in the warehouse management system and assigns it to the nearest automated guided vehicle (AGV). Materials exceeding their timeout status undergo material eligibility removal from the manufacturing execution system (MES) side, material locking from the warehouse management system side, and an automatic baking assessment work order is generated and pushed to the baking equipment control system.

[0131] In one embodiment of the humidity monitoring and moisture-proof storage management method for SMD reel materials in this application, see [link to relevant documentation]. Figure 6 It can also specifically include the following: Step S601: The material status classification results are encoded into status synchronization data packets according to a preset synchronization cycle and pushed to the warehouse management system and the manufacturing execution system via the descriptive status transfer interface or the unified process control architecture interface. Materials in the warning status are given priority in production by the manufacturing execution system and marked with priority outbound by the warehouse management system.

[0132] In this embodiment, the confirmation waiting session is initiated at the moment the state synchronization data packet is sent. After each data packet is sent, the system assigns a session record containing a session identifier, data packet sequence number, sending time, and number of retransmissions. The confirmation receipt is returned by the receiver after successfully parsing the data packet. The receipt contains a reception status code and reception time field. The receiver must return the receipt within a preset confirmation time limit to complete a full round of synchronization interaction.

[0133] A retransmission is triggered when the confirmation timer reaches its preset timeout duration without receiving a valid acknowledgment. The number of retransmissions is incremented by one and compared with a preset threshold for the maximum number of retransmissions. If the threshold is not reached, the original data packet is retransmitted with the same sequence number to support idempotent deduplication by the receiver. The retransmission interval increases exponentially using a backoff strategy, doubling each time from the preset base backoff duration until a preset backoff limit is reached, preventing exacerbation of congestion in network congestion situations.

[0134] Once the maximum retransmission count is reached, retransmission stops and the session is closed, while three parallel degradation processes are executed. The communication failure event is recorded in the fault log, including the target system identifier, the fault start time, and the range of failed data packet sequence numbers. The audible and visual alarm devices at the warehouse site are activated to transmit alarm information locally, compensating for information delivery failures caused by remote communication interruptions. The communication status indicator on the warehouse dashboard is switched from green to red, and a communication interruption message is displayed at the top.

[0135] The measurement path for the linkage response delay starts from the timestamp of the material status change event within this system and ends at the receiving time in the confirmation receipt returned by the recipient. After each linkage action is completed, the delay is calculated and compared with the preset upper limit threshold condition. If the delay is within the threshold condition, the linkage is completed normally. If the delay exceeds the threshold condition, a linkage timeout alarm event is registered and a degraded linkage path is triggered.

[0136] Linkage timeout alarms are categorized by alarm level and summarized in a linkage response timeliness statistics table in periodic reports. The statistics table includes three metrics: the average and maximum linkage response delays, and the timeout rate, used to assess communication link quality and the reliability of the linkage mechanism. When the timeout rate exceeds a preset maintenance threshold, a communication link health check work order is automatically generated and pushed to the system administrator.

[0137] The confirmation status of the material status synchronization data packet is read by the baking process in step S602 after determining that the material locking has been completed. The baking process uses the confirmation of the material locking as a prerequisite for starting, and the locking confirmation information comes from the locking success receipt returned by the warehouse management system.

[0138] Step S602: Classify the alarm events generated in each stage into levels according to their severity. Material timeout lockout and equipment failure are classified as the highest level and pushed to the responsible person's terminal in real time, requiring a response within a preset time limit. Material near expiration and abnormal humidity are classified as the middle level and pushed to the dashboard for display. Material warning and energy saving reminder are classified as the lowest level and only written to the log for summary in the periodic report.

[0139] This embodiment aggregates alarm events from event logs at each stage and performs hierarchical routing according to a preset hierarchy. Material timeout lockout events, dehumidifier unit failure events, inert gas generator failure events, and large-scale sensor offline events are classified as the highest level. Alarm information is pushed in real time to the mobile terminals of the on-duty supervisor and equipment maintenance personnel, triggering audible and visual alarms in the warehouse. The push message includes the alarm type, associated material or equipment identifier, and required response time limit fields. The response time limit is set according to the urgency of the alarm type.

[0140] Material nearing expiration and abnormal humidity areas persisting beyond the preset duration are categorized as intermediate level alerts. These alerts are pushed to the warehouse dashboard and color-coded to indicate severity: red for near expiration, orange for abnormal humidity, and green for resolved issues. Resolved issues must be completed within the preset timeframe. Intermediate level alerts that are not resolved within the timeframe are automatically escalated to the highest level and re-pushed to the responsible person's terminal according to the escalated rules.

[0141] Material warning events, periodic energy-saving reminders, and sensor calibration expiration reminders are categorized as the lowest level. Alarms at this level are only written to the alarm log and summarized in periodic reports, without real-time push notifications to reduce disruption to operators. This three-tiered push strategy balances alarm timeliness with the information load on maintenance personnel, ensuring that alarms of varying severity receive response speeds and processing priorities commensurate with their risk levels.

[0142] The human-machine interface provides three types of operation entry points: alarm confirmation, manual review, and automatic system decision-making cancellation. When a cancellation operation is executed, the operator's employee number and the reason for cancellation are forcibly collected. The reason for cancellation can be selected from a preset set, including options such as sensor reading error, tag information error, system misjudgment, or production urgency. After cancellation, the system records the status values ​​before and after the operation to the audit log, providing a complete operation trajectory for subsequent review of the reasonableness of the handling.

[0143] The audit log uses the operation timestamp as the primary key and records five fields: employee ID, operation type, operation summary, pre-operation status value, and post-operation status value. Logs are archived in separate tables daily, and the retention period meets preset audit compliance requirements. The audit log employs an immutable append-only policy; written records cannot be modified or deleted. The audit log also supports keyword searches based on operator, operation type, and time range.

[0144] For materials in an expired state, after the material lock takes effect, the baking parameter table is queried to obtain the baking temperature and baking time parameters corresponding to the humidity sensitivity level and cumulative exposure time. The baking work order includes the material code, baking temperature, baking time, and baking equipment number fields, and is pushed to the baking equipment control system and the warehouse manager's terminal. After baking is completed, the operator performs a confirmation operation on the human-machine interface, and the system automatically resets the material exposure cumulative time field to zero, clears the lock flag, and updates the material status to normal.

[0145] To integrate distributed sensing acquisition, three-dimensional humidity field modeling, dynamic exposure time tracking, moisture control, and system linkage technologies into a complete humidity monitoring and moisture-proof storage management device, this application provides an embodiment of a humidity monitoring and moisture-proof storage management device for SMD reel materials, which implements all or part of the aforementioned humidity monitoring and moisture-proof storage management method. See [link to embodiment]. Figure 7 The humidity monitoring and moisture-proof storage management device for SMD reel materials specifically includes the following components:

[0146] The multi-point sensing module deploys temperature and humidity sensor nodes in the storage area at a preset spatial density. It collects temperature and humidity data via a wireless network, aggregating it into a raw environmental data stream. It also receives RFID readers to extract material attributes from material tags and outputs material attribute records based on historical cumulative exposure time. This module uses a temperature and humidity sensor node array and a wireless gateway as its hardware foundation, and edge computing nodes as data aggregation and preprocessing units. Through sensor management middleware, it aligns multi-source sensing data under a unified time base and then completes data verification, aggregation, and sorting.

[0147] The humidity field modeling and anomaly detection module is used to filter the raw environmental data stream using an exponentially weighted moving average and then construct a three-dimensional humidity field based on sensor spatial coordinates and radial basis function interpolation. Anomalies are marked using spatial gradients. This module uses a radial basis function interpolation engine as its core algorithm component, outputting a continuous humidity field function covering the entire storage area, as well as the circumscribed cuboid coordinates of the anomaly region and the maximum gradient magnitude within the region for downstream modules to query.

[0148] The exposure time tracking module is used to read the humidity sensitivity level from the material attribute record, query the time threshold table to determine the upper limit of the exposure time, and determine the exposure coefficient based on the local humidity of the storage location. After updating the cumulative exposure time, it calculates the remaining effective time. This module performs batch time update operations for all materials in a background offline thread with a preset calculation cycle, and obtains real-time local humidity values ​​through the humidity field interpolation interface to maintain an accurate match between the exposure coefficient and the current microenvironment.

[0149] The status grading and control decision module is used to generate early warnings based on the proportion of effective time for material status grading, determine the target humidity based on the highest sensitivity level of materials in each moisture-proof storage area, and calculate the humidity intrusion compensation based on door opening events to generate control instructions. This module uses a state machine to build material lifecycle management, triggers door opening humidity intrusion compensation calculations in an event-driven manner, and outputs control instructions containing target humidity and control mode to the execution module.

[0150] The dehumidification and inert gas control module is used to parse the control commands and adjust the dehumidifier power through a proportional-integral-derivative controller or adjust the inert gas flow rate through a proportional valve to bring the humidity of the storage area close to the target value. This module consists of a drying cabinet control submodule, an inert gas cabinet control submodule, a circulating fan control submodule, and a cabinet door interlock protection submodule. It operates according to a preset control cycle, and the submodules exchange status information via a control bus to ensure coordinated switching of control modes.

[0151] The system linkage and anomaly management module is used to synchronize the material status classification results and the early warnings to the warehouse management system and manufacturing execution system via a communication interface to trigger linkage actions. It also pushes alarm events in a tiered manner and records manual operations to the audit log. This module supports two interface protocols—representative state transition and unified process control architecture—using a dual-channel communication architecture. An alarm classification engine routes alarms to corresponding notification channels according to preset levels, and an audit log engine archives manual operation records daily and maintains data compliance for a preset retention period.

[0152] As can be seen from the above description, the humidity monitoring and moisture-proof storage management device for SMD reel materials provided in this application embodiment constitutes a fully automated closed-loop management from humidity environment perception to material disposal through the synergy of six modules: multi-point sensing, humidity field modeling and anomaly detection, exposure time tracking, state classification and control decision, dehumidification and inert gas execution, and system linkage and anomaly management.

[0153] This invention also 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 program, it implements the above-mentioned method for humidity monitoring and moisture-proof storage management of SMD reel materials.

[0154] This invention also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the above-described method for humidity monitoring and moisture-proof storage management of SMD reel materials.

[0155] This invention also provides a computer program product, which includes a computer program that, when executed by a processor, implements the above-mentioned method for humidity monitoring and moisture-proof storage management of SMD reel materials.

[0156] In this embodiment, a high-density distributed sensor array and three-dimensional radial basis function interpolation humidity field modeling are used to continuously sense and rapidly locate microenvironment humidity. Dynamic exposure coefficient time tracking based on local humidity of the storage location replaces manual ledger recording for precise time-sensitive control of materials. Dynamic setting of target humidity driven by the distribution of material humidity sensitivity level and feedforward compensation for door opening events balance the accuracy of moisture control and energy consumption. The warehouse management system and manufacturing execution system driven by material status classification automatically link to quickly lock and automate the disposal process of expired materials.

[0157] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above descriptions are merely specific embodiments of the present invention and are not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for humidity monitoring and moisture-proof storage management of SMD reel materials, characterized in that, The method includes: Temperature and humidity sensor nodes are deployed in the storage area according to a preset spatial density. Temperature and humidity data are collected through a wireless network and aggregated into a raw environmental data stream. The humidity sensitivity level and packaging status are extracted from the material tags read by the RFID reader and the material attribute record is output based on the historical cumulative exposure time. After filtering the raw environmental data stream using an exponentially weighted moving average, a three-dimensional humidity field is constructed based on the sensor spatial coordinates and interpolated using radial basis functions. Anomalies are marked using spatial gradients. The humidity sensitivity level in the material attribute record is read, and the exposure time upper limit is determined by looking up the time threshold table. The exposure coefficient is determined by interpolating the local humidity of the cargo location, and the cumulative time is updated. The remaining effective time is then calculated. The material status classification is executed to generate an early warning based on the proportion of effective time. The target humidity is determined based on the highest sensitivity level of the materials in the area, and the humidity intrusion compensation is calculated based on the door opening event to generate a control command. After parsing the control instructions, the power of the dehumidifier unit is adjusted through a proportional-integral-derivative controller or the flow rate of inert gas is adjusted through a proportional valve to make the humidity of the storage area approach the target humidity. The material status classification results and the early warning are synchronized to the warehouse management system and the manufacturing execution system through the communication interface to trigger priority outbound or material locking linkage. Alarm events are pushed in a graded manner and manual operations are recorded in the audit log.

2. The method for humidity monitoring and moisture-proof storage management of SMD reel materials according to claim 1, characterized in that, The process of deploying temperature and humidity sensor nodes in the storage area at a preset spatial density to collect temperature and humidity data via a wireless network and aggregating it into a raw environmental data stream includes: The temperature and humidity sensor nodes are installed under the shelf panels with the probes facing the material space. Each node is synchronized with the wireless network time base beacon and collects temperature and relative humidity values ​​according to a preset sampling period. The data includes node identifier, collection time and temperature and humidity values, which are relayed to the edge gateway hop by hop via the wireless network according to a preset time delay limit. After the RFID reader at the warehousing station performs an inventory check on the material tags at a preset distance, it extracts the material code and batch number from the tag user storage area and queries the material database to obtain the historical cumulative time and the humidity sensitivity level. When the material is first put into storage and is still sealed, the cumulative time is set to zero. When the material is returned to storage, the historical cumulative time is used as the starting value to continue accumulating the time.

3. The method for humidity monitoring and moisture-proof storage management of SMD reel materials according to claim 1, characterized in that, The process of constructing a three-dimensional humidity field based on sensor spatial coordinates using radial basis function interpolation after filtering the original environmental data stream with exponential weighted moving average and marking anomalies using spatial gradients includes: The temperature and humidity time series of each sensor node in the original environmental data stream are filtered by exponential weighted moving average and the smoothing factor is set to a preset value to suppress transient noise. The spatial coordinates of each node and the filtered humidity value are used as scatter sources and Gaussian kernel radial basis function interpolation is used to construct a continuous humidity field function covering the entire storage area. The humidity gradient vector field is obtained by taking the first-order partial derivative of the continuous humidity field function along three spatial directions. Spatial locations where the gradient magnitude exceeds a preset threshold are marked as humidity abnormal areas. The coordinates of the material location are used as query points to perform interpolation to obtain local humidity values, and the exposure coefficient is determined accordingly to update the cumulative exposure time.

4. The method for humidity monitoring and moisture-proof storage management of SMD reel materials according to claim 1, characterized in that, The step of reading the humidity sensitivity level from the material attribute record and looking up the time threshold table to determine the upper limit of exposure time includes: Extract the humidity sensitivity level and opening mark from the material attribute record, query the time threshold table based on the humidity sensitivity level to obtain the upper limit of workshop exposure time, set the cumulative exposure time to zero when the material is first put into the warehouse and is still sealed, and read the historical cumulative time from the database when it is returned to the warehouse and continue to count; The exposure coefficient is multiplied by the calculation cycle increment every second and accumulated to the cumulative exposure time. The updated cumulative exposure time is subtracted from the upper limit of workshop exposure time to obtain the remaining effective time. The ratio of the remaining effective time to the total effective time is compared with a preset ratio threshold condition to determine the current state of the material.

5. The method for humidity monitoring and moisture-proof storage management of SMD reel materials according to claim 1, characterized in that, The step of generating early warnings based on the proportion of effective time for material status classification includes: The remaining effective time ratio is obtained by calculating the ratio of the remaining effective time to the upper limit of the workshop exposure time. It is then compared with the first and second preset ratio threshold conditions in sequence. When it is higher than the first threshold condition, it is normal. When it is between the two, it triggers an early warning and priority outbound reminder. When it is between the second threshold condition and zero, it triggers an emergency outbound or dryer transfer instruction. The distribution of humidity sensitivity levels of materials in each moisture-proof storage cabinet is statistically analyzed according to a preset grid, and the highest value is taken as the highest sensitivity level in the area. The corresponding target humidity setting value is queried from the preset mapping table. The door status sensor level jump is detected to calibrate the door opening time and record the duration. The humidity intrusion compensation is calculated according to the exponential decay model and superimposed on the target humidity setting value.

6. The method for humidity monitoring and moisture-proof storage management of SMD reel materials according to claim 1, characterized in that, The step of parsing the control command and adjusting the dehumidifier power through a proportional-integral-derivative controller or adjusting the inert gas flow rate through a proportional valve to bring the humidity of the storage area closer to the target humidity includes: The target humidity value and control mode field in the control command are analyzed. In the drying cabinet mode, the output power of the semiconductor dehumidifier is adjusted by the proportional integral derivative controller and the proportional gain, integral time and derivative time are set to preset values. In the inert gas cabinet mode, the gas flow rate is adjusted by the proportional valve and the oxygen content is controlled not to exceed the preset concentration threshold. The maximum modulus of the humidity field spatial gradient in the current cabinet is read and the internal circulation fan speed is adjusted according to the preset mapping relationship to promote the uniformity of humidity in the cabinet. When the cabinet door is opened, the inert gas inlet valve is closed and the fan speed is increased to a preset ratio to form an airflow barrier at the door.

7. The method for humidity monitoring and moisture-proof storage management of SMD reel materials according to claim 1, characterized in that, The step of synchronizing the material status classification results and the early warning to the warehouse management system and manufacturing execution system via a communication interface to trigger priority outbound or material locking linkage includes: According to a preset synchronization cycle, the material status classification results are encoded into status synchronization data packets and pushed to the warehouse management system and the manufacturing execution system via a descriptive status transfer interface or a unified process control architecture interface. Materials in the early warning status are given priority in production by the manufacturing execution system and marked with a priority outbound tag by the warehouse management system. The alarm events generated in each stage are classified into levels according to their severity. Material timeout lockout and equipment failure are classified as the highest level and pushed to the responsible person's terminal in real time, requiring a response within a preset time limit. Material near expiration and abnormal humidity are classified as the middle level and pushed to the dashboard for display. Material early warning and energy saving reminder are classified as the lowest level and are only written to the log and summarized in the periodic report.

8. A humidity monitoring and moisture-proof storage management device for SMD reel materials, characterized in that, The device includes: The multi-point sensing module is used to deploy temperature and humidity sensor nodes in the storage area according to a preset spatial density, collect temperature and humidity data through a wireless network, aggregate them into a raw environmental data stream, receive RFID readers to read material tags to extract material attributes, and output material attribute records by combining historical cumulative exposure time. The humidity field modeling and anomaly detection module is used to construct a three-dimensional humidity field based on the sensor spatial coordinates and interpolation according to the radial basis function after filtering the original environmental data stream by exponential weighted moving average and marking anomalies by spatial gradient. The exposure time tracking module is used to read the humidity sensitivity level query time threshold table in the material attribute record to determine the upper limit of the exposure time, and then calculate the remaining effective time after determining the exposure coefficient based on the local humidity of the cargo location, updating the cumulative exposure time, and calculating the remaining effective time. The status classification and control decision module is used to generate early warnings by classifying material status according to the proportion of effective time, determine the target humidity based on the highest sensitivity level of materials in each moisture-proof storage area, and generate control instructions by calculating humidity intrusion compensation based on door opening events. The dehumidification and inert gas execution module is used to parse the control command and adjust the power of the dehumidifier unit through a proportional-integral-derivative controller or adjust the inert gas flow rate through a proportional valve to make the humidity of the storage area approach the target value. The system linkage and anomaly management module is used to synchronize the material status classification results and the early warning to the warehouse management system and manufacturing execution system via the communication interface to trigger linkage actions, push alarm events in a classified manner, and record manual operations to the audit log.

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 humidity monitoring and moisture-proof storage management method for SMD reel materials according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the computer program implements the steps of the humidity monitoring and moisture-proof storage management method for SMD reel materials as described in any one of claims 1 to 7.

Images