An intelligent archive warehouse monitoring system based on the Internet of Things

Abstract

The present invention relates to the field of intelligent monitoring technology and discloses an intelligent archive warehouse monitoring system based on the Internet of Things. The system comprises: a plurality of edge sensing units deployed in the warehouse, which dynamically learn personalized stable-state baselines of environmental parameters, monitor parameter drift in real time, and trigger local perturbation events; the units achieve collaborative judgment through perturbation wake-up signals, and report events of interest to a central platform only when preset conditions are met. The present invention forms a full-spectrum perception of slow-changing and sudden risks through baseline learning and collaborative wake-up mechanisms of edge units, combined with instantaneous event template matching and targeted enhanced monitoring; dynamic health fingerprint analysis based on perturbation resonance characteristic spectra can proactively warn of stability degradation trends of the archive microenvironment, significantly improving the intelligence level and preventive protection capabilities of the monitoring system.

Description

Technical Field

[0001] The present invention relates to an intelligent archive warehouse monitoring system based on the Internet of Things, belonging to the technical field of Internet of Things environment monitoring. Background Art

[0002] Currently, environmental monitoring in archive warehouses primarily relies on a centralized IoT monitoring architecture. This involves deploying a large number of sensors (such as those for temperature, humidity, gas, and vibration) to collect environmental data and upload it to a central platform for unified analysis. While this model can achieve full coverage, its actual operation exposes inherent contradictions between information redundancy and superficial understanding, specifically manifested as follows:

[0003] 1. Sensor overload and ineffective attention: Massive sensors continuously generate data, most of which is redundant information under normal conditions. This results in ineffective consumption of the central platform's computing power. However, subtle abnormal signals that truly require attention are often ignored or delayed due to the data overload. For example, persistent small drifts in temperature and humidity in a local area are often treated as noise by the system because they do not exceed the global fixed threshold.

[0004] 2. Mismatch between static thresholds and dynamic risks: Existing systems rely on preset rigid thresholds to identify anomalies, making it difficult to adapt to dynamic risk contexts caused by factors such as archival materials and seasonal changes. For example, the same humidity fluctuation has significant differences in the impact on ordinary paper and precious silk archives, but existing technologies cannot conduct differentiated assessments.

[0005] 3. Lack of correlation and insufficient early foresight: Risk assessment is usually based on independent analysis of a single parameter and lacks the ability to identify the synergistic effects of multiple factors (such as the accelerated deterioration of acidic paper due to the coupling of temperature and humidity). In addition, sudden instantaneous disturbances (such as impact vibration and gas leakage) are often missed by traditional continuous verification logic due to their short duration.

[0006] The industry has attempted to address this by increasing sensor density or introducing complex algorithm optimizations, but such improvements further exacerbate the data transmission and computing burdens and still fail to resolve the core contradiction between the lack of edge intelligence and inefficient collaborative perception. For example, some solutions use high-frequency sampling to capture instantaneous events, but due to the limited computing power of edge nodes, this leads to a surge in false alarm rates and energy consumption. Therefore, how to achieve early and accurate identification of environmental disturbances on low-power, low-cost edge devices, and distinguish between local anomalies and global interference through distributed collaborative mechanisms, while simultaneously addressing both gradual accumulation and sudden instantaneous risks, has become a key technical challenge in this field. Summary of the Invention

[0007] The present invention provides an intelligent archive warehouse monitoring system based on the Internet of Things, the main purpose of which is to solve the problems of missed detection of weak anomalies, misjudgment of dynamic risks and insufficient collaborative identification of multi-source disturbances caused by information redundancy and superficial cognition in existing centralized monitoring systems.

[0008] To achieve the above objectives, the present invention provides an intelligent archive warehouse monitoring system based on the Internet of Things, the system comprising:

[0009] Multiple edge sensing units are configured in the archive warehouse, and each edge sensing unit is configured as follows:

[0010] Dynamically learn and continuously maintain a personalized steady-state baseline of at least one environmental parameter within its monitoring range, the personalized steady-state baseline being obtained by applying a weighted moving average or exponential smoothing algorithm to historical sampled values of the environmental parameter within a predetermined time window;

[0011] monitoring the current value of the environmental parameter in real time, and determining that a local perturbation event has occurred and entering a local alert state when the current value deviates from the personalized steady-state baseline for a predetermined duration and deviates from a predetermined drift threshold;

[0012] When entering the local alert state, a perturbation wake-up signal containing only its own identification and perturbation wake-up state is sent to its predefined adjacent edge sensing unit without specific environmental parameter data;

[0013] receiving a perturbation wake-up signal from its neighboring edge sensing unit, and when a preset cooperative perturbation determination condition is met, entering a cooperative attention state and sending attention event information to a central processing platform, wherein the preset cooperative perturbation determination condition includes that the edge sensing unit itself is in a local alert state and receives at least one perturbation wake-up signal from its neighboring edge sensing unit, or that the edge sensing unit receives perturbation wake-up signals from at least two different neighboring edge sensing units; and a central processing platform, wherein the central processing platform is configured to:

[0014] Receive and process attention event information from edge sensing units that enter the collaborative attention state to monitor the archive warehouse;

[0015] Based on the event information of concern, relevant edge sensing units are requested to upload more detailed historical data of environmental parameters on demand for in-depth risk analysis;

[0016] Based on statistics of event information and in-depth risk analysis results, the predetermined drift threshold of the edge perception unit or the parameters in the preset cooperative perturbation judgment condition are dynamically adjusted.

[0017] Preferably, each edge perception unit is further configured to: store at least one predefined feature sequence template, each corresponding to a specific type of instantaneous environmental disturbance event; while continuously monitoring the current value of the environmental parameter and performing personalized steady-state baseline learning and comparison, match the current sampling data of the environmental parameter or the short-term feature sequence formed therefrom with at least one feature sequence template in parallel and in real time; when the current sampling data or the short-term feature sequence formed therefrom successfully matches any feature sequence template, it is determined that an instantaneous key disturbance event corresponding to the template has occurred locally, and independent of whether the edge perception unit enters a local alert state or a collaborative attention state, directly send instantaneous alarm information indicating the type, occurrence time and location of the instantaneous key disturbance event to the central processing platform.

[0018] Preferably, the central processing platform is further configured to: after receiving the instantaneous alarm information, send targeted enhanced monitoring instructions to the edge sensing unit where the instantaneous critical disturbance event occurs and its predefined adjacent edge sensing units, and the instructions are configured to enable the instructed edge sensing unit to temporarily increase its monitoring sensitivity or adjust its parameters for determining the local perturbation event in step (b) within a predetermined short time limit.

[0019] Preferably, the power supply module of the edge sensing unit is configured to support long-life battery power supply or energy harvesting technology power supply.

[0020] Preferably, the communication module of the edge sensing unit is configured to communicate with adjacent edge sensing units and / or the central processing platform via Bluetooth Low Energy (BLE) or LoRa wireless technology.

[0021] Preferably, the power consumption of each edge sensing unit in the local alert state is higher than that in the silent monitoring state, and the power consumption in the collaborative attention state is higher than that in the local alert state.

[0022] Preferably, the central processing platform is also configured to: for one or more specific monitoring areas in a preset archive warehouse, or for one or more specific edge sensing units, long-term record and statistical analysis of the frequency of entering the collaborative attention state within a predetermined time period, the combination pattern and quantity characteristics of the perturbation wake-up signals sent by the adjacent edge sensing units required to trigger the collaborative attention state, the response lag time for entering the collaborative attention state, or the duration of the collaborative attention state, which together constitute the resonance mode time series characteristic spectrum data.

[0023] Preferably, the central processing platform is further configured to: construct and continuously update a dynamic health fingerprint for a specific monitoring area or a specific edge sensing unit based on the evolution of the resonance mode time series characteristic spectrum data over time, which can characterize the comprehensive response characteristics and internal stability of the corresponding archival microenvironment to external disturbances, wherein the characteristic vector of the dynamic health fingerprint is It can be expressed in the following form:

[0024] ,

[0025] in, is the average number of triggered neighbors, is the average resonance incubation period, is the average disturbance duration, is the recent frequency; is the preset weight coefficient.

[0026] Preferably, the central processing platform is also configured to: analyze the long-term evolution trend of the dynamic health fingerprint, and when it is detected that the dynamic health fingerprint presents a predetermined negative evolution pattern indicating that the potential risks of the archive microenvironment are accumulating or its inherent stability is continuously decreasing, generate a forward-looking trend risk warning information that is different from the one based on the abnormalities of the immediate environmental parameters.

[0027] Preferably, after generating forward-looking trend risk warning information, the central processing platform is also configured to dynamically adjust the file inspection priority, optimize the allocation of preventive maintenance resources, or guide the scientific partitioning and storage of files of different sensitivity levels.

[0028] Compared with the background technology problems, the beneficial effects of the present invention are:

[0029] 1. Through continuous learning of the personalized baseline of monitoring parameters by edge-sensing units, the system can capture weak, persistent environmental drifts that are difficult to detect using traditional fixed thresholds. When multiple adjacent units interact collaboratively to trigger perturbation wake-up signals based on local baseline drift, the system can autonomously identify potential risk areas with spatial correlation. This dual spatiotemporal verification mechanism not only avoids the misjudgment of isolated disturbances, but also achieves precise positioning of progressive environmental anomalies with low communication overhead through the logical superposition of minimalist signals between units, providing a critical time window for risk intervention.

[0030] 2. While continuously monitoring baseline drift, the edge units match predefined instantaneous disturbance feature templates in parallel to build independent alarm channels for sudden, highly dynamic risks (such as impact vibration and gas leakage). When the central platform receives an instantaneous alarm, it immediately triggers the targeted enhanced monitoring mechanism and dynamically adjusts the detection sensitivity and judgment parameters of the relevant areas. This nested design of fast and slow monitoring logic enables the system to capture the trend characteristics of slowly changing risks and respond immediately and conduct secondary verification of sudden disturbances, forming a closed loop of multi-dimensional risk perception.

[0031] 3. Based on long-term recorded collaborative attention event feature data (such as the number of triggered neighbors and response lag time), the system constructs dynamic health fingerprints for different areas that reflect their environmental stability. By analyzing the temporal evolution of perturbation resonance modes, this fingerprint reveals the attenuation of the archival carrier's tolerance to external disturbances or the trend of internal structural degradation. Compared with direct monitoring of environmental parameters, this reverse inference mechanism based on system behavior patterns can detect the accumulation of microenvironmental vulnerability caused by hidden factors such as material aging and improper storage density earlier, providing forward-looking guidance for preventive maintenance.

[0032] 4. Through edge intelligent processing such as local baseline calculation and collaborative judgment of perturbation signals, the system triggers detailed data inspection and communication link activation in specific areas under collaborative attention or instantaneous alarm states. This event-triggered dynamic power consumption strategy, combined with low-power power supply and communication technology, enables the system to maintain high monitoring accuracy while supporting ultra-large-scale node deployment and continuous operation for more than several years, significantly superior to centralized architectures that rely on continuous data transmission.

[0033] 5. The central platform remotely optimizes the baseline tolerance and collaborative judgment threshold of edge units based on historical event data and manual verification results. This feedback mechanism not only avoids the interference of seasonal environmental changes on fixed parameters, but also enables differentiated configuration based on the archival characteristics of different regions (such as material sensitivity). The system thus forms an autonomous evolutionary capability of monitoring-analysis-optimization, maintaining stable monitoring performance in complex and changing warehouse environments, and avoiding the sensitivity drift problem caused by parameter fixation in traditional systems. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Figure 1 This is a timing diagram of the interactive process of the present invention for identifying and triggering alarms and enhancing monitoring based on instantaneous disturbance feature templates;

[0035] Figure 2 This is a schematic diagram of the trend warning generation process based on feature extraction and dynamic health fingerprint analysis of the present invention;

[0036] Figure 3 Schematic diagram of the edge perception node perturbation event triggering and collaborative attention process of the present invention.

[0037] The purpose, features and advantages of the present invention will be further described with reference to the accompanying drawings and in conjunction with the embodiments. DETAILED DESCRIPTION

[0038] It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0039] The present application provides an Internet of Things-based intelligent archive warehouse monitoring system, which includes:

[0040] Multiple edge sensing units are configured in the archive warehouse, and each edge sensing unit is configured as follows:

[0041] Dynamically learn and continuously maintain a personalized steady-state baseline of at least one environmental parameter within its monitoring range, the personalized steady-state baseline being obtained by applying a weighted moving average or exponential smoothing algorithm to historical sampled values of the environmental parameter within a predetermined time window;

[0042] monitoring the current value of the environmental parameter in real time, and determining that a local perturbation event has occurred and entering a local alert state when the current value deviates from the personalized steady-state baseline for a predetermined duration and deviates from a predetermined drift threshold;

[0043] When entering the local alert state, a perturbation wake-up signal containing only its own identification and perturbation wake-up state is sent to its predefined adjacent edge sensing unit without specific environmental parameter data;

[0044] receiving a perturbation wake-up signal from its neighboring edge sensing unit, and when a preset cooperative perturbation determination condition is met, entering a cooperative attention state and sending attention event information to a central processing platform, wherein the preset cooperative perturbation determination condition includes that the edge sensing unit itself is in a local alert state and receives at least one perturbation wake-up signal from its neighboring edge sensing unit, or that the edge sensing unit receives perturbation wake-up signals from at least two different neighboring edge sensing units; and a central processing platform, wherein the central processing platform is configured to:

[0045] Receive and process attention event information from edge sensing units that enter the collaborative attention state to monitor the archive warehouse;

[0046] Based on the event information of concern, relevant edge sensing units are requested to upload more detailed historical data of environmental parameters on demand for in-depth risk analysis;

[0047] Based on statistics of event information and in-depth risk analysis results, the predetermined drift threshold of the edge perception unit or the parameters in the preset cooperative perturbation judgment condition are dynamically adjusted.

[0048] Preferably, each edge perception unit is further configured to: store at least one predefined feature sequence template, each corresponding to a specific type of instantaneous environmental disturbance event; while continuously monitoring the current value of the environmental parameter and performing personalized steady-state baseline learning and comparison, match the current sampling data of the environmental parameter or the short-term feature sequence formed therefrom with at least one feature sequence template in parallel and in real time; when the current sampling data or the short-term feature sequence formed therefrom successfully matches any feature sequence template, it is determined that an instantaneous key disturbance event corresponding to the template has occurred locally, and independent of whether the edge perception unit enters a local alert state or a collaborative attention state, directly send instantaneous alarm information indicating the type, occurrence time and location of the instantaneous key disturbance event to the central processing platform.

[0049] Preferably, the central processing platform is further configured to: after receiving the instantaneous alarm information, send targeted enhanced monitoring instructions to the edge sensing unit where the instantaneous critical disturbance event occurs and its predefined adjacent edge sensing units, and the instructions are configured to enable the instructed edge sensing unit to temporarily increase its monitoring sensitivity or adjust its parameters for determining the local perturbation event in step (b) within a predetermined short time limit.

[0050] Preferably, the power supply module of the edge sensing unit is configured to support long-life battery power supply or energy harvesting technology power supply.

[0051] Preferably, the communication module of the edge sensing unit is configured to communicate with adjacent edge sensing units and / or the central processing platform via Bluetooth Low Energy (BLE) or LoRa wireless technology.

[0052] Preferably, the power consumption of each edge sensing unit in the local alert state is higher than that in the silent monitoring state, and the power consumption in the collaborative attention state is higher than that in the local alert state.

[0053] Preferably, the central processing platform is also configured to: for one or more specific monitoring areas in a preset archive warehouse, or for one or more specific edge sensing units, long-term record and statistical analysis of the frequency of entering the collaborative attention state within a predetermined time period, the combination pattern and quantity characteristics of the perturbation wake-up signals sent by the adjacent edge sensing units required to trigger the collaborative attention state, the response lag time for entering the collaborative attention state, or the duration of the collaborative attention state, which together constitute the resonance mode time series characteristic spectrum data.

[0054] Preferably, the central processing platform is further configured to: construct and continuously update a dynamic health fingerprint for a specific monitoring area or a specific edge sensing unit based on the evolution of the resonance mode time series characteristic spectrum data over time, which can characterize the comprehensive response characteristics and internal stability of the corresponding archival microenvironment to external disturbances, wherein the characteristic vector of the dynamic health fingerprint is It can be expressed in the following form:

[0055] ,

[0056] in, is the average number of triggered neighbors, is the average resonance incubation period, is the average disturbance duration, is the recent frequency; is the preset weight coefficient.

[0057] Preferably, the central processing platform is also configured to: analyze the long-term evolution trend of the dynamic health fingerprint, and when it is detected that the dynamic health fingerprint presents a predetermined negative evolution pattern indicating that the potential risks of the archive microenvironment are accumulating or its inherent stability is continuously decreasing, generate a forward-looking trend risk warning information that is different from the one based on the abnormalities of the immediate environmental parameters.

[0058] Preferably, after generating forward-looking trend risk warning information, the central processing platform is also configured to dynamically adjust the file inspection priority, optimize the allocation of preventive maintenance resources, or guide the scientific partitioning and storage of files of different sensitivity levels.

[0059] At the same time, to further improve the system's response accuracy and judgment consistency in actual application scenarios, the central processing platform introduces a multi-cycle difference and normalized offset analysis mechanism in combination with historical vector data when processing the evolution trend of dynamic health fingerprints. Specifically, if any two core indicators (such as the average disturbance duration and the collaborative incubation period) in three consecutive statistical periods have a negative deviation of more than 10% compared with the historical average, it is considered a significant evolution trend. The system immediately generates a forward-looking warning prompt to assist manual intervention judgment; and in the enhanced monitoring stage, targeted enhanced monitoring instructions, for example, not only include an increase in sampling frequency (for example, from once every 60 seconds to once every 20 seconds), but also synchronously adjust the judgment parameters. For example, the perturbation event trigger threshold is temporarily tightened to ±5%, the baseline update window is shortened to 60 minutes, and high-sensitivity template matching logic is enabled. The specific settings of such parameters are dynamically optimized based on historical response data of similar events to ensure that the execution of instructions is targeted and effective. These are all extended implementation methods known to ordinary technicians in this field.

[0060] Example 1: The smart archive warehouse monitoring system based on the Internet of Things provided in this embodiment, the edge sensing unit can adopt a modular embedded design, and internally integrate a multi-parameter composite sensor array with the ability to detect temperature and humidity, gas composition (including carbon dioxide and volatile organic compounds), vibration and light. The data acquisition module wakes up the sensor array at a preset time interval through high-stability sampling control logic. In the silent monitoring state, the system acquires environmental parameters at a sampling frequency of once per minute by default. The frequency can be dynamically adjusted according to the alert state or enhanced monitoring needs. In the default state, the system maintains ultra-low power consumption operation, and the timed wake-up mechanism is maintained only by a low-power microprocessor. At the same time, in order to achieve early perception of subtle environmental changes, the edge The perception unit embeds a lightweight time series modeling mechanism locally to generate and continuously update personalized steady-state baselines for independent environmental parameters. This baseline does not rely on the central platform for unified distribution, but is a dynamic mean curve independently constructed by the edge unit based on the self-collected historical data window. Each environmental parameter is processed by weighted sliding average or exponential smoothing. The smoothing coefficient is set between 85% and 95%, for example. The specific value is automatically determined by the local algorithm during the system initialization phase based on the fluctuation characteristics of the environmental parameter itself. During the baseline maintenance process, the system adopts a non-uniform weighting strategy, giving higher weights to recent data to improve the response to short-term changes and avoid the interference of extreme values on the overall trend judgment.

[0061] When the system is running, the edge unit continuously compares the deviation between the current perception value and its baseline. If a parameter deviates in the same direction for three consecutive sampling periods, and the maximum deviation value exceeds the preset tolerance range of the baseline (the range is set to plus or minus seven percent, determined based on the historical fluctuation range of a typical archive warehouse environment), the preliminary judgment logic of the perturbation event is triggered, and the unit immediately enters the local alert state. In this state, the edge unit does not need to activate the complete data upload process, but only sends a minimalist perturbation wake-up signal to its three predefined neighboring edge units. The signal contains the unit's unique identifier, event trigger timestamp and current status flag, and does not contain specific environmental parameter data. , in order to complete the initial wake-up action of collaborative identification of local events with extremely low power consumption. After receiving such perturbation wake-up signals, adjacent edge units make judgments based on the local current state and preset collaborative judgment rules. If the current unit itself is also in the local alert state and receives at least one wake-up signal from a neighboring unit in this state, it enters the collaborative attention state and determines that there is a possible regional risk. Another trigger mechanism is: even if the unit is not yet in the alert state, if it receives wake-up signals from two different neighboring units continuously within ten minutes, it will directly enter the collaborative attention state and send an attention event report containing the event time, location and related edge unit information to the central processing platform.

[0062] The system has a built-in independent transient disturbance recognition mechanism to address high-dynamic risk events such as vibration shocks, transient gas leaks, or sudden changes in light intensity. The edge sensing unit pre-stores characteristic sequence templates for various typical transient disturbance events, including features such as sudden frequency changes in vibration signals and abnormal increases in gas concentration slopes. These templates can be extracted from historical data, for example. In actual operation, the edge unit performs real-time sliding processing on the current sensing data, generating short-term sequences, which are then compared against pre-set templates. Feature matching is typically performed using a sliding window of four to six sampling periods. When the similarity between the current sequence and any template exceeds a preset matching threshold (typically 80%), it is determined to have occurred as a transient critical disturbance event of the corresponding type. An alarm message is immediately reported to the central platform, including the event type, trigger time, and geographic location code. Upon receiving this transient alarm message, the central platform immediately issues a targeted enhanced monitoring command to the unit where the event occurred and its neighboring units. This command instructs the target unit to increase the sampling frequency (to once every 20 seconds) and appropriately relax the threshold for determining micro-disturbance events over the next five to ten minutes to capture possible subsequent disturbances or residual interference. The system also establishes a dynamic health assessment mechanism for the archive warehouse environment through archiving and analysis of long-term monitoring data. The central platform automatically compiles weekly statistics on the resonance characteristics of coordinated attention events within each monitoring area or unit. These include parameters such as the average number of triggering neighboring units involved in each event, the coordinated incubation period (i.e., the time from the onset of the first alert state to the formation of coordinated attention), the average duration of the disturbance, and the frequency of events per unit time. These indicators are normalized and organized into structured vectors, which serve as important reference information for reflecting the stability of the microenvironment in that area or unit. For example, if a region experiences ten coordinated attention events within a month, with an average number of triggering neighboring units of 2.6, a coordinated incubation period of seven minutes, an average duration of five minutes, and a frequency of 2.5 events per unit time per week, the region's current environmental stability state vector is: [2.6, 7, 5, 2.5]. The platform performs trend modeling and change analysis on these vectors. If a significant negative evolutionary trend is identified, it will issue a trend warning, prompting operations and maintenance personnel to promptly conduct unplanned inspections of the area, adjust archive density, or optimize the operation strategies of local air conditioning and dehumidification equipment.

[0063] In terms of key parameter setting, this embodiment analyzes the historical monitoring data of four typical document archive warehouses to derive the key judgment threshold of system micro-disturbance events. The initial judgment threshold of temperature and humidity disturbances is recommended to be set within 7% of the baseline. The effective identification template of vibration disturbances is mainly concentrated on the instantaneous peak change characteristics in the range of 10 to 20 Hz. The abnormality criterion for gas disturbances is a concentration increase rate exceeding 8 units per minute. The above parameter settings have been doubly verified by laboratory simulation platforms and field deployment feedback data, and have sufficient engineering basis and technical feasibility. In terms of system energy consumption, the typical power consumption of the edge sensing unit in the silent state is 10 to 30 microwatts; the power consumption rises to less than 200 microwatts in the local alert state; and the peak power consumption does not exceed 500 microwatts in the enhanced monitoring state. All units are powered by energy collection components such as photovoltaic panels or thermoelectric power generation devices, which are all extended implementation methods known to ordinary technicians in this field.

[0064] Example 2: In the actual application scenario of the archive warehouse microenvironment monitoring system, this example constructs a typical document archive warehouse environment, simulates common microenvironment fluctuations and typical disturbance events, and evaluates the system's response capability, judgment accuracy, and false alarm control level under actual deployment conditions.

[0065] This example selects a document warehouse on the second floor of an archive as the test site. The warehouse has internal dimensions of 28 meters by 12 meters and a floor height of 3.2 meters. It is equipped with a centralized ventilation system and zoned air-conditioning control devices. The warehouse is used to store paper originals, manuscripts, and some precious documentary materials for a long time. The environmental control index requirements are: the temperature is maintained in the range of 18°C to 25°C, and the relative humidity is maintained between 40% and 60%. A total of 12 edge sensing units are deployed in the warehouse, evenly spaced at intervals of 3 meters. Four of the units are set as intervention control nodes to simulate human disturbance situations. Each edge unit integrates a multi-parameter composite sensor array, including temperature and humidity, vibration, and volatile organic compound (TVOC) detection modules. The data collection frequency is set to once every 60 seconds by default. It has local processing capabilities and can implement functions such as baseline learning, drift determination, proximity communication, and template matching. The parameter settings are shown in Table 1:

[0066] Table 1: Parameter setting table

[0067]

[0068] During the first 24 hours of the test, constant environmental conditions (temperature 22.5°C, relative humidity 48%) were maintained in the warehouse to test the system's baseline modeling capabilities in a stable environment. The experimental phenomenon showed that all edge sensing units completed preliminary baseline modeling of their respective environmental parameters within 6 hours after system initialization. The modeling curve exhibited monotonic convergence characteristics, and the deviation of the collected values from the baseline was then stably controlled within the range of ±1.2%. The test concluded that the dynamic modeling algorithm based on exponential smoothing has good convergence and anti-interference capabilities, and is suitable for the needs of dynamic baseline construction in actual engineering environments.

[0069] Starting at the 36th hour, the local environment of node #04 was manually perturbed by temperature and humidity, raising the temperature by 1.8°C and the relative humidity by 5%. The perturbation lasted for 90 minutes. The system responded by detecting that the deviation from the baseline at node #04 exceeded 7% in three consecutive sampling periods, and it entered a local alert state. It then sent a perturbation wake-up signal to neighboring nodes #03 and #05. Node #03 entered a collaborative attention state at the 12th minute, and node #05 did not trigger a response due to a stable environment. The central platform received the event report information at the 15th minute. The analysis concluded that the local baseline deviation trigger mechanism effectively identified slowly changing temperature and humidity events, the response delay was within the expected range, and the collaborative perception mechanism had the ability to identify regional anomalies.

[0070] At the 48th hour, nodes #07, #08, and #09 simultaneously applied disturbances (temperature increased by 1.6°C and TVOC concentration increased by 80 ppm), with a disturbance duration set to 25 minutes. The system responded by the three nodes successively entering a local alert state between the 4th and 6th minutes, and after being awakened by each other's micro-perturbations, all entered a coordinated attention state. The central platform received the event report at the 10th minute and initiated enhanced monitoring instructions. In the subsequent 10 minutes, the sampling frequency of the relevant nodes was increased to once every 20 seconds, successfully capturing the rising trend of TVOC and micro-vibration signals. The analysis concluded that the system has the ability to identify regional resonance risks under the conditions of multi-point synchronous disturbances, the event response mechanism is efficient, and the issuance of instructions is well synchronized with the on-site status.

[0071] At the 60th hour, a ground vibration event was artificially simulated, with an input disturbance frequency of 12Hz, a peak acceleration of 0.6g, and a duration of 6 seconds. The system responded by instantly identifying the vibration event and generating an instantaneous alarm at node #10. The matching template similarity reached 0.88, exceeding the preset threshold. The central platform completed the command issuance within 5 seconds. The system did not detect any subsequent anomalies, and the node status automatically recovered. The analysis concluded that the preset template mechanism accurately identified instantaneous vibration events, the system had a fast response time, and good false alarm control.

[0072] Table 2: Summary of test data

[0073]

[0074] The experimental data clearly verified the system's baseline modeling capabilities for dynamic changes in environmental parameters in actual deployment scenarios, the triggering and identification effect of local anomalies, and the collaborative response logic under multi-point disturbances. The collaborative perception mechanism constructed by the system has the ability to accurately identify regional micro-disturbances, and cooperates with the event-driven enhanced monitoring process to achieve closed-loop processing. In addition, it achieves rapid response to high-dynamic disturbance events through template matching, effectively compensating for the delay problem of the static judgment mechanism in emergency situations. The overall false alarm rate of the system is controlled within 2%, and the instantaneous response delay is significantly lower than that of traditional centralized monitoring systems, with excellent real-time performance and engineering feasibility.

[0075] Example 3: This example combines Figures 1 to 3 , this paper describes the implementation of an intelligent archive warehouse monitoring system based on the Internet of Things. Figure 1 As shown in the figure, when edge perception unit A continuously monitors the current value of the environmental parameter and matches the short-term feature sequence formed by the current sampling data with at least one feature sequence template in parallel and in real time, it can determine that an instantaneous key disturbance event corresponding to the template has occurred locally under the condition of successfully matching the instantaneous disturbance feature template, and immediately send the instantaneous alarm information (type, occurrence time and location) to the central processing platform. After receiving the instantaneous alarm information, the central processing platform will send targeted enhanced monitoring instructions to the adjacent edge perception unit B and the adjacent edge perception unit C respectively to trigger them to temporarily increase the monitoring sensitivity. At the same time, the central processing platform will also send targeted enhanced monitoring instructions to edge perception unit A. Therefore, edge perception unit A temporarily increases the monitoring sensitivity and adjusts the parameters for determining local perturbation events to improve the recognition ability of subsequent disturbances. If the feature sequence template is not matched, edge perception unit A will continue baseline learning and comparison.

[0076] like Figure 2 As shown in the figure, through the feature extraction module, key parameters including the number of triggered neighbors, resonance incubation period, duration and frequency of occurrence are extracted from the collaborative attention state recorded long-term by the edge sensing unit. These parameters are sent to the dynamic health fingerprint calculation module as multi-dimensional input to construct a dynamic health fingerprint that can reflect the response characteristics of a specific monitoring area or edge sensing unit to external disturbances and the evolution of its inherent stability in a perturbation resonance event. Subsequently, the dynamic health fingerprint calculation result is input into the trend analysis module. By identifying the changing trend of the fingerprint data in the time dimension, the early identification of the potential risk evolution path is achieved. If the trend analysis result shows that the system stability has significantly decreased or the potential risk is accumulating, the warning signal generation link is entered, and the system outputs the corresponding trend risk warning signal.

[0077] like Figure 3As shown in the figure, nodes #01, #02, #05, and #07 are all in silent monitoring state and no abnormality is detected; node #03 enters the local alert state based on the judgment logic of parameter deviation ≥7% and sends a perturbation wake-up signal to the neighboring nodes; node #06 also enters the local alert state after independently judging the abnormality. After that, because the collaborative judgment conditions are met, node #04 enters the collaborative attention state and reports the attention event to the central processing platform. After receiving the attention event, the central processing platform sends targeted enhanced monitoring instructions to the relevant nodes, thereby guiding the local area to improve the monitoring sensitivity. The propagation path of the perturbation wake-up signal is marked by solid arrows in the figure, the sending direction of the enhanced monitoring instruction is indicated by dotted arrows, and the solid arrows indicate the data flow of the attention event report.

[0078] Example 4: In a typical application scenario, such as when an archives management agency implements intelligent transformation of its first-level security-level special collection archive warehouse, the system deploys multiple composite environmental perception units with edge intelligent processing capabilities, and builds a full-spectrum intelligent monitoring system with collaborative perception, fingerprint modeling and dynamic early warning functions through the central processing platform. During implementation, edge sensing units were installed at key monitoring locations within the archive warehouse, including below the central ventilation outlet, near exterior walls, and in aisles and skylights between high-density storage racks. These locations are subject to various environmental disturbances during daily operation, including fluctuating wind speeds, fluctuating temperature gradients, localized moisture accumulation, and sudden changes in light intensity. Therefore, the sensing units must possess strong adaptability to these scenarios and the ability to self-regulate dynamic parameters. Each edge sensing unit incorporates a lightweight local data modeling logic for continuous monitoring of multiple environmental parameters, including temperature and humidity, gas concentrations, and microvibration signals. It constructs a personalized steady-state baseline based on the most recent 90-minute data window. This modeling process utilizes an exponential smoothing algorithm combined with a non-uniform weighting update strategy, assigning a higher weight to the most recent 30 minutes of data in the time series to enhance the ability to identify the early stages of sudden disturbances. The smoothing factor is set between 85 and 95 percent. The specific value is automatically set by the sensing unit during system initialization based on the historical volatility of each sensor. It can also be adjusted during subsequent operation based on data feedback from the central platform, effectively avoiding modeling errors caused by varying sensor accuracy or age.

[0079] Taking a high temperature and high humidity environment as an example, node A is deployed near the air outlet of an air conditioner. The area where it is located is prone to periodic small fluctuations in humidity due to the condensation effect. During the modeling process, node A identifies the frequency change characteristics of the humidity parameter, and then automatically adjusts the smoothing factor to a lower value. It also increases the judgment threshold of humidity disturbance by 1% on the basis of the original 7%, in order to enhance the tolerance to such background disturbances, thereby effectively suppressing false alarms caused by the start and stop or switching of the air conditioner in the actual measurement environment. In the default state of the system operation, if an edge unit has an offset in the same direction relative to its steady-state baseline and exceeds the tolerance range in three consecutive sampling periods (about three minutes), the unit determines that a local perturbation event has occurred and enters a local alert state. In this state, the node will send a message to its three predefined Neighboring nodes synchronously send perturbation wake-up signals. The above signals adopt a fixed-length low-byte structure and only contain the sender identification, trigger timestamp and current status flag. They are broadcast and transmitted through the low-power Bluetooth communication channel to reduce communication bandwidth occupancy and ensure the efficiency of collaborative response in scenarios with dense node deployment. If one of the neighboring nodes is in a local alert state when receiving the wake-up signal, or a node receives wake-up signals from two different neighboring nodes continuously within ten minutes, the node enters a collaborative attention state and uploads a data packet including location code, collaborative relationship information and event summary content to the central processing platform through an encrypted communication channel. The above collaborative judgment mechanism is designed based on the common risk propagation path model of document archives, and prioritizes the identification of multi-point disturbance patterns with spatial correlation trends.

[0080] In an actual deployment test, the area where Node C is located experienced a local temperature rise due to a sudden failure of the air-conditioning equipment. The node detected that the temperature offset value exceeded the tolerance range within four minutes, and then sent a perturbation wake-up signal to the adjacent nodes B and D. Node B had already entered the local alert state in advance due to detecting a slight change in humidity, so the two nodes quickly formed a collaborative relationship and uploaded event data to the central processing platform. After receiving the event information, the central platform identified through the event recording module that two collaborative events had occurred in the area within an hour, and immediately issued an enhanced monitoring instruction, requiring all relevant nodes in the area to increase the sampling frequency to once every fifteen seconds and temporarily shorten the baseline update time window to sixty minutes. In order to improve the perception of trend disturbance changes, in addition to the above-mentioned detection mechanism based on baseline drift, the edge perception unit also presets several instantaneous disturbance feature templates, including vibration pulse width distribution characteristics, gas concentration slope mutation form, etc., and performs parallel feature matching on the monitoring data during operation. The above templates are derived from the features extracted by professional archival institutions after statistical analysis of spectrum, slope, duration and multi-dimensional behavior based on environmental accident data in the past ten years. The system uses a short-term sequence window of fixed length for matching, and sets the template similarity judgment threshold to 82% by default. The matching algorithm adopts time-weighted composite comparison logic to ensure the recognition ability of approximate but non-standard disturbances.

[0081] In the simulated earthquake interference scenario, node E successfully matched the high-frequency pulse interference template within two seconds after the vibration occurred, immediately generated an instantaneous alarm information and uploaded it to the central processing platform. The report content included the event type, occurrence time and node location code. After receiving the information, the platform immediately issued a targeted enhanced monitoring instruction to node E and its three neighboring nodes. The instruction content included enabling the high-frequency vibration detection subroutine and expanding the detection bandwidth of the gas sensor. The system then recorded a slight upward trend in TVOC concentration within fifteen minutes, confirming the possibility of structural disturbance accompanied by volatilization of substances. The platform immediately triggered the manual review process of the area. The central processing platform included the above events in the dynamic health fingerprint database and constructed a feature vector based on key parameters such as the number of responding neighbors, average triggering lag time, disturbance duration period and event frequency per unit time. The vector was compared with the historical mean. Trend comparison: if any two parameters have a negative deviation of more than ten percent in three consecutive statistical periods, a trend warning will be automatically triggered, prompting operation and maintenance personnel to perform preemptive maintenance on the relevant areas. This warning mechanism enables the system to identify the trend of declining regional stability in advance before obvious parameter abnormalities occur, providing an early warning basis for archival resource protection. At the same time, the system also adopts an information-physical fusion control technology architecture, with real-time status perception, edge intelligent processing and centralized coordination feedback as the core structure, and has dynamic and adaptive operation capabilities. In specific implementation, for example, a low-power embedded computing platform can be combined with a low-energy communication module, supplemented by energy harvesting and power supply methods, so that the system can achieve long-term, continuous and efficient intelligent operation without the need for additional wiring or frequent maintenance. These are all extended implementation methods that can be known to ordinary technicians in this field.

[0082] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0083] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not limiting. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. An intelligent archive warehouse monitoring system based on the Internet of Things, characterized in that: The system comprises: Multiple edge sensing units are configured in an archive warehouse, each edge sensing unit being configured to: dynamically learn and continuously maintain a personalized steady-state baseline of at least one environmental parameter within its monitoring range; monitor the current value of the environmental parameter in real time, and when the current value drifts from the personalized steady-state baseline for a predetermined duration and deviates from a predetermined drift threshold, determine that a local perturbation event has occurred and enter a local alert state; upon entering the local alert state, transmit a perturbation wake-up signal containing only its own identification and perturbation wake-up state and no specific environmental parameter data to its predefined neighboring edge sensing units; receive perturbation wake-up signals from its neighboring edge sensing units, and when a preset collaborative perturbation determination condition is met, enter a collaborative attention state and transmit attention event information to a central processing platform; Furthermore, the system further includes a central processing platform configured to: receive and process attention event information from edge sensing units that have entered a collaborative attention state; request related edge sensing units to upload more detailed historical environmental parameter data based on the attention event information; and dynamically adjust predetermined drift thresholds of edge sensing units or parameters in preset collaborative perturbation determination conditions based on statistics of the attention event information and in-depth risk analysis results; The central processing platform is further configured to: for one or more specific monitoring areas in a preset archive warehouse, or for one or more specific edge sensing units, long-term record and statistically analyze the frequency of entering a collaborative attention state within a predetermined time period, the combination pattern and quantity characteristics of the perturbation wake-up signals sent by the adjacent edge sensing units required to trigger the collaborative attention state, the response lag time for entering the collaborative attention state, or the duration of the collaborative attention state, thereby forming a resonance pattern time series characteristic spectrum data; The central processing platform is also configured to: based on the evolution of the resonance mode time series characteristic spectrum data of a specific monitoring area or a specific edge sensing unit over time, construct and continuously update a dynamic health fingerprint for the area or unit that can characterize the comprehensive response characteristics and internal stability of the corresponding archival microenvironment to external disturbances, where the characteristic vector of the dynamic health fingerprint is It can be expressed in the following form: ， in, is the average number of triggered neighbors, is the average resonance incubation period, is the average disturbance duration, is the recent frequency; is the preset weight coefficient.

2. The intelligent archive warehouse monitoring system based on the Internet of Things according to claim 1 is characterized in that: Each edge perception unit is further configured to: store at least one predefined feature sequence template, each corresponding to a specific type of instantaneous environmental disturbance event; while continuously monitoring the current value of the environmental parameter and conducting personalized steady-state baseline learning and comparison, match the current sampling data of the environmental parameter or the short-term feature sequence formed by it with at least one feature sequence template in parallel and in real time; when the current sampling data or the short-term feature sequence formed by it successfully matches any feature sequence template, it is determined that an instantaneous key disturbance event corresponding to the template has occurred locally, and independent of whether the edge perception unit enters a local alert state or a collaborative attention state, it directly sends an instantaneous alarm information indicating the type, occurrence time and location of the instantaneous key disturbance event to the central processing platform.

3. The intelligent archive warehouse monitoring system based on the Internet of Things according to claim 2 is characterized in that: The central processing platform is also configured to: after receiving the instantaneous alarm information, send targeted enhanced monitoring instructions to the edge sensing unit where the instantaneous critical disturbance event occurs and its predefined adjacent edge sensing units, and the instructions are configured to temporarily increase the monitoring sensitivity of the instructed edge sensing unit within a predetermined short period of time.

4. The intelligent archive warehouse monitoring system based on the Internet of Things according to claim 1 is characterized in that: The communication module of the edge sensing unit is configured to communicate with neighboring edge sensing units and / or the central processing platform via low-power Bluetooth or LoRa wireless technology.

5. The intelligent archive warehouse monitoring system based on the Internet of Things according to claim 1 is characterized in that: The central processing platform is also configured to analyze the long-term evolution trend of the dynamic health fingerprint and generate trend risk warning information that is different from the abnormalities based on the immediate environmental parameters when it detects that the dynamic health fingerprint shows a predetermined negative evolution pattern indicating that the potential risks of the archive microenvironment are accumulating or its inherent stability is continuously decreasing.

6. The intelligent archive warehouse monitoring system based on the Internet of Things according to claim 5 is characterized in that: After generating trend risk warning information, the central processing platform is also configured to dynamically adjust the file inspection priority, optimize the allocation of preventive maintenance resources, or guide the scientific partitioning and storage of files of different sensitivity levels.

Images