Intelligent tea storage temperature, humidity and atmosphere coordinated regulation and control system

Abstract

The invention discloses an intelligent tea storage temperature, humidity and atmosphere coordinated regulation and control system, and relates to the technical field of tea storage control, and the system is characterized in that sampling ports are arranged in a door area, a cold bridge area, an air return path area and a stack, and an air pump and a multi-way valve group poll and send samples to a centralized dew point measurement unit to obtain a dew point temperature sequence; a door area disturbance event mark is generated in combination with the door magnet and the door opening duration; comparing the dew point temperature with the key surface temperature to obtain a condensation margin, and forming and grading a high-risk node library according to a hysteresis rule; according to the node library, attached dry air scouring, pulse type penetrating circulation and gate area compensation are executed on the target node, and the execution quantity change step length is limited; and after a stable window is met, gradually retreating and quitting, recording an event and processing action sequence, updating strategy parameters, realizing condensation risk positioning and node-based treatment, reducing excessive dehumidification and improving operation and maintenance consistency.

Description

Technical Field

[0001] This invention relates to the field of tea storage control technology, specifically to a smart tea storage temperature, humidity, and atmosphere coordinated control system. Background Technology

[0002] Tea is highly sensitive to environmental temperature and humidity conditions and airflow organization during storage, turnover, and medium- to long-term storage. Tea is typically stored in bags, boxes, or pallets. Warehouses are usually equipped with ventilation, dehumidification, fresh air or circulating fans, and temperature and humidity sensors for constant temperature and humidity control or time-based control to meet the needs of different types of tea for preservation and quality stability. With the increasing scale and frequency of warehousing operations, warehouses often experience disturbances such as door opening and closing, forklift access, and personnel entry and exit. During the rainy season, periods of high humidity, or when there are large diurnal temperature variations, the outside air humidity is high and changes rapidly. Factors such as different thermal conductivity of the enclosure structure, columns and metal components, ground cooling, and air ducts and return air paths can lead to localized cold bridges or dead air zones. Insufficient ventilation penetration inside the stacks and near the ground can cause moisture retention and humidity gradients. While the overall warehouse environment is controllable, local microenvironments may deviate from this controllable pattern.

[0003] In the aforementioned situations, when hot and humid air enters through open doors and is either backflowed or recirculated, or when moisture inside the stack enters a lower-temperature boundary layer, the temperature may drop below the critical temperature for air saturation, potentially leading to condensation. This typically occurs first in concealed locations such as corners, return channels in door areas, the bottom of pallets, and inside the stack, making it difficult to detect. Currently, common control methods involve controlling and alarming relative humidity and temperature thresholds. Sensors are often located at representative positions within the warehouse or near walls. Controllers can start and stop dehumidification or temperature control equipment and implement measures such as regular inspections and stack-turning ventilation. However, relative humidity is significantly affected by temperature and cannot directly reflect the critical temperature for air saturation. Furthermore, the surface temperature of localized cold bridges and the gas state inside the stack often lack continuous coverage, making it difficult to detect condensation thresholds in a timely manner and incorporate them into control decisions.

[0004] Therefore, the technical issues that need to be addressed in the current tea storage environment control are: due to the non-uniform microenvironment caused by door area operation disturbances, cold bridges and internal moisture retention in stacks, the critical state of condensation may be hidden in the location and persist. The overall temperature and humidity indicators do not meet the requirements. If not identified in time, it will cause tea to become damp, moldy, have off-flavors and batch differences, resulting in problems such as scrapping or downgrading sales and increased costs for tracing customer complaints. Summary of the Invention

[0005] (a) Technical problems to be solved To address the shortcomings of existing technologies, this invention provides a smart tea storage temperature and humidity atmosphere coordinated control system. It generates door zone disturbance event markers by combining door magnets and door opening duration; compares dew point temperature with key surface temperatures to obtain condensation margin; and forms and classifies a high-risk node library according to hysteresis rules. Based on the node library, it performs attached dry air flushing, pulsed penetration circulation, and door zone compensation on target nodes, while limiting the step size of execution changes; it re-collects dew points and recalculates condensation margin, gradually exiting after meeting a stable window; records event and handling action sequences, and updates strategy parameters, thereby achieving condensation risk location and node-based management, solving the technical problems described in the background art.

[0006] (II) Technical Solution To achieve the above objectives, the present invention provides the following technical solution: A smart tea storage temperature and humidity atmosphere coordinated control system is implemented using dehumidification equipment. It includes setting up sampling ports in the warehouse door area, cold bridge area and inside the stack, and using an air pump and a multi-way valve group to periodically poll and send air samples to a centralized dew point measurement unit to obtain the dew point temperature sequence, and collecting door magnets to generate door area disturbance event markers. The condensation margin is obtained by comparing the dew point temperature sequence with the key surface temperature. A high-risk node library is generated and classified according to the condensation margin and hysteresis rules. When the high-risk node database triggers the disposal conditions, a combination of targeted dew reduction actions is performed on the target node according to the classification, including dry air flushing of the cold bridge surface by the low dew point dry air branch, pulsed penetration cycle for nodes inside the stack, and door area compensation process when the door area disturbance event is triggered. The system continuously recovers dew point temperature and recalculates condensation margin. When the exit conditions are met, it exits the directional dew reduction action combination according to the gradual withdrawal strategy, records the event and handling action sequence, and updates the strategy parameters of the high-risk node library.

[0007] Furthermore, sampling ports are arranged in zones according to door area, cold bridge area at the corner, return air path area, and representative points inside the stack, and are assigned node numbers. The multi-way valve group polls each node number according to a preset cycle. Before switching nodes, the air pump performs purging and delayed sampling, and generates valid sampling marks to correspond to the dew point temperature sequence.

[0008] Furthermore, the door zone disturbance event flag is generated by the dew point temperature sequence of the door magnetic state, door opening duration, and door zone node number. The controller calculates the door zone change intensity based on the dew point temperature difference between adjacent polling cycles. The door zone disturbance event flag is set when the door magnetic state is open, the door zone change intensity exceeds the threshold, and the minimum duration of continuous door opening is reached.

[0009] Furthermore, the critical surface temperature is collected by anchor temperature measurement points set at the corners of the walls, columns, and the ground, and a node mapping table is established. The node mapping table includes the node number, anchor number, mapping weight, and update timestamp. The mapping weight is determined according to the distance between the node and the anchor. When the anchor is missing, the critical surface temperature is updated based on the wall thermal response estimate.

[0010] Furthermore, a risk score is generated based on the condensation margin and a hysteresis rule is applied. The risk score is obtained based on the cumulative decay over time when the condensation margin is lower than the safety margin. The entry process uses a combination of the entry threshold and the minimum duration for determination, while the exit process uses a combination of a stable window that is below the exit threshold and continuously meets the valid sampling markers, and the node level is kept from bouncing back between entry and exit.

[0011] Furthermore, the high-risk node database classifies and sorts the nodes entering the database into three levels: Level 1, Level 2, and Level 3, and records the level field for each level. The level is determined based on the deterioration trend of condensation margin, recovery speed, and number of repeated triggers. The sorting results are output as a list of target nodes and a treatment priority. The list of target nodes is used as the input for selecting a combination of targeted condensation reduction actions.

[0012] Furthermore, the low dew point dry air branch consists of a dry air source, a branch air duct, an electric branch valve, and an attached air outlet component. The attached air outlet component is fixed to the critical surface temperature corresponding position of the cold bridge area in the corner of the wall. The controller starts and stops the electric branch valve one by one according to the target node list and maintains the preset flushing time, while recording the branch start and stop sequence and writing it into the action sequence.

[0013] Furthermore, the pulse-type penetration cycle is executed on the corresponding nodes of the representative points inside the stack. The controller alternately switches the penetration fan speed and return air path valve position according to the preset pulse cycle, and re-polles the sampling port of the representative point inside the stack after each switch to update the dew point temperature sequence. At the same time, the pulse cycle and valve position switching sequence are written into the treatment action sequence.

[0014] Furthermore, the door zone compensation process is executed during the period when the door zone disturbance event flag is set. The controller opens the door zone preset dry air buffer and adjusts the return air guiding component to make the airflow in the door zone enter the return air path area. At the same time, it sets the upper limit of the change step size for the dehumidification equipment execution quantity, branch valve position and fan speed. After the door magnet changes from open to closed, it continues to execute the preset delayed dehumidification process and then removes the door zone disturbance event flag.

[0015] Furthermore, the exit conditions include that the condensation margin of all target nodes in the high-risk node library continuously meets the safety margin and the sampling validity markers are continuously valid. The gradual exit strategy exits step by step in the order of low dew point dry air branch, pulsed penetration circulation, and normal dehumidification. During the exit process, the interlocking of the gate area disturbance event markers remains valid. After the interlocking is released, the event and the sequence of handling actions are associated and archived, and the strategy parameter version number is updated.

[0016] (III) Beneficial Effects This invention provides a smart tea storage temperature and humidity atmosphere coordinated control system, which has the following beneficial effects: Sampling ports are set up at representative points in the door area, cold bridge area at the corner, return air path area, and inside the stack. A vacuum pump and multi-way valve group alternately deliver samples to a centralized dew point measurement unit to generate a dew point temperature sequence, thus collecting and aligning the data with the concealed microenvironment and door opening fluctuations. Door magnets and door opening duration are combined with the door area dew point temperature sequence to form a door area disturbance event marker, independent of the overall warehouse average temperature and humidity, distinguishing the risks of backflow and cold bridge condensation in the door area and reducing the probability of false triggering.

[0017] Condensation margin is obtained by comparing key surface temperature with dew point temperature. A high-risk node library is generated and classified according to hysteresis rules. Condensation risk is treated by solidifying nodes and outputting treatment priorities. The high-risk node library drives a combination of targeted decondensation removal actions, including attaching dry air flushing to cold bridge surfaces, pulsed penetration circulation to the inside of the stack, and door area compensation when the door area is disturbed. This ensures that the treatment actions correspond to the risk nodes and avoids deep dehumidification of the entire warehouse.

[0018] By setting the step size of the dehumidification equipment's execution volume and guiding the return air, the atmosphere changes during the switching of dry air branches, penetration circulation, and normal dehumidification are regulated to avoid re-humidification and excessive dryness, thus improving operational consistency. By re-collecting the dew point temperature sequence and recalculating the condensation margin, when a stable window is obtained, the gradual withdrawal strategy exits the targeted dew reduction action combination, records the event and handling action sequence, and updates the strategy parameters of the high-risk node library, achieving a reusable governance process under similar operating conditions. Attached Figure Description

[0019] Figure 1 This is a block diagram of the overall system structure of the present invention; Figure 2 This is a schematic diagram of the sampling point layout within the library of the present invention; Figure 3 This is a schematic diagram of the multi-channel polling sampling gas path and isolation structure of the present invention; Figure 4 This is a flowchart of the polling sampling fixed timing and steady-state discrimination process of the present invention; Figure 5 This is a schematic diagram of the gate disturbance identification and event frame encapsulation of the present invention; Figure 6 This is a flowchart illustrating the generation process of condensation margin and node risk in this invention. Figure 7 This is a schematic diagram of the processing sequence and three types of action paths of the present invention; Figure 8 This is a schematic diagram of the timeline for the data acquisition verification, gradual exit, and event review of this invention. Detailed Implementation

[0020] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0021] Please see Figures 1-8 This invention provides a smart tea storage temperature and humidity atmosphere coordinated control system, comprising: Step 1: Using a centralized dew point measurement unit in conjunction with multi-path polling sampling, obtain the dew point temperature sequence covering the door area, cold bridge area at the corner, return air path area, and representative points inside the stack, and generate door area disturbance event markers and valid sampling markers so that Step 2 can directly reference them under a unified timestamp.

[0022] The process involves first solidifying the differences in the microenvironment into a layout of sampling ports that allow for repeated sampling, and then using physical isolation to ensure that the readings from each sampling port do not contaminate each other. The location coverage and accuracy calibration are separated: the sampling front end is responsible for coverage, the centralized dew point measurement unit is responsible for consistency, and maintenance is concentrated on the filter element and seals.

[0023] The loading and unloading at the doorway introduces hot and humid air from the outside. The cold bridge area at the corner and the boundary layer near the pillar are prone to surface cooling. Moisture accumulates in the return air path area. The representative point inside the stack forms a moisture-retaining layer due to insufficient penetrating air volume. If the temperature and humidity at the center point of the warehouse are relied upon only, it is difficult to reflect the local critical state.

[0024] Sampling ports are set up at the door area, corners, return air, and stacked sampling probes. Each sampling port is connected to a multi-way valve group through sampling pipelines. The common end of the multi-way valve group is connected to an air pump. The outlet of the air pump is connected to the air inlet of the centralized dew point measurement unit, forming a gas path chain that covers multiple locations with a single measurement unit.

[0025] The sampling port in the door area is fixed at the confluence of the return flow channel inside the door; the sampling port in the corner is fixed at the boundary layer where the corner meets the column; the sampling port in the return air area is fixed in front of the return air inlet; the stacked sampling probe is inserted into the middle layer or near the ground layer of the stack along the guide hole on the side of the tray and is fastened to the tray frame by a positioning collar. Each sampling pipeline is equipped with a one-way check valve structure, an isolation chamber is set in front of the multi-way valve group, the sampling port is equipped with a removable filter element, and the end of the stacked sampling probe is equipped with an anti-clogging filter head and locked with a sealing ring.

[0026] When in use, the sampling port covers the door area, cold bridge boundary layer, return air intake area and stacked moisture retention layer, and the dew point temperature sequence can reflect the differences in microenvironment; the centralized dew point measurement unit is uniformly calibrated, reducing the drift calibration of multi-point high-precision devices; the isolation chamber, check valve structure and filter element enable the sampling chain to operate for a long time and the reading source is clear.

[0027] Furthermore, a fixed timing sequence is established for valve switching, cleaning, steady-state discrimination, data acquisition, and abnormal jumps, ensuring that dew point samples from each sampling port are traceable to the steady-state window. Dew point temperature is sensitive to changes in water vapor; transition readings can mix residues from the previous channel into the next, leading to condensation margin errors in step two. The controller operates on a polling cycle... The multi-way valve assembly opens sequentially according to the sampling ports; after valve switching, the cleaning phase begins, with a cleaning duration of [duration missing]. The valve position remains unchanged, and the suction pump is driven to displace the intake chamber; then, the steady-state discrimination phase begins, and the controller operates according to the time step. Read continuous dew point temperature and calculate stability index When the stability index Not greater than the stability threshold The data collection phase begins, generating dew point temperature sequences. With sampling valid marker ; If within the maximum waiting time If the sampling fails the steady-state test, an anomaly is recorded and the process moves to the next sampling port. The stability index takes the following form:

[0028] Among them, stability index The value range is 0 to 1, used to characterize the normalized change intensity of the dew point reading within the discrimination window; dew point temperature The value range is -20 to 40 degrees Celsius, used to represent the current channel dew point reading; discrimination window The value range is 3 to 12, used to determine the number of readings involved in the judgment; the time step... The value range is 1 to 10 seconds, used to match the response of the measurement unit; Attenuation coefficient The value ranges from 0.1 to 1, and it is suitable for attenuating earlier readings; stable threshold. The value range is 0.01~0.2; the polling period is... The value range is 30~600s, applicable to coverage frequency; cleaning duration The value range is 3~60s, applicable to replacement chambers; maximum waiting time. The value range is 10~120s, which is suitable for slowing down the entire wheel in a single channel; Valid sampling marker The value can be 0 or 1, where 1 indicates the sample comes from the steady-state window and 0 indicates the transition period or anomaly; the controller only handles... The sample was written into the dew point temperature sequence The sampling port number and timestamp are provided.

[0029] When in use, the sweep and steady-state discrimination separate the transition segment and the effective segment, and the dew point temperature sequence is effective; the attenuation weight makes the steady-state discrimination more sensitive to gate disturbances and does not slow down the polling; abnormal jumps ensure that a single channel failure will not prevent the entire round of coverage.

[0030] Furthermore, the backflow from the door opening is transformed into an event label that can be referenced by subsequent rules, and data frame verification is used to ensure that the label is consistent with the dew point sequence timing, so that the door area disturbance can be explicitly referenced in the hysteresis determination in step two.

[0031] The dew point temperature at the sampling port of the door zone may rise rapidly when the door is opened. If door magnetic sensor information and a description of the intensity of the change are lacking, the node library will keep the door zone at the top for an extended period, masking the true risks in corners and inside stacked structures. The controller collects door magnetic sensor status. With door opening time and from the door zone dew point temperature Structural gate region variation intensity When the gate magnet is in state And the intensity of gate region changes Exceeding the event threshold Write gate disturbance event flags Otherwise, write .

[0032] In this case, the intensity of the gate zone change is expressed in the following form:

[0033] Among them, the intensity of gate region change The value range is 0 to 1, used to uniformly characterize the jump intensity under different dew point baselines; gate zone dew point temperature The value range is -20 to 40 degrees Celsius, used to represent the dew point reading of the door zone passage; time interval The value range is 1 to 60 seconds, used to define the disturbance detection scale; door magnetic state. The value can be 0 or 1, used to indicate whether the door is open or closed; the duration of the open door is also specified. The value range is 0 to 600 seconds, used to describe the duration of a disturbance; event threshold. The value ranges from 0.02 to 0.3, and is used as the threshold for triggering gate disturbances and is written into the configuration table during the debugging phase.

[0034] The controller encapsulates the sampling port number, timestamp, dew point temperature, valid sampling marker, door magnetic status, and door zone disturbance event marker into an event frame, and adds a cyclic redundancy check (CRC) code. If the check fails, the frame is discarded and retransmitted. Once the retransmission reaches the upper limit, it is written to the communication error log. Event frames are stored monotonically in ascending order of timestamp to avoid playback misalignment. The door zone disturbance event marker distinguishes between short-term backflow and continuous risk, facilitating the use of different hysteresis rules in step two. Event frame verification and retransmission ensure consistency between the dew point sequence and the event tag. The installation and playback actions in this embodiment can be directly reproduced and used for debugging.

[0035] To address reading distortions caused by sampling pipeline blockage, valve misalignment, and probe leakage, a channel health diagnosis and fallback mechanism is established. This ensures that step one can still output dew point sequences and anomaly tags under abnormal conditions, preventing step two from losing its continuity criterion due to missing measurements. The high-risk node library relies on the continuity of the dew point sequence; if a node frequently misses measurements, it will be incorrectly downgraded or incorrectly prioritized. The controller maintains the channel health status for each sampling port. With channel maintenance event flags Channel health status The value is either 0 or 1, and it is a channel maintenance event flag. The value is either 0 or 1; the controller combines the steady-state judgment with the number of timeouts and the stability index. Anomalies are identified based on convergence speed and the reachability range of dew point readings; in case of anomalies, [further action is taken]. And place And write it to the event log.

[0036] The first layer of fallback is downgraded sampling, achieved by extending the cleaning time. The system reduces the pumping volume to minimize the impact of leaks; the second fallback is a bypass replacement, where the polling sequence skips abnormal channels and prompts for filter replacement or seal inspection; the third fallback is a backup mapping, where, after activating the reserved backup sampling port in the same area, the backup sampling port number is mapped to the original node identifier and the dew point temperature sequence continues to be generated. To ensure alignment in step two, step one specifies that each sampling port operates within one polling cycle. At most one sample valid flag can be output. The samples are used, and monotonically increasing timestamps are used as a unified index across steps; if the duty terminal needs to display continuously, it can... The samples are piecewise linearly interpolated while keeping the timestamps from rolling back.

[0037] When in use, channel health diagnosis treats blockages, leaks, and valve malfunctions as maintainable events to prevent distorted data from entering the node library; backup mapping maintains the continuity of the regional risk sequence and reduces the damage of missing measurements to the criteria; output alignment constraints ensure that step two can directly enter the condensation margin calculation chain based on the node identifier and timestamp.

[0038] Step 2: Under the constraints of the dew point temperature sequence, valid sampling markers, and gate disturbance event markers output in Step 1, generate the condensation margin and node risk level for each sampling node, and include nodes that meet the conditions of persistence and deterioration trend into the high-risk node library so that they can be directly called according to priority in Step 3.

[0039] The dew point temperature sequence from step one is transformed into a node state that can participate in condensation detection, avoiding the direct driving of subsequent actions by transition readings, channel anomalies, and short-term gate disturbances. Condensation is not triggered by a single reading, but rather by the relationship between dew point temperature and surface temperature continuously approaching a critical value over a certain period. If timestamps are not aligned and invalid samples are not removed first, node risk will be amplified by noise, leading to frequent fluctuations in the high-risk node pool. The controller uses the timestamp within the polling cycle as an index to read the sampled and marked as valid dew point temperatures from the dew point temperature sequence of each node, and uses the gate disturbance event marker as a disturbance window constraint to first generate the node dew point state, then merges it with the node's critical surface temperature to generate the node condensation margin state, and writes this state into the subsequent persistence and deterioration trend calculation queue.

[0040] The controller assigns a number to each node. Within a polling cycle, only valid dew point temperature readings marked as valid are accepted. If multiple valid readings exist within the cycle, the reading closest to the end of the cycle is taken as the representative dew point temperature of the node in that cycle. If no valid readings exist within the cycle, a missing reading flag is written based on the channel health status, and the dew point status of the node in the previous cycle is kept unoverwritten.

[0041] The controller then aligns the dew point temperature and the critical surface temperature to the same timestamp, forming a condensation margin with a gate disturbance window label, and finally outputs the node condensation margin state for calculating the node risk level.

[0042] Key surface temperature The value ranges from -10 to 60 degrees Celsius and is used to characterize the nodes. The instantaneous temperature of a surface prone to condensation can be estimated from the surface temperature measurement point or the virtual surface temperature. node dew point temperature The value ranges from -20 to 40 degrees Celsius and is used to characterize nodes. The air dew point temperature is output by the centralized dew point measurement unit in step one and filtered by the valid sampling marker; Condensation margin The value ranges from -30 to 60 degrees Celsius and is used to characterize nodes. The degree of approaching the critical condensation point, A negative value indicates that the conditions for condensation, where the surface temperature is below the dew point temperature, have been met. A positive value indicates that there is still a safety margin; When in use, the selection of valid sampling markers and the alignment of polling cycles ensure that the source of condensation margin is consistent with the timestamp, preventing false criticality in the transition section readings; the continuity of node status is maintained by missing measurement retention and missing measurement markers to prevent disruption by abnormal channels; and the gate area disturbance window label is used to distinguish between short-term backflow and structural risks in subsequent persistence criteria.

[0043] Furthermore, without increasing the number of surface temperature measurement points, an actionable key surface temperature source is provided for each sampling node, so that the condensation margin is not limited to a few fixed surface points.

[0044] The surface temperatures of cold bridge areas at wall corners, near-ground layers of pallets, and metal components of columns are often out of sync with the air temperature inside the warehouse due to the influence of thermal conductivity and airflow organization of the enclosure structure. If only air temperature is used to replace surface temperature, the condensation judgment will be reduced to relative humidity judgment, and the condensation mechanism will be lost.

[0045] First, the surface temperatures of a small number of surface temperature measurement points are read as anchor point temperatures. Then, based on the node mapping table formed during the installation and commissioning phase, each sampling node is mapped to one or two anchor points, and the critical surface temperature of the node is generated accordingly. When a node is not directly covered by an anchor point or the anchor point fails, the controller initiates wall thermal response estimation, updates the critical surface temperature of the node with the historical critical surface temperature and the current node dew point state, thereby maintaining the computability of condensation margin.

[0046] To ensure each sampling node All have available key surface temperatures It is recommended to add the following rule-based description to the instruction manual (to avoid vagueness): Establish a node mapping table during the installation and debugging phase, and map each sampling node... Associated with one or two surface temperature measurement anchors The surface temperature measurement anchor point Fixed installation on surfaces prone to thermal bridging, such as corners, columns, floor cooling zones, or the inside of doorways. The node mapping table shows each pair of nodes. — Anchor point Configure mapping weights and satisfy ,and The range of values ​​is The weights are fixed during debugging based on the geometric distance from the node to the anchor point, the dominant airflow direction, and the occlusion relationship. A clear calculation formula is provided accordingly.

[0047] Key surface temperature Value range: -10 to 60 degrees Celsius; function as a node. The input is the surface temperature of the area prone to condensation. It is obtained through mapping calculation or estimation update; anchor point surface temperature. Value range: -10 to 60 degrees Celsius; function as an anchor point. The measured surface temperature was collected from the attached surface temperature measuring points; mapping weights : Value range 0 to 1; function is the scaling factor that maps anchor point temperature to node temperature; fixed by debugging. Related anchor point set. : for nodes The associated set of anchor indexes; its function is to limit the range of anchors participating in the calculation, and it usually contains 1 or 2 anchors.

[0048] The node mapping table is generated during the installation phase by on-site surveying. The survey content includes the sampling port location, adjacent anchor point numbers, relative position coefficients, and typical airflow direction. The controller maps the nodes according to the mapping table. The key surface temperature is determined as a weighted combination of anchor point temperatures, with the weights given by relative position coefficients and locked during the commissioning phase. The wall thermal response estimation uses a first-order heat dissipation inertia update: when the node... When the anchor point is missing, the controller updates the critical surface temperature by combining the critical surface temperature of the node in the previous cycle with the dew point state of the node in the current cycle. The dew point state is used to characterize the impact of changes in water vapor content near the node on boundary layer heat transfer. The updated critical surface temperature is still aligned with the polling cycle timestamp.

[0049] To avoid conceptualization, the engineering implementation of wall thermal response estimation can choose three paths: First, use linear interpolation and first-order inertial updates, which is suitable for continuous surfaces such as wall corners and the ground; second, use monotonic cubic interpolation to complete the anchor point temperature along the wall height direction, which is suitable for multi-level storage locations; third, use handheld infrared thermometers to generate a node mapping table and solidify weights during empty warehouse debugging, which is suitable for warehouses with complex cold bridge distribution.

[0050] When anchor point temperature is briefly missing or node mapping table is temporarily unavailable, to avoid To address the issue of fracture, it is recommended to provide a first-order thermal inertia update as a specific form for estimating the wall's thermal response:

[0051] thermal inertia coefficient The value ranges from 0.6 to 0.98; its function is to characterize the hysteresis and inertia of the surface temperature of the building envelope. A larger value indicates a greater reliance on the previous round of surface temperature memory. This value can be obtained through configuration adjustments or by selecting based on material experience. Polling cycle The value range is 30 to 600 seconds; its function is the same as the polling in step one, making... Aligned with the dew point temperature timestamp.

[0052] The fallback rule for missing anchor point measurements: When a certain anchor point... If the surface temperature measurement point is offline, then... Remove the anchor point and then remove the remaining anchor points. Perform normalization; if If all anchor points are missing measurements, then... A missing surface temperature marker is written to ensure that step two can maintain the continuity of condensation margin without erroneously exiting the decision process.

[0053] In use, a small number of anchor point temperatures, combined with a node mapping table, can cover multiple sampling nodes, avoiding the stacking of surface temperature measurement points; the wall thermal response estimation can still provide key surface temperature inputs when anchor points are missing or cold bridges migrate, making condensation margin continuously available; the equivalent implementation path provides feasible options for different warehouse structures, avoiding limiting surface temperature acquisition to a single hardware form.

[0054] Furthermore, the single condensation margin is transformed into a persistent risk, and short-term backflow is suppressed within the gate disturbance window, thus focusing the high-risk node pool on the structural risks caused by cold bridging and moisture retention. Condensation risk often manifests as a continuous decrease in condensation margin over several polling cycles or repeated fluctuations in the low margin range; if only... The instantaneous threshold determination equates occasional disturbances with structural risks. The controller maintains a risk score for each node, which is obtained by accumulating the gap amount below the safety margin exponentially within a time window. A suppression factor is applied within the gate disturbance window to limit the disturbance of the node pool to short-term backflow within the gate area within the event window. Subsequently, the controller adopts a delayed entry rule. When the risk score exceeds the entry threshold, the node is added to the high-risk node pool and assigned a risk level. Only when the risk score falls below the exit threshold and the exit conditions are continuously met is the node allowed to be downgraded or removed.

[0055] Inbound threshold With outbound threshold and satisfy .when And the corresponding sampling valid marker At that time, the node Add to the high-risk node library; when the node is already in the library and Degradation or removal is allowed if the requirements are met for a certain number of consecutive polling cycles. Inbound threshold. The value ranges from 1 to 20, and is used as the upper threshold for nodes to enter the high-risk node database; the lower threshold for nodes to leave the database. The value ranges from 0 to 15, and is used as the lower threshold for nodes to leave or exit the high-risk node database; valid sampling marker. The value range is 0 or 1, and it is used to make the threshold judgment based only on valid samples.

[0056] Controller setting safety margin As a safety boundary for condensation margin, and within the time window Calculate a risk score for each node. Meanwhile, the weights within the gate area disturbance window are adjusted by marking the gate area disturbance event, so that backflow into the gate area is recorded but does not directly increase the risk level of non-gate area nodes; The risk score calculation uses an index kernel to highlight recent changes and masks invalid samples with valid sampling markers, as detailed below:

[0057] Where: Node risk score The value ranges from 0 to 30, and is used to represent nodes. The cumulative intensity of insufficient condensation margin within the window is used for warehousing, grading, and sorting; safety margin. The value ranges from 1 to 5 degrees Celsius and is used to specify the safety boundary for condensation allowance. The larger the value, the more conservative the protection against cold bridging and moisture retention. Condensation margin The value ranges from -30 to 60 degrees Celsius and is used to characterize nodes. The critical proximity is used to calculate the gap; time step The value ranges from 1 to 10 seconds, and is used to define the time interval for backtracking sampling within the window; the attenuation coefficient. The value range is from 0.1 to 1, used to define the decay rate of the exponent against historical terms; window length The value ranges from 3 to 24, defining the number of historical items participating in the accumulation, which, together with the polling period, determines the time scale of risk accumulation; valid sampling markers. The value is 0 or 1, used to mask invalid samples so that invalid dew point readings are not included in the risk score; gate disturbance event flag. The value is 0 or 1, used to identify backflow disturbance windows in the door zone, and to suppress misjudgments of structural risks within the disturbance window; disturbance suppression coefficient. The value ranges from 0 to 1, and is used to specify the degree of weakening of the risk score weight within the disturbance window. The larger the value, the less sensitive it is to backflow in the gate area; In addition, node classification and sorting are required. The controller prioritizes nodes based on their risk scores, with node risk levels divided into discrete levels corresponding to risk score ranges, to select the intensity of the targeted dew reduction action in step three. Sorting can be maintained using a min-heap. Nodes, to avoid the computation time consumed by full sorting, When the number of execution branches is set to an integer in the settings table.

[0058] Node risk level The set of values ​​is ; where 1 represents mild, 2 represents moderate, and 3 represents severe; by Obtained through segmented mapping. The action intensity level is directly taken as... No new rank symbols will be introduced to avoid terminology drift. It can be written as:

[0059] Node risk level : Set of values Its function is to drive the intensity of the action item in step three; grading threshold. : Range of values to Its function is to define the boundary between mild and moderate severity; grading threshold. : Range of values to Its function is to demarcate between moderate and severe cases.

[0060] When used, the risk score incorporates the persistence of low margin and the trend of deterioration into the same dimension to avoid the jitter caused by instantaneous thresholds; the gate area disturbance event marker and the disturbance suppression coefficient together limit the impact of backflow on the structural risk assessment, making the node library more focused on cold bridges and stacking moisture retention.

[0061] Furthermore, the high-risk node library and its derived information are encapsulated into processing inputs that can be directly read in step three, and the interface remains stable during missing tests, disturbances, and mapping changes, so as to avoid ambiguity introduced in step three due to changes in the input structure.

[0062] Step three requires selecting directional decondensation branches and airflow paths based on node priority sequences. If the node database only provides node numbers but lacks condensation margin, risk score, and disturbance label, it is difficult to unify the action intensity and exit conditions. Simultaneously, stacking positions may be adjusted during warehousing operations, and the node mapping table may be updated; it is essential to ensure that the update process does not cause node identifier drift. The controller encapsulates the condensation margin, risk score, node risk level, door disturbance window label, and channel health status of each node into node entries, and forms a high-risk node database with all existing node entries. When the node mapping table is updated, the controller first calculates the condensation margin corresponding to the old and new mappings in parallel within a polling cycle and compares their consistency. Only after passing the consistency check is the controller switched to the new mapping to avoid instantaneous node identifier drift.

[0063] The node entries in the high-risk node library must include at least the node number, node risk level, node risk score, current condensation margin, the most recent entry timestamp, and the most recent gate area disturbance window label. For nodes with missing data, the node entry retains the condensation margin and risk score of the previous cycle and adds a missing data marker, allowing step three to choose a conservative action or wait for the next update. For gate area nodes, the node entry also retains the door opening duration so that step three can set the duration when entering the gate area compensation process. In terms of interface constraints, the controller stipulates that a node library snapshot is published once at the end of each polling cycle and frozen until the end of the next cycle. Step three only consumes the snapshot and does not directly read the intermediate state, thereby avoiding the action process being rewritten midway. The publication can use the same event frame format and add a cyclic redundancy check code. If the check fails, step three maintains the snapshot of the previous cycle and enters a conservative waiting state.

[0064] When in use, the structured output of node library entries allows step three to select the intensity of targeted dew reduction actions based on the node risk level, and to set exit conditions based on the dew margin and risk score; mapping switching consistency verification avoids node identifier drift caused by stacking adjustments and can be traced across cycles; snapshot publishing and verification retransmission stabilize the input of step three and avoid ambiguity in the action chain due to communication jitter.

[0065] Step 3: Using the high-risk node database snapshot output in Step 2 as the action entry point, subject to condensation margin. Hard constraints, low dew point dry air branch, pulsed penetration circulation, return air path switching and door area compensation are treated as a single chain of treatment sequences, and each treatment is written into a traceable action frame for review in step four.

[0066] Because step two defines the high-risk node database snapshot as including node number, node risk level, and node risk score. Condensation margin Gate disturbance event marker Because of the structured input of timestamps, adding a second set of trigger sources in step three would cause the processing logic to fork, reducing the interpretability of the node library driver.

[0067] The controller reads a snapshot of the high-risk node database after each polling cycle and first scores the nodes according to their risk. The node priority sequence is obtained by descending order, and then the dew margin is read for each node. With safety margin The relative relationship will The node was identified as a candidate for disposal and will The minimum value is used as the hard constraint boundary for this round of processing; then, the controller maps the processing candidates to three types of action paths based on the node type identifier: corner cold bridge path, stacking penetration path, and door area compensation path, and generates a single-chain rule for each path with entry condition, hold condition, and exit condition, with the hold condition based on condensation margin. Based on the upward trend, the exit condition is... The condensation margin is determined by the continuous satisfaction of the target. Hard constraints are implemented as the start and end boundaries of the action sequence.

[0068] Each candidate node is assigned a disposal entry, which includes the node number, action path type, action duration limit, action intensity level, return air path target, and door area compensation mark. The action intensity level is directly mapped from the node risk level, avoiding the introduction of new scoring symbols in step three. The action duration limit is jointly limited by the polling cycle and the node entry timestamp, ensuring that the action does not lose input consistency by crossing the next snapshot. The return air path target is fixedly mapped from the node's location to the corresponding return air vent or return air duct inlet, avoiding ambiguity about where the air is sent.

[0069] When in use, the high-risk node database snapshot is the sole entry point for the disposal sequence; disposal can be derived from the node entries to determine the exposure margin. and node risk scoring Safety margin As hard constraint boundaries for entry, maintenance, and exit, the handling intensity does not deviate from the condensation mechanism; the structured output of action entries provides fields for subsequent action frame encapsulation. Furthermore, by lowering the local air dew point instead of relying on whole-warehouse deep dehumidification, the condensation margin is increased. Return to safety margin Above. Among them, the dry air must adhere to a thin layer of air near the surface where condensation easily forms, and form a closed loop with the return air path to avoid ineffective diffusion of dry air within the storage.

[0070] Key surface temperature of cold bridge zone at corner Often lower than the air temperature inside the warehouse, if only the overall dehumidification output is increased, the rate at which the air dew point decreases is affected by the warehouse volume, and condensation may occur first at the boundary layer. However, attached dry air flushing can form a low dew point air film near the surface, lowering the node dew point temperature. First decrease, thereby increasing condensation margin. .

[0071] For entries mapped to corner cold bridge paths, the controller first switches the low dew point dry air branch to the corresponding corner nozzle and adjusts the nozzle direction to tangentially supply air along the surface, creating a continuous air film at the junction of the corner and the column. Then, the controller synchronously adjusts the return air path target, guiding the humid air near the corner to the return air inlet, forming a closed loop of dry air supply – moisture recovery – centralized dehumidification – dry air resupply. Subsequently, the controller receives the condensation margin of the corner nodes according to a polling cycle. Updated, when condensation margin is increased Still less than the safety margin While maintaining the attached flushing and keeping the return air path unchanged, the condensation margin is increased. The air volume is gradually reduced only when the safe zone is entered and the exit conditions are continuously met, in order to avoid boundary layer rebound.

[0072] The low dew point dry air branch is a separate branch from the dehumidification equipment's outlet side. A slit nozzle or guide vane is installed at the end of the branch, and a fixed distance is maintained between the nozzle and the corner surface to form a repeatable adhesion flow field. The return air path is switched by an electric air valve. The switching sequence is to first open the corner return air valve and then limit the opening of the non-target return air valve to ensure that moisture near the corner is recovered first. At the same time, the controller limits the variation range of the overall dehumidification output to prevent abrupt changes in the relative humidity of the entire warehouse. This implements the intention of coordinated temperature and humidity atmosphere control, which prioritizes local treatment and gradually follows the overall warehouse, at the execution level.

[0073] As a supplement: For each controlled execution variable (damper opening, fan speed, main air branch valve position), set the maximum change step size within a unit polling cycle. This causes the control commands to change gradually. The target value and allowed step size are explicitly carried in the action frame field, and the controller side splits them into multiple action frame sequences according to the step size and issues them.

[0074] Among them, execution volume Value range: 0~1; used to characterize the opening degree of the damper or the normalized fan speed. Values ​​are obtained from controller setpoints and actuator feedback values, with a maximum variation step size. The value ranges from 0 to 0.2; it is used to limit the change in the amount of execution within a polling cycle to avoid sudden changes.

[0075] And provide the splitting rules (no complex algorithms are needed, just implementable ones):

[0076] Target execution volume Values ​​range from 0 to 1; the function is the target valve position or target gear position given by the action entry. (Symbol function) Its function is to indicate the direction of change, and its value is... , , Cutting operator Its function is to limit the execution volume to The specific form is .

[0077] When using this method, the attached flushing system first establishes a low dew point air film near the surface prone to condensation, thus maximizing condensation allowance. The recovery does not depend on the complete replacement of the entire air in the warehouse; the return air path coupling brings the moisture to a controllable return air inlet, reducing the ineffective diffusion of dry air and maintaining the traceability of the atmosphere path.

[0078] Furthermore, by combining short-term penetrating air supply with intermittent recovery to form a pulsed penetrating cycle, the water vapor exchange between the inside and outside of the stack is transformed from slow infiltration to controlled replacement.

[0079] Nodal dew point temperature of representative points inside the stack The persistently high level of airflow, coupled with the lack of significant external return air nodes, indicates that the moisture retention layer is blocked by the stacking structure and packaging. If only the overall circulating air volume of the warehouse is increased, the airflow will flow around the stack, resulting in slow changes in the internal state. High-risk nodes will remain in the warehouse for a long time, causing waste disposal resources to be occupied.

[0080] For processing entries mapped to stack penetration paths, the controller first enables the penetration air supply interface and selects the corresponding tray base or side air inlet. Then, it executes a pulsed penetration cycle according to three segments: supply-retention-recovery. In the supply segment, low dew point dry air is supplied into the stack in the penetration direction. In the retention segment, the supply valve is closed and the pressure difference between the inside and outside of the stack is maintained to dissipate. In the recovery segment, the target return air path is opened so that the moisture inside the stack is guided to the return air inlet. This three-segment cycle can be executed multiple times within a polling cycle, but the valid sampling marker must be checked before each execution. Whether it is effective, avoid blindly increasing the intensity of the action under conditions of insufficient measurement; when the condensation allowance inside the stack is... When the recovery speed is insufficient, the controller prioritizes extending the recovery section rather than extending the feeding section, thereby making the exhaust of moisture the main cause and effect, and avoiding fluctuations in the atmosphere inside the storage caused by prolonged blowing.

[0081] The penetrating air supply interface can be either a pallet base air cavity interface or a side-inserted air gun interface. The pallet base air cavity interface is suitable for standardized pallet storage locations, while the side-inserted air gun interface is suitable for boxed stacking. Return air guidance is achieved by opening a set of return air valves near the stack and limiting the opening of the far-end return air valve, making the moisture recovery path short and definite. To ensure that the cycle can be implemented, the controller writes the three cycle segments into the task scheduling table. The task scheduling table is bounded by the polling cycle and is bound to the snapshot timestamp of the high-risk node library. If a new snapshot is received during the execution of the scheduling table and the node has exited the disposal conditions, the next cycle will stop after the current cycle ends and the process will gradually fade out.

[0082] During use, the pulsed penetration cycle transforms the stagnant layer inside the stack into a controllable replacement process, controlling the dew point temperature at the nodes inside the stack. It can be guided to the return air inlet and descend along with the recovery section; cycle time scheduling and effective sampling marking The linkage ensures that the penetration action is only executed when the input is reliable, thus avoiding miscontrol due to missing measurements; the causal chain of the recovery section is extended first to strengthen the main role of moisture discharge and reduce passive fluctuations in the atmosphere inside the storage.

[0083] Furthermore, use gate disturbance events as markers. As the trigger condition, and based on the duration of door opening. With condensation margin By jointly limiting the duration, we can prevent gate area compensation from evolving into a long-term whole-database strategy.

[0084] Step 1 has been marked with a gate zone disturbance event. To separate backflow from normal fluctuations, step two writes this marker to the node library snapshot. Therefore, step three should limit door zone compensation to event-driven short-chain actions, and ensure that the dry air buffer and rapid dehumidification of the door zone can be reviewed in step four. When the door zone disturbance event is marked in the node library snapshot... or condensation margin of gate nodes Continuously falling into the safety margin At this point, the controller enters the door zone compensation process, first opening the dry air nozzles in the door zone to form a dry air buffer zone, then switching the door zone return air path to guide the backflowing moisture into the return air inlet, and then adjusting the response based on the door opening duration. Set the compensation window to maintain airflow and dehumidification even after the door is closed, until condensation is achieved in the door area. The exit conditions are restored and continuously met. Throughout the compensation process, the controller encapsulates each valve position switch and fan speed change into an action frame and sends it to the actuator. The action frame requires the actuator to return an acknowledgment frame. The acknowledgment frame contains a sequence number and a cyclic redundancy check code. If no acknowledgment is received, the signal is retransmitted up to the maximum number of retransmissions. If the maximum number of retransmissions is exceeded, the signal enters the fallback chain.

[0085] The action frame field must include at least the node number, action type, target valve position, target fan speed, hold time, sequence number, and cyclic redundancy check code. The confirmation frame field must include at least the sequence number, execution result code, and cyclic redundancy check code. The first layer of the fallback chain maintains the previous stable valve position and continues to collect condensation margin in the gate area. The second layer is to switch to a conservative return air path and limit the range of dehumidification output changes. The third layer is to prompt the terminal to manually check the status of the door zone damper and door magnet. In order to align with the node library snapshot, the controller binds the action frame timestamp and the snapshot timestamp and writes them to the log, so that step four can review the complete chain of opening the door-compensation entry-compensation exit according to the timeline.

[0086] When in use, gate compensation is marked by gate disturbance events. Triggered and subject to door opening duration With condensation margin The common constraints ensure that the compensation process can be controlled to enter and exit; the sequence numbers and cyclic redundancy check codes of the action frames and confirmation frames make the execution chain verifiable, reducing the failure of handling caused by valve position not being in place; the layered fallback chain keeps the single chain of handling from splitting when communication or actuator is abnormal, and provides clear direction for manual inspection.

[0087] Step 4: During and after the treatment process in Step 3, continue to collect the dew point temperature sequence and recalculate the condensation margin. With a safety margin To verify the exit conditions for hard constraints, the action frame log is used to complete the gradual exit and event review, and the data is written back to the high-risk node library level and strategy parameters. The next handling will directly use the same terminology system and timestamp.

[0088] Furthermore, the condition of no condensation is transformed from a subjective judgment into a verifiable exit condition, avoiding boundary layer rebounds immediately after the action is completed, and also preventing the action from being maintained indefinitely and consuming execution resources. This is based on the effective sampling markers from step one. As a reliable constraint for the recovery, and with the condensation margin in step two... With safety margin As an exit boundary, the decision chain is fully aligned with the preceding steps. The targeted dew point treatment focuses on the cold bridge area at the corner, the representative point inside the stack, and the return air path area at the door. The dew point temperature and surface temperature at these locations may change rapidly during the treatment. If the treatment is exited based solely on a single reading reaching the threshold, a short-term drop-and-rise cycle may easily occur after the door is reopened or the return air path is restored, leading to fluctuations in the high-risk node library level.

[0089] At the end of each polling cycle, the controller reads the dew point temperature sequence updated in step one, and combines it with the sampled valid markers. After removing invalid samples, the condensation margin at each node in this round is recalculated. The controller establishes an exit candidate set for the set of nodes covered by the disposal entry; the controller only exits if each node in the exit candidate set satisfies the conditions within a consecutive number of polling cycles. And the corresponding sampling valid marker Only when the gate disturbance event is marked as exitable, otherwise the processing continues and data retrieval continues; when the gate disturbance event is marked When the door zone disturbance window protection state is entered, the number of consecutive satisfactions is recalculated in the next polling cycle after the door zone disturbance window ends, thus avoiding misjudging the short-term improvement during the backflow period as a stable recovery.

[0090] The window for consecutive satisfaction counts is given by the configuration table and bound to the polling cycle, allowing the exit decision to be directly implemented in engineering. For nodes with missing data, the controller does not replace satisfaction with missing data, but maintains the previous exit candidate set state and writes a missing data flag until the node recovers valid data collection. For stacked nodes, the controller requires that the exit candidate set simultaneously cover the representative point inside the stack and its adjacent return air path area nodes, and only when both are satisfied will the exit candidate set be valid. Only then is exit allowed, thus binding the completion of internal replacement and external recycling under the same exit condition.

[0091] When using it, the exit judgment is as follows: and The relationship is the unique boundary, ensuring that the exit condition does not deviate from the condensation mechanism; valid sampling markers. Constraint exit determination only proceeds when data is reliable, avoiding false exits due to missing tests; gate disturbance event marking. The participation in the exit determination ensures that short-term backflow in the gate area will not induce a false stability in the exit determination.

[0092] Furthermore, the exit process was changed from a one-time shutdown to a gradual downgrade in a fixed sequence to ensure that the switching between dry air branches, local circulation and normal dehumidification does not introduce new atmospheric disturbances.

[0093] Directional dew reduction involves switching between low-dew-point dry air branches and return air paths, as well as the timing of stacking and penetration. If the exit sequence is lacking, the return air may resume before the dry air continues, or the dry air may stop before the return air remains restricted, leading to short-term localized moisture accumulation and impacting... The stability is maintained. When the candidate set for exiting is determined to be in an exitable state, the controller does not immediately restore the entire warehouse to normal operation, but instead enters a gradual exit sequence: first, the low dew point dry air branch is gradually downgraded from the target node specified in the treatment item and eventually shut down, while keeping the return air path target unchanged to continue recovering residual moisture; then, the pulse-type penetration cycle feed section is stopped, leaving only the recovery section for a period of time to release the residual moisture inside the stack; then, the return air path valve position is gradually restored from the target return air inlet to the normal return air valve position, and recovery is continued during the restoration process. To prevent boundary layer rebound, the system eventually restores normal dehumidification and normal circulating air settings, and writes the exit completion marker to the action frame log. The controller still sends out action frames for each gradual retreat action and waits for confirmation frames. The confirmation frame contains a sequence number and a cyclic redundancy check code to match the action frame. If no confirmation frame is received within the specified time, it enters a conservative hold state: maintaining the previous stable valve position and the previous stable fan speed while continuing to backtrack. At the same time, the terminal marks the fade-out action as unconfirmed; when multiple unconfirmed actions occur consecutively, the controller will pause the fade-out sequence at the current stage to avoid further uncertainty caused by continuing to switch when the actuator status is unclear.

[0094] When in use, the gradual exit sequence ends the dry air, penetration and return air in a fixed order to avoid localized re-humidification caused by shutdown; the closed loop of action frame and confirmation frame ensures that each gradual exit action has proof of execution; the conservative retention freezes the valve position and gear position when confirmation is missing to avoid changes in airflow path caused by unclear actuator status.

[0095] Furthermore, the process of handling a single incident is transformed from a completed task into a reusable experience. This involves creating a unified event record that includes the triggering cause, the evolution of the event, the handling items, and the exit process, and providing rules for labeling the causes. This allows subsequent similar scenarios to be directly referenced without altering the terminology of the claims.

[0096] The updating of the high-risk node database's level and the self-correction of its strategy parameters rely on a consistent description of why it was triggered, where it was triggered first, and how it exited. Without a unified event log, subsequent actions must start from scratch, making it impossible to utilize existing action paths and explaining the reasons for changes in the node database's level. The controller creates an event log at the start of each action, binding the trigger snapshot timestamp, the set of triggering node numbers, and the node risk score at the time of triggering. Condensation margin at the time of triggering Gate disturbance event marker With door opening time During the processing, action frame logs and confirmation frame results are continuously appended, and data is collected back according to the polling cycle. Sequence fragments; upon exit completion, an exit completion marker and a fade-out sequence stage marker are written to form an event loop.

[0097] Subsequently, the controller writes a cause label to the event according to the labeling rules: when a disturbance event in the door zone of the handling window is triggered, a cause label is added. When the gate node is at the top of the node priority sequence, it is marked as "open gate backflow"; when the condensation margin of the representative point inside the stack... When a node in the adjacent return air path zone enters the low margin zone before the node in the adjacent return air path zone and consistently scores higher, it is marked as "stagnant moisture." When a node in the corner cold bridge zone is repeatedly triggered over a long period and does not overlap with the door zone disturbance window, it is marked as "enhanced cold bridge." Event records are written to local storage with structured fields and can be synchronized to the duty terminal. The fields include event number, timestamp, node set, etc. Snapshot Trajectory summary and The event log includes action frame sequence and confirmation results, channel health status markers, fade-out phase markers, and exit completion markers. To prevent record loss, event logs are written to a temporary area first and then committed. Only when the commit is successful is the log marked as queryable. If the commit fails, the temporary area is retained and the log is retried in the next query cycle. Once the retry limit is reached, a manual export prompt is displayed.

[0098] When using it, triggering, handling, exiting, and recollecting evidence will be included in the same event log for retrieval; the cause labeling rules are based on... , , and Based on the temporal sequence, subjective attribution is avoided.

[0099] The conclusions of the event review are then written into the level and strategy parameters of the high-risk node library to ensure that the next action follows the correct action path. Instability issues arising from self-correction are avoided through parameter version management and fallback chains.

[0100] The frequency of door area operations, stacking patterns, and distribution of cold bridges vary among different warehouses, and a fixed safety margin is required. Disturbance suppression coefficient In some scenarios, the polling cycle may cause the node library to remain at the top for a long time or exit too slowly; however, if the parameters are adjusted without boundaries, the condensation margin determination may lose consistency.

[0101] After the event logging is closed, the controller rewrites the risk level of the high-risk node database based on the cause tags and handling process results: for nodes that repeatedly appear at the top of the priority sequence during an event and have a long exit time, their risk level is increased and their priority attention mark in the node mapping table is upgraded; for door zone nodes that frequently trigger within the door zone disturbance window but exit quickly, their risk level remains unchanged, but the trigger sensitivity of the door zone compensation process is increased; for nodes triggered by stacked moisture retention, the pulse-type penetration cycle cycle configuration is adjusted first without increasing the overall dehumidification output variation. Furthermore, at the parameter self-calibration level, the controller only allows adjustments within preset boundaries to maintain a safety margin. Disturbance suppression coefficient The polling cycle is set, and each adjustment is marked with a parameter version number. The adjustment takes effect in the next polling cycle. If abnormal fluctuations or frequent exit judgments occur in the node library within several polling cycles after the adjustment takes effect, a version rollback is triggered to restore the parameters of the previous version and the rollback reason is written to the event log.

[0102] Parameter self-calibration can employ two engineering paths: bounded grid search or bounded stepwise search. , Enumerating within the discrete candidate set of the polling cycle, selecting candidate versions that reduce the number of repeated triggers and the number of repetitions in the exit phase during event review; bounded stepwise search adjusts only one parameter after each event, prioritizing the adjustment of... To mitigate the impact of gate disturbance windows, the polling cycle is adjusted to balance coverage and computational load. Finally, the safety margin is fine-tuned as necessary. This involves changing the hard constraint boundaries. Both paths require that each adjustment be written to a version number and that a rollback point be retained, which can be queried in the terminal.

[0103] When used, the level writeback transforms the event review conclusions into level evolution conclusions usable by the node library, allowing the next action to quickly lock the action path without adding new symbols; parameter version management and fallback chain can ensure that self-correction is within the boundary and can be restored, without policy drift, and parameter adjustment can have a clear process rather than a vague conceptual description.

[0104] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0105] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0106] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0107] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0108] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A smart tea storage temperature and humidity atmosphere coordinated control system, implemented using dehumidification equipment, characterized in that: This includes setting up sampling ports in the warehouse door area, cold bridge area and inside the stack, and using an air pump and multi-way valve group to periodically poll and send air samples into a centralized dew point measurement unit to obtain the dew point temperature sequence, and collecting door magnets to generate door area disturbance event markers. The condensation margin is obtained by comparing the dew point temperature sequence with the key surface temperature. A high-risk node library is generated and classified according to the condensation margin and hysteresis rules. When the high-risk node database triggers the disposal conditions, a combination of targeted dew reduction actions is performed on the target node according to the classification, including dry air flushing of the cold bridge surface by the low dew point dry air branch, pulsed penetration cycle for nodes inside the stack, and door area compensation process when the door area disturbance event is triggered. The system continuously recovers dew point temperature and recalculates condensation margin. When the exit conditions are met, it exits the directional dew reduction action combination according to the gradual withdrawal strategy, records the event and handling action sequence, and updates the strategy parameters of the high-risk node library.

2. The intelligent tea storage temperature and humidity atmosphere coordinated control system according to claim 1, characterized in that: Sampling ports are arranged in zones according to door area, cold bridge area at the corner, return air path area, and representative points inside the stack, and are assigned node numbers. The multi-way valve group polls each node number according to a preset cycle. Before switching nodes, the air pump performs purging and delayed sampling, and generates valid sampling marks to correspond to the dew point temperature sequence.

3. The intelligent tea storage temperature and humidity atmosphere coordinated control system according to claim 2, characterized in that: The door zone disturbance event flag is generated by the dew point temperature sequence of the door magnetic state, door opening duration, and door zone node number. The controller calculates the door zone change intensity based on the dew point temperature difference between adjacent polling cycles. The door zone disturbance event flag is set when the door magnetic state is open, the door zone change intensity exceeds the threshold, and the minimum duration of the door opening is reached.

4. The intelligent tea storage temperature and humidity atmosphere coordinated control system according to claim 3, characterized in that: The critical surface temperature is collected by anchor temperature measurement points set at the corners of the walls, columns and the ground, and a node mapping table is established. The node mapping table includes the node number, anchor number, mapping weight and update timestamp. The mapping weight is determined according to the distance between the node and the anchor. When the anchor is missing, the critical surface temperature is updated according to the wall thermal response estimate.

5. The intelligent tea storage temperature and humidity atmosphere coordinated control system according to claim 4, characterized in that: A risk score is generated based on the condensation margin and a hysteresis rule is applied. The risk score is obtained based on the cumulative decay over time when the condensation margin is lower than the safety margin. The entry process uses a combination of the entry threshold and the minimum duration for determination, while the exit process uses a combination of a stable window that is below the exit threshold and continuously meets the valid sampling markers, and the node level is kept from bouncing back between entry and exit.

6. The intelligent tea storage temperature and humidity atmosphere coordinated control system according to claim 5, characterized in that: The high-risk node database classifies and sorts the nodes entering the database into three levels: Level 1, Level 2, and Level 3, and records the level field for each level. The level is determined based on the deterioration trend of condensation margin, recovery speed, and number of repeated triggers. The sorting results are output as a list of target nodes and a treatment priority. The list of target nodes is used as the input for selecting a combination of targeted decondensation reduction actions.

7. The intelligent tea storage temperature and humidity atmosphere coordinated control system according to claim 6, characterized in that: The low dew point dry air branch consists of a dry air source, a branch air duct, an electric branch valve, and an attached air outlet component. The attached air outlet component is fixed to the critical surface temperature corresponding position of the cold bridge area in the corner of the wall. The controller starts and stops the electric branch valve one by one according to the target node list and maintains the preset flushing time, while recording the branch start and stop sequence and writing it into the handling action sequence.

8. The intelligent tea storage temperature and humidity atmosphere coordinated control system according to claim 7, characterized in that: The pulse-type penetration cycle is executed on the corresponding node of the representative point inside the stack. The controller alternately switches the penetration fan speed and return air path valve position according to the preset pulse cycle. After each switch, the sampling port of the representative point inside the stack is re-polulated to update the dew point temperature sequence. At the same time, the pulse cycle and valve position switching sequence are written into the treatment action sequence.

9. A tea intelligent storage temperature and humidity atmosphere coordinated control system according to claim 8, characterized in that: The door zone compensation process is executed during the period when the door zone disturbance event flag is set. The controller opens the door zone preset dry air buffer and adjusts the return air guiding component to make the airflow in the door zone enter the return air path area. At the same time, it sets the upper limit of the change step size for the dehumidification equipment execution quantity, branch valve position and fan speed. After the door magnet changes from open to closed, it continues to execute the preset delayed dehumidification process and then removes the door zone disturbance event flag.

10. The intelligent tea storage temperature and humidity atmosphere coordinated control system according to claim 9, characterized in that: The exit conditions include that the condensation margin of all target nodes in the high-risk node library continuously meets the safety margin and the sampling validity markers are continuously valid. The gradual exit strategy exits step by step in the order of low dew point dry air branch, pulsed penetration circulation, and normal dehumidification. During the exit process, the interlocking of the gate area disturbance event markers remains valid. After the interlocking is released, the event and the sequence of handling actions are associated and archived, and the strategy parameter version number is updated.

Images