Iot-based communication machine room comprehensive monitoring method and device and storage medium

Abstract

The present application relates to the technical field of communication machine room monitoring, and particularly relates to a communication machine room comprehensive monitoring method and device based on the Internet of Things and a storage medium. The method comprises: collecting machine room data through a sensing node of the communication machine room; constructing a three-dimensional correlation model based on the machine room data to obtain an Internet of Things event topology graph; establishing a logic rule engine based on the Internet of Things event topology graph to perform event attribution; and performing hierarchical early warning based on the result of event attribution. The present application realizes comprehensive and deep monitoring of the running state of the communication machine room by constructing a four-layer Internet of Things architecture comprising an environment sensing layer, a device state layer, a physical security layer and a comprehensive analysis platform.

Description

Technical Field

[0001] This invention relates to the field of communication equipment room monitoring technology, and in particular to a method, device and storage medium for comprehensive monitoring of communication equipment rooms based on the Internet of Things. Background Technology

[0002] As communication networks continue to expand, communication data centers, as core nodes for data transmission and equipment operation, are directly affected by their internal environment and equipment status, impacting network stability. Data centers contain numerous servers, air conditioners, power distribution equipment, and other devices, operating under high loads and generating significant heat. They are also sensitive to parameters such as temperature, humidity, cleanliness, vibration, and gas concentration. Traditional methods relying on manual inspections or independent alarm systems are insufficient to meet real-time and collaborative requirements. The development of IoT technology provides a new technological path for comprehensive data center monitoring.

[0003] Existing communication equipment room monitoring technologies have the following main shortcomings: First, each sensor operates independently, lacking the ability to correlate and analyze data, making it difficult to identify complex anomalies or attribute events. Second, there is a lack of unified spatial and logical modeling methods, making it impossible to effectively integrate physical location, equipment affiliation, and time-series data, resulting in difficulties in anomaly localization. Third, the early warning mechanism is simplistic, relying solely on fixed thresholds to trigger alarms, ignoring the degree of deviation and deterioration trends, and easily leading to missed or false alarms. Fourth, it cannot distinguish between routine maintenance actions and genuine equipment failures or security incidents, requiring maintenance personnel to spend a significant amount of time troubleshooting invalid alarms, resulting in low response efficiency. Summary of the Invention

[0004] The purpose of this invention is to provide a method, device, and storage medium for comprehensive monitoring of communication equipment rooms based on the Internet of Things, so as to solve at least one of the problems existing in the prior art.

[0005] To achieve the above objectives, the present invention adopts the following technical solution:

[0006] A comprehensive monitoring method for communication equipment rooms based on the Internet of Things (IoT), comprising:

[0007] Data is collected from the communication equipment room through sensing nodes;

[0008] A three-dimensional correlation model is constructed based on data from the computer room to obtain an IoT event topology map;

[0009] A logical rule engine is built based on the IoT event topology map to perform event attribution;

[0010] Graded early warning is based on the results of event attribution.

[0011] Furthermore, an IoT event topology map is established. For the spatial index layer, the physical space of the communication room is divided into a cubic grid with a side length of 0.5 meters, and each grid is assigned a spatial index ID. The physical coordinates of all sensing nodes are mapped to the corresponding grid to bind the sensing nodes with the spatial index ID.

[0012] For the logical association layer, association tuples are generated based on the logical affiliation of the sensing nodes to classify sensing nodes belonging to the same device.

[0013] For time slices and event streams, a circular buffer with a preset buffer duration is maintained. Sensing nodes belonging to the same monitoring object store the latest sensor readings and state change events in the circular buffer in chronological order. When the sensor value of any sensing node of a monitoring object exceeds its preset baseline or a state jump occurs, it is marked as a potential event and a time tracking window is set to continuously monitor the data change trend of all sensing nodes belonging to the same monitoring object within the time tracking window.

[0014] Furthermore, collaborative analysis is performed on the associated tuples in the IoT event topology graph, and judgment is made through a preset logical rule engine. The preset logical rule engine matches the data change patterns of each monitored object within the event tracking window to output event attribution.

[0015] Furthermore, after completing the attribution analysis, the deviation and deterioration rate of different events are analyzed. The deviation is calculated to be between 0 and 2. A deviation greater than 2 is counted as 2. A deviation less than 0.2 is marked as slight deviation, a deviation between [0.2, 0.6] is marked as moderate deviation, and a deviation greater than 0.6 is marked as severe deviation.

[0016] The method for analyzing the deterioration rate is as follows: starting from the event origin, take the sensor readings of the five most recent sampling points, and use linear least squares fitting to obtain the slope of change; for vibration anomalies, use the impact frequency to replace its deterioration rate.

[0017] Furthermore, a tiered early warning system is implemented based on the event attribution results, deviation, and rate of deterioration, including Level 1, Level 2, and Level 3 early warnings.

[0018] Furthermore, a Level 1 warning is determined when one of the following conditions is met:

[0019] Condition 1: The event attribution is classified as high-risk and critical, and is marked as severely deviated;

[0020] Condition 2: The rate of deterioration reaches its corresponding first rate threshold;

[0021] Condition 3: The vibration and impact frequency is greater than or equal to 5 times / 10 seconds and the peak value of a single impact is greater than or equal to 0.5g;

[0022] Condition 4: TVOC and ozone simultaneously trigger a Level 1 or Level 2 warning.

[0023] Furthermore, for a Level II warning, a warning is defined as one that meets one of the following conditions but does not reach the Level I warning level:

[0024] Condition 1: The event attribution is of the critical type and is marked as moderate deviation, and the rate of deterioration has not reached its corresponding first rate threshold;

[0025] Condition 2: The rate of deterioration reaches its corresponding second rate threshold;

[0026] Condition 3: Deviation is greater than or equal to 0.4 but does not meet the conditions for a Level 1 warning.

[0027] Furthermore, for a Level 3 warning, a warning is defined as one that meets one of the following conditions but does not reach the level of a Level 1 or Level 2 warning:

[0028] Condition 1: Marked as slightly deviated and the rate of deterioration has not reached its corresponding second rate threshold;

[0029] Condition 2: The time attribution is due to temperature rise caused by normal load fluctuations or routine maintenance actions;

[0030] Condition 3: The cumulative area of ​​gas concentration exceeding the limit is less than 10 ppm·s, and the peak growth rate is less than 0.3.

[0031] On the other hand, the present invention also provides an integrated monitoring device for communication equipment rooms based on the Internet of Things, comprising:

[0032] The data acquisition unit is used to collect data from the communication equipment room through the sensing nodes in the equipment room;

[0033] The correlation analysis unit is used to build a three-dimensional correlation model based on data center data to obtain an IoT event topology map;

[0034] The event attribution unit is used to build a logical rule engine based on the IoT event topology map for event attribution.

[0035] The tiered early warning unit is used to issue tiered early warnings based on the results of event attribution.

[0036] On the other hand, the present invention also provides a storage medium storing instructions that, when run on a computer, cause the computer to execute the IoT-based integrated monitoring method for communication equipment rooms as described above.

[0037] The beneficial effects of this invention are as follows: By collecting data from the data center through sensing nodes, a three-dimensional correlation model is constructed to form an IoT event topology map. Combined with a logical rule engine, event attribution and hierarchical early warning are performed, improving the collaborative analysis capabilities of data center monitoring. It can organically integrate spatial location, device logical attribution, and time series data to achieve accurate location and cause tracing of abnormal events. Through quantitative analysis of deviation and deterioration rate, the severity and trend of anomalies are effectively distinguished, avoiding false alarms or missed alarms caused by single threshold alarms. The hierarchical early warning mechanism can output different levels of alarms according to the severity of the event, enabling maintenance personnel to prioritize high-risk events and improving maintenance response efficiency and resource utilization. At the same time, the logical rule engine can identify the differences between routine maintenance actions and equipment failures and security events, reducing interference from invalid alarms, lowering maintenance costs, and enhancing the overall operational reliability and security of the communication data center. Attached Figure Description

[0038] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0039] Figure 1 This is a flowchart of the IoT-based integrated monitoring method for communication equipment rooms in this embodiment.

[0040] Figure 2 This is a schematic diagram of the structure of the IoT-based integrated monitoring device for communication equipment rooms in this embodiment. Detailed Implementation

[0041] The following detailed description, in conjunction with the accompanying drawings and specific embodiments, provides a more comprehensive overview of the IoT-based integrated monitoring method, apparatus, and storage medium for communication equipment rooms disclosed in this invention. It should be noted that the technical features or combinations of technical features described in the following embodiments should not be considered isolated; they can be combined to achieve better technical results. In the accompanying drawings of the following embodiments, the same reference numerals in each drawing represent the same features or components, which can be applied to different embodiments. Therefore, once an item is defined in one drawing, it does not need to be further discussed in subsequent drawings.

[0042] It should be noted that the structures, proportions, sizes, etc., illustrated in the accompanying drawings are merely for illustrative purposes and to aid those skilled in the art in understanding and reading the invention. They are not intended to limit the conditions under which the invention can be implemented. Any modifications to the structure, changes in proportions, or adjustments to size, provided they do not affect the effectiveness or purpose of the invention, should fall within the scope of the technical content disclosed in the invention. The scope of the preferred embodiments of the present invention includes other implementations, wherein functions may be performed not in the order stated or discussed, including substantially simultaneously or in reverse order, depending on the functions involved. This should be understood by those skilled in the art to which the embodiments of the present invention pertain.

[0043] Techniques, methods, and apparatus known to those skilled in the art may not be discussed in detail, but where appropriate, such techniques, methods, and apparatus should be considered part of the specification. In all examples shown and discussed herein, any specific values ​​should be interpreted as merely exemplary and not as limitations. Therefore, other examples of exemplary embodiments may have different values.

[0044] In the description of the embodiments of this application, " / " means "or", and "and / or" is used to describe the relationship between related objects, indicating that there can be three relationships. For example, "A and / or B" means: A and B exist alone, B exists alone, and A and B exist simultaneously. In the description of the embodiments of this application, "multiple" refers to two or more embodiments.

[0045] Please see Figure 1 As shown, this embodiment illustrates a comprehensive monitoring method for communication equipment rooms based on the Internet of Things, including:

[0046] Please continue reading. Figure 1 As shown, the IoT-based integrated monitoring method for communication equipment rooms also includes:

[0047] Step S1: Collect data from the communication room through the sensing nodes in the communication room.

[0048] Specifically, in step S1 of this embodiment, sensing nodes are deployed in the communication equipment room of the science and technology park. The sensing nodes include environmental monitoring nodes and physical sensing nodes. For example, for a communication equipment room with an area of ​​150 square meters to 300 square meters, 20 to 50 standard server racks, as well as corresponding network switching equipment, air conditioning systems, and power distribution units are deployed inside.

[0049] For environmental monitoring nodes, environmental monitoring nodes are deployed at the air inlet and outlet of each cabinet, at the floor grille every 2 meters in the cold aisle, at the ceiling every 2 meters in the hot aisle, and at the return air vent of each precision air conditioner. The environmental monitoring nodes include the following sensors:

[0050] The digital temperature and humidity sensor has a sampling frequency of once every 10 seconds. It uses a chip that integrates a capacitive or resistive humidity sensing element and a bandgap temperature sensing element. The measurement range is temperature -20℃ to +80℃ with an accuracy of ±0.3℃, and humidity 0 to 100%RH with an accuracy of ±2%RH.

[0051] The micro differential pressure sensor uses a MEMS differential pressure chip with a range of -50Pa to +50Pa and an accuracy of ±0.5Pa. The two ends of the micro differential pressure sensor are connected to the inside of the front and rear doors of the cabinet through silicone tubes, respectively, to monitor the air pressure difference between the front and rear of the cabinet in real time. The normal range of the pressure difference is [3Pa, 12Pa]. If the cabinet is new or the filter is being cleaned, the normal range of the pressure difference is [8Pa, 12Pa]. As the filter becomes clogged or the fan ages, the air pressure difference between the front and rear of the cabinet gradually decreases to below 3Pa.

[0052] The dust particle counter uses the principle of laser scattering to measure the concentration of airborne particles with a diameter ≥0.5μm. The range is 0 to 10^6 particles / m³, and the accuracy is ±10% of the reading. The sampling frequency is once every 5 minutes, or an encrypted sampling is initiated immediately when the vibration / door magnetic sensor in the computer room is triggered. Under normal computer room cleanliness, the particle concentration should be less than 10^4 particles / m³.

[0053] For physical sensing nodes, the following sensors are included:

[0054] The miniature vibration and tilt sensor employs a triaxial MEMS accelerometer with a configurable range, such as ±2g to ±16g, a sensitivity of 0.1mg, and a sampling frequency of 100Hz. The sensor is attached to the non-ventilation area on the side of the chassis of each server, network switch, and disk array via a magnetic base or 3M adhesive to record acceleration values ​​along three axes and differentiate different behaviors on the waveform: during normal operation, the root mean square of the vibration value should be less than 0.005g; the peak impact generated by the normal opening and closing of the cabinet door is approximately 0.05g to 0.2g, lasting 0.2 seconds; unauthorized disassembly of the equipment will generate a non-periodic impact of 0.5g to 1.5g, lasting for more than 1 second.

[0055] The system consists of door magnetic switches and infrared array sensors. Wireless door magnetic switches are installed at corresponding positions on the door frames and door panels of the air inlet and outlet doors of each cabinet to detect the door's open / closed status and immediately upload events when the status changes. An 8×8 pixel low-resolution infrared array sensor with a field of view of 60°×60° is installed on the top of each cabinet, covering an area of ​​0.5 meters to 2 meters in front of the cabinet. The infrared array sensor outputs a frame of heat source distribution matrix every 2 seconds to identify the presence, movement trajectory and contour features of personnel. If the heat source temperature is between 30°C and 37°C and occupies multiple consecutive pixels, it is identified as a person.

[0056] A gas sensor array, comprising a combination of gas-sensitive elements deployed at the return air vent of each precision air conditioner, including:

[0057] Ozone sensor, range 0 to 10 ppm, accuracy ±0.02 ppm;

[0058] The TVOC sensor uses the metal-oxide-semiconductor principle, has a measurement range of 0 to 50 ppm, is calibrated with equivalent isobutylene, and has an accuracy of ±5%.

[0059] Hydrogen sulfide sensor, used to detect acid mist generated by battery leakage, with a range of 0 to 20 ppm;

[0060] The sensors in the aforementioned gas sensor array sample every 30 seconds. Under normal conditions, the TVOC concentration in the computer room should be below 0.5 ppm and the ozone concentration should be below 0.05 ppm.

[0061] All sensing nodes communicate with the central analysis platform using low-power wide-area network communication protocols such as LoRaWAN or NB-IoT. Each sensing node has a unique identifier and stores its physical coordinates and logical affiliation. The physical coordinates are rectangular coordinates with the southwest corner of the communication room as the origin and the unit being centimeters. The logical affiliation is the monitoring object to which the sensing node belongs, such as a cabinet or channel. The cabinet refers to the aforementioned standard server cabinet.

[0062] Specifically, in step S1 of this embodiment, by deploying multiple environmental monitoring nodes and physical sensing nodes, covering temperature, humidity, micro-pressure difference, dust particles, vibration, door magnets, infrared heat sources, and various gas concentrations, comprehensive data on the operational status of the communication room can be collected. The combination of high-precision sensor selection and a reasonable sampling frequency ensures the continuity and accuracy of data acquisition. The application of low-power wide-area network communication protocols reduces deployment and maintenance costs. Simultaneously, each sensing node possesses a unique identifier and physical coordinates and logical attribution information, providing a high-quality data foundation for subsequent 3D modeling and event correlation analysis, significantly improving the system robustness and information integrity of the data acquisition layer.

[0063] Step S2: Construct a three-dimensional association model based on the data center data to obtain an IoT event topology map.

[0064] Specifically, in step S2 of this embodiment, data cleaning is performed on the data in the computer room. For single-point data from each sensor, a sliding window midpoint filter is used with a window width of 5 sampling points to remove outliers caused by communication interference. For missing data, if the duration is less than 30 seconds, linear interpolation is used to supplement it; if it exceeds 30 seconds, the sensing node is marked as having a communication anomaly, and a node self-check is triggered. The node self-check is a self-diagnosis enabled for sensing nodes with communication anomalies to ensure data integrity.

[0065] Specifically, in step S2 of this embodiment, the method for establishing an IoT event topology diagram includes:

[0066] For the spatial index layer, the physical space of the communication room is divided into a cubic grid with a side length of 0.5 meters, and each grid is assigned a spatial index ID. The physical coordinates of all sensing nodes are mapped to the corresponding grid to bind the sensing nodes with the spatial index ID.

[0067] For the logical association layer, association tuples are generated based on the logical affiliation of the sensing nodes to classify sensing nodes belonging to the same device. For example, a cabinet tuple is formed by classifying the inlet air temperature sensor, outlet air temperature sensor, differential pressure sensor, vibration sensor of all devices in the cabinet, door magnetic switch, infrared array sensor, and environmental temperature and humidity nodes of the aisle to which the cabinet belongs. A cabinet tuple is formed by classifying the return air temperature and humidity of a precision air conditioner, return air vent gas sensor, and all environmental temperature and humidity nodes in the three cold aisles covered by the air conditioner.

[0068] For time slices and event streams, a circular buffer with a preset buffer duration is maintained. Sensing nodes belonging to the same monitored object store the latest sensor readings and state change events in the circular buffer in chronological order. When the sensor value of any sensing node of a monitored object exceeds its preset baseline or a state jump occurs, it is marked as a potential event and a time tracking window is set to continuously monitor the data change trends of all sensing nodes belonging to the same monitored object within the time tracking window. In this embodiment, the preset buffer duration is set to 30 minutes. This embodiment does not specify a specific limit, but its value range should be within [20 minutes, 60 minutes]. The time tracking window setting should not exceed half of the preset buffer duration and should be greater than or equal to 5 minutes. For example, in this embodiment, the time tracking window can be set to 15 minutes. The state change events include door opening and closing, vibration data exceeding a threshold, etc. For the preset baseline of sensor readings, it is calculated by the sliding average of the same period in the past 24 hours, allowing a fluctuation range of ±15%. By constructing an IoT event topology map, each state change event will obtain a spatial coordinate, a time coordinate, and an event-related data stream, forming a three-dimensional data association foundation.

[0069] Specifically, in step S2 of this embodiment, a three-dimensional association model including a spatial index layer, a logical association layer, and time slices and event streams is constructed to form an IoT event topology map, achieving a unified expression of the physical space of the data center, the ownership relationship of equipment, and time series data. Spatial grid partitioning and the assignment of spatial index IDs enable any sensing node to quickly locate its physical position; logical association tuples bind multiple sensors of the same device, facilitating collaborative analysis; the design of the circular buffer and event tracking window effectively captures data change trends before and after anomalies, providing a complete temporal context for event attribution and greatly enhancing the fusion and analysis capabilities of multi-source heterogeneous data.

[0070] Step S3: Establish a logical rule engine based on the IoT event topology map to perform event attribution.

[0071] Specifically, in step S3 of this embodiment, the associated tuples in the IoT event topology graph are analyzed collaboratively, and a preset logical rule engine is used to make judgments. The preset logical rule engine matches the data change patterns of each monitored object in the event tracking window to output event attribution.

[0072] Specifically, in step S3 of this embodiment, the preset logical rule engine is described in detail, taking a typical abnormal scenario as an example:

[0073] Attribution analysis of abnormally high temperature at the rack exhaust vents:

[0074] Suppose that the temperature To value of the air outlet temperature sensor of cabinet R1 rises from 28℃ to 34℃ within three consecutive sampling points, exceeding the normal operating limit of the cabinet, and the event attribution analysis process is triggered as follows:

[0075] 1. Determine ventilation and heat dissipation capacity based on micro-differential pressure data. Read the average value ΔP of the micro-differential pressure sensor in the same rack group over the past 10 seconds. If ΔP is less than 2.5Pa and the rate of decrease exceeds 0.5Pa / minute, it is determined to be a ventilation system failure; if ΔP is in the range of [2.5Pa, 5Pa], it is determined to be a partial blockage or fan performance degradation; if ΔP is greater than 5Pa, it is determined to be a clear heat dissipation channel, and proceed to step 2.

[0076] 2. Based on equipment load and gas data, distinguish between internal heat sources and the external environment. When the heat dissipation channel is unobstructed, obtain the real-time total power consumption Pt of all servers in the rack and the historical average power consumption Pavg of the past hour from the IT operation and maintenance system via an interface. If Pt > Pavg × If step 1.2 is initially determined to be a temperature rise caused by normal load fluctuations, only this abnormal event is recorded. If the difference between Pt and Pavg is within ±10%, the server load is determined to be not the main cause, and the peak value of the TVOC sensor in the gas sensor array for the most recent minute is read. If the TVOC concentration exceeds 1.6 times the preset baseline, and the Pearson correlation coefficient between the TVOC concentration change rate and the temperature rise rate is greater than 0.6, it is determined to be component thermal aging or overheating gas release. After determining that it is component thermal aging or overheating gas release, the cause is further subdivided. If the ozone concentration exceeds 0.08ppm at the same time, it is determined to be a power module abnormality; otherwise, it is determined to be aging of the electrolytic capacitor array, and the aging index is analyzed. The aging index = (To-Tb) / Tb×(TVOC / TVOCb), where Tb represents the preset baseline of the temperature sensor value, and TVOCb represents the preset baseline of the TVOC sensor reading. If the aging index exceeds 0.5, a high-priority maintenance warning is triggered; if it is below 0.3, a medium-priority attention is triggered. If the TVOC concentration is normal, step 3 is executed.

[0077] 3. Correlate the intake air environment and dust data, read the intake air temperature Ti and the dust particle concentration D in the intake area of ​​the rack. If To-Ti is greater than 15℃ and D exceeds 5×10^4 particles / m³, it is determined that the intake air filter is severely clogged. If To-Ti is not greater than 15℃, but Ti exceeds its recommended operating temperature limit, it is determined that the data center cooling is insufficient, and the return air temperature of the air conditioning in the air conditioning area of ​​the cold aisle where the rack is located is checked. If multiple racks show this phenomenon at the same time, an alarm for degraded air conditioning system performance is triggered.

[0078] The above three steps are used to determine the cause of the abnormally high temperature at the rack exhaust vent.

[0079] Attribution analysis of physical safety incidents based on vibration and door magnets:

[0080] When the miniature vibration sensor in rack R2 detects a peak axial composite acceleration exceeding 0.1g, the event analysis process is initiated:

[0081] 1. Confirm the synchronization between the door status and the event, immediately read the status change history of the door magnetic switch, and establish an observation window of [-2 seconds, +2 seconds] centered on the vibration peak. If the door magnetic switch changes from closed to open within the observation window, it is marked as door opening accompanied by vibration, and step 2 is executed; if the door magnetic switch remains closed, it is marked as abnormal vibration of non-open door, and step 3 is executed.

[0082] 2. Based on infrared array and vibration waveform recognition, human behavior pattern recognition is performed. When the door is confirmed to be open, waveform features from the vibration sensor are extracted within 5 seconds after the door is opened.

[0083] Calculate the effective vibration value (RMS) and duration (Tv);

[0084] Analyze the changes in heat sources of the infrared array sensor within 10 seconds before and after the door is opened, including the centroid movement trajectory and area changes of the pixel clusters of heat sources of people;

[0085] If the vibration waveform is a single narrow pulse with a half-width at half-maximum (FWHM) of less than 0.3 seconds and an RMS of less than 0.15g, and infrared data shows that a human heat source remains near the rack for less than 15 seconds, it is considered a rapid plug-in / plug-out operation, recorded as a routine maintenance action, and no alarm is triggered. If the vibration waveform is multiple continuous pulses with an interval of 0.2 to 0.5 seconds, an RMS between 0.2g and 0.8g, and a duration exceeding 3 seconds, and infrared data shows that the heat source remains in front of the rack for more than 1 minute, it is considered a prolonged operation, triggering a mid-level audit alarm, and the operator's entry and exit work order information is recorded. If the vibration waveform is low-frequency with high amplitude (peak value greater than 0.8g, frequency less than 5Hz), and infrared data shows two or more heat sources, it is considered a multi-person collaborative equipment handling operation, triggering a high-level audit alarm, requiring management personnel to confirm whether it is a planned change.

[0086] 3. Abnormal vibration analysis when the door is not open: When vibration occurs while the door magnet remains closed, analyze the vibration frequency components: If the frequency is concentrated between 20Hz and 200Hz, and the amplitude shows a regular periodicity, such as once every 0.1 seconds, it is determined that the fan or hard drive inside the cabinet is vibrating abnormally, and a hardware health check is notified; If the vibration is a single impact with a peak value exceeding 1g and a duration of less than 0.1 seconds, it is determined to be an external impact, and a high-priority alarm is immediately pushed to security personnel, and the cameras in the area are linked to record and play back the video; If the vibration occurs more than 3 times within 10 seconds, but the peak value is between 0.1g and 0.3g, it is determined that there is a vibration source nearby.

[0087] Specifically, in step S3 of this embodiment, the logical rule engine established based on the IoT event topology map can collaboratively match and judge the data change patterns of each monitored object within the event tracking window, and output clear event attributions. Through multi-step reasoning analysis of typical abnormal scenarios such as abnormal cabinet air outlet temperature, vibration, and door magnetic linkage safety events, it can distinguish different causes such as ventilation failure, load fluctuation, component aging, filter blockage, external impact, and routine maintenance operations, realizing an automated closed loop from data anomaly to cause location. This rule engine has good scalability and interpretability, significantly improving the accuracy and intelligence level of anomaly diagnosis.

[0088] Step S4: Develop a tiered early warning system based on the results of event attribution.

[0089] Specifically, in step S4 of this embodiment, after completing the attribution analysis, the deviation and deterioration rate of different events are analyzed:

[0090] For temperature-related anomalies, set the deviation to (current temperature - upper limit of normal operating temperature) / upper limit of normal operating temperature;

[0091] For minor pressure difference abnormalities, set the deviation as (lower limit of normal pressure difference - current pressure difference) / lower limit of normal pressure difference;

[0092] For abnormal gas concentrations, set the deviation as (current concentration - upper limit of normal concentration) / upper limit of normal concentration;

[0093] For abnormal vibration, set the deviation as (peak vibration - normal vibration threshold) / normal vibration threshold;

[0094] The deviation is calculated to be between 0 and 2. A deviation greater than 2 is counted as 2. A deviation less than 0.2 is marked as slight deviation, a deviation between 0.2 and 0.6 is marked as moderate deviation, and a deviation greater than 0.6 is marked as severe deviation.

[0095] The method for analyzing the deterioration rate is as follows: starting from the event origin, take the sensor readings of the five most recent sampling points, and use linear least squares fitting to obtain the slope of change; for vibration anomalies, the impact frequency is used to replace its deterioration rate.

[0096] Specifically, in step S4 of this embodiment, a graded early warning is performed based on the event attribution results, deviation, and deterioration rate:

[0097] A Level 1 warning is defined as one that meets one of the following conditions:

[0098] Condition 1: The event attribution is classified as high-risk and critical, and is marked as severely deviated;

[0099] Condition 2: The rate of deterioration reaches its corresponding first rate threshold;

[0100] Condition 3: The vibration and impact frequency is greater than or equal to 5 times / 10 seconds and the peak value of a single impact is greater than or equal to 0.5g;

[0101] Condition 4: TVOC and ozone simultaneously trigger a Level 1 or Level 2 warning;

[0102] For a Level II warning, a warning is defined as one that meets one of the following conditions but does not reach the Level I warning level:

[0103] Condition 1: The event attribution is of the critical type and is marked as moderate deviation, and the rate of deterioration has not reached its corresponding first rate threshold;

[0104] Condition 2: The rate of deterioration reaches its corresponding second rate threshold;

[0105] Condition 3: Deviation is greater than or equal to 0.4 but does not meet the conditions for a Level 1 warning;

[0106] A Level 3 warning is defined as one that meets one of the following conditions but does not meet the requirements of a Level 1 or Level 2 warning:

[0107] Condition 1: Marked as slightly deviated and the rate of deterioration has not reached its corresponding second rate threshold;

[0108] Condition 2: The time attribution is due to temperature rise caused by normal load fluctuations or routine maintenance actions;

[0109] Condition 3: The cumulative area of ​​gas concentration exceeding the limit is less than 10 ppm·s, and the peak growth rate is less than 0.3.

[0110] The event attribution falls under the high-risk / critical category, including:

[0111] If the TVOC concentration exceeds 0.8 ppm or the ozone concentration exceeds 0.08 ppm, and the peak growth rate is greater than 1, the peak growth rate is calculated by setting a sliding time window with a width of 2 minutes, recording the maximum concentration value Cp1 within the sliding time window, and the maximum concentration value Cp2 of the previous window, and setting the peak growth rate = (Cp1-Cp2) / Cp2. It is valid when Cp2 is greater than 0; if Cp2=0 and Cp1 is greater than 0, the peak growth rate is considered to be infinite.

[0112] Peak acceleration greater than or equal to 0.5g, and impact frequency greater than or equal to 3 times / 10 seconds;

[0113] ΔP is less than 0.5 Pa, and the short-term mutation rate of ΔP is greater than or equal to 1.5 Pa / 10 seconds;

[0114] The deterioration rate reaching its corresponding first rate threshold includes:

[0115] The rate of temperature deterioration is greater than or equal to 0.5℃ / min;

[0116] The peak growth rate of TVOC concentration is greater than or equal to 1 and the current concentration exceeds 0.5 ppm;

[0117] The peak ozone concentration growth rate is greater than or equal to 1 and the current concentration exceeds 0.05 ppm;

[0118] The short-term mutation rate of ΔP is greater than or equal to 1.5 Pa / 10 seconds, and the pressure difference is less than 2.0 Pa after it decreases.

[0119] The cumulative over-limit area is greater than or equal to 30 ppm·second. The method for calculating the cumulative over-limit area is to multiply the over-limit amount of each sampling point by the sampling interval and then sum them up from the first time the gas concentration exceeds the normal upper limit until the concentration falls back below the upper limit, so as to obtain the cumulative over-limit area.

[0120] The event attribution falls under the critical type, including:

[0121] The long-term trend rate of ΔP is greater than or equal to 1.0 Pa / hour, or ΔP is between 0.5 Pa and 2.5 Pa;

[0122] The temperature deviation is greater than or equal to 0.4, but there is no gas release;

[0123] The rack door is open for 5 minutes or more, and the vibration is regular with a peak value of less than 0.5g;

[0124] The deterioration rate reaching its corresponding second rate threshold includes:

[0125] The rate of temperature deterioration is between 0.1 and 0.5 °C / min;

[0126] The long-term trend rate of ΔP is greater than or equal to 1.0 Pa / hour for 2 consecutive hours;

[0127] The short-term mutation rate of ΔP is between 0.5 and 1.5 Pa / 10 seconds, and is below 3 Pa after the pressure difference decreases;

[0128] The peak growth rate of TVOC is between 0.3 and 1.0, and the current concentration exceeds 0.5 ppm;

[0129] The peak ozone growth rate is between 0.3 and 1.0, and the current concentration exceeds 0.05 ppm;

[0130] The cumulative area exceeding the limit is between 10 and 30 ppm·s;

[0131] The Level 1 warning is an emergency warning that requires immediate action; the Level 2 warning is an important warning that requires prompt action; and the Level 3 warning is a general warning that only records a notification.

[0132] Specifically, in step S4 of this embodiment, after completing the attribution analysis, the event deviation degree and deterioration rate are further calculated to construct a three-level graded early warning mechanism. The deviation degree quantifies the degree of deviation of the anomaly from the normal range, and the deterioration rate reflects the event development trend through linear fitting. The combination of the two makes the early warning no longer dependent on a single static threshold, but dynamically assesses the severity and urgency of the event. Level 1 early warning is used for high-risk and critical events that deviate significantly or deteriorate rapidly, Level 2 early warning is used for medium-risk events, and Level 3 early warning is used for minor or routine events. This enables graded responses to events of different levels, helps maintenance personnel to rationally allocate processing priorities, reduces unnecessary emergency response costs, and improves overall maintenance efficiency and data center security capabilities.

[0133] Please see Figure 2 As shown, this is an IoT-based integrated monitoring device for communication equipment rooms in this embodiment, including:

[0134] The data acquisition unit is used to collect data from the communication equipment room through the sensing nodes in the equipment room;

[0135] The correlation analysis unit is used to build a three-dimensional correlation model based on data center data to obtain an IoT event topology map;

[0136] The event attribution unit is used to build a logical rule engine based on the IoT event topology map for event attribution.

[0137] The tiered early warning unit is used to issue tiered early warnings based on the results of event attribution.

[0138] This application also provides a computer-readable storage medium storing instructions that, when run on a computer, cause the computer to execute the IoT-based integrated monitoring method for communication equipment rooms as described in the above method embodiments.

[0139] It will be understood by those skilled in the art that all or some of the steps and systems in the methods disclosed above can be implemented as software, firmware, hardware, and suitable combinations thereof. Some or all of the physical components can be implemented as software executed by a processor, such as a central processing unit, digital signal processor, or microprocessor, or as hardware, or as an integrated circuit, such as an application-specific integrated circuit. Such software can be distributed on a computer-readable medium, which can include computer storage media (or non-transitory media) and communication media (or transient media). As is known to those skilled in the art, the term computer storage media includes volatile and non-volatile, removable and non-removable media implemented in any method or technology for storing information (such as computer-readable programs, data structures, program modules, or other data). Computer storage media includes, but is not limited to, RAM, ROM, EEPROM, flash memory or other memory technologies, CD-ROM, digital versatile disc (DVD) or other optical disc storage, magnetic cartridges, magnetic tape, disk storage or other magnetic storage devices, or any other medium that can be used to store desired information and is accessible to a computer. Furthermore, as is known to those skilled in the art, communication media typically contain computer-readable programs, data structures, program modules, or other data in modulated data signals such as carrier waves or other transmission mechanisms, and may include any information delivery medium.

[0140] In the above description, the disclosure of this invention is not intended to limit itself to these aspects. Rather, within the scope of the objectives of this disclosure, components can be selectively and operationally combined in any number. Furthermore, terms such as “comprising,” “encompassing,” and “having” should be interpreted by default as inclusive or open-ended, rather than exclusive or closed, unless explicitly defined as such. All technical, scientific, or other terms are to be understood by those skilled in the art, unless defined as such. Public terms found in dictionaries should not be interpreted in the context of the relevant technical documents in an overly idealistic or impractical manner, unless explicitly defined as such in this disclosure. Any modifications or alterations made by those skilled in the art based on the foregoing disclosure are within the scope of the claims.

Claims

1. An Internet of Things-based communication machine room integrated monitoring method, characterized in that, include: Data is collected from the communication equipment room through sensing nodes; A three-dimensional correlation model is constructed based on data from the computer room to obtain an IoT event topology map; A logical rule engine is built based on the IoT event topology map to perform event attribution; Graded early warning is based on the results of event attribution. 2.The Internet of Things based communication machine room integrated monitoring method according to claim 1, characterized in that, Establish an IoT event topology map. For the spatial index layer, divide the physical space of the communication room into a cubic grid with a side length of 0.5 meters, and assign a spatial index ID to each grid. Map the physical coordinates of all sensing nodes to the corresponding grid to bind the sensing nodes with the spatial index ID. For the logical association layer, association tuples are generated based on the logical affiliation of the sensing nodes to classify sensing nodes belonging to the same device. For time slices and event streams, a circular buffer with a preset buffer duration is maintained. Sensing nodes belonging to the same monitoring object store the latest sensor readings and state change events in the circular buffer in chronological order. When the sensor value of any sensing node of a monitoring object exceeds its preset baseline or a state jump occurs, it is marked as a potential event and a time tracking window is set to continuously monitor the data change trend of all sensing nodes belonging to the same monitoring object within the time tracking window. 3.The Internet of Things based communication machine room integrated monitoring method according to claim 2, characterized in that, Collaborative analysis is performed on the associated tuples in the IoT event topology graph, and judgment is made through a preset logical rule engine. The preset logical rule engine matches the data change pattern of each monitored object within the event tracking window to output event attribution. 4.The Internet of Things based communication machine room integrated monitoring method according to claim 3, characterized in that, After completing the attribution analysis, the deviation and deterioration rate of different events are analyzed. The deviation is calculated between 0 and 2. A deviation greater than 2 is counted as 2. A deviation less than 0.2 is marked as slight deviation, a deviation between [0.2, 0.6] is marked as moderate deviation, and a deviation greater than 0.6 is marked as severe deviation. The method for analyzing the deterioration rate is as follows: starting from the event origin, take the sensor readings of the five most recent sampling points, and use linear least squares fitting to obtain the slope of change; for vibration anomalies, use the impact frequency to replace its deterioration rate. 5.The Internet of Things based communication machine room integrated monitoring method according to claim 4, characterized in that, The early warning system is tiered based on the outcome of event attribution, deviation, and rate of deterioration, including Level 1, Level 2, and Level 3 warnings. 6.The Internet of Things based communication machine room integrated monitoring method according to claim 5, characterized in that, A Level 1 warning is defined as one that meets one of the following conditions: Condition 1: The event attribution is classified as high-risk and critical, and is marked as severely deviated; Condition 2: The rate of deterioration reaches its corresponding first rate threshold; Condition 3: The vibration and impact frequency is greater than or equal to 5 times / 10 seconds and the peak value of a single impact is greater than or equal to 0.5g; Condition 4: TVOC and ozone simultaneously trigger a Level 1 or Level 2 warning. 7.The Internet of Things based communication machine room integrated monitoring method according to claim 6, characterized in that, For a Level II warning, a warning is defined as one that meets one of the following conditions but does not reach the Level I warning level: Condition 1: The event attribution is of the critical type and is marked as moderate deviation, and the rate of deterioration has not reached its corresponding first rate threshold; Condition 2: The rate of deterioration reaches its corresponding second rate threshold; Condition 3: Deviation is greater than or equal to 0.4 but does not meet the conditions for a Level 1 warning.

8. The method for integrated monitoring of communication equipment rooms based on the Internet of Things according to claim 7, characterized in that, A Level 3 warning is defined as one that meets one of the following conditions but does not meet the requirements of a Level 1 or Level 2 warning: Condition 1: Marked as slightly deviated and the rate of deterioration has not reached its corresponding second rate threshold; Condition 2: The time attribution is due to temperature rise caused by normal load fluctuations or routine maintenance actions; Condition 3: The cumulative area of ​​gas concentration exceeding the limit is less than 10 ppm·s, and the peak growth rate is less than 0.

3.

9. A communication machine room integrated monitoring device based on Internet of Things, applied to the communication machine room integrated monitoring method based on Internet of Things in any one of claims 1-8, characterized in that, include: The data acquisition unit is used to collect data from the communication equipment room through the sensing nodes in the equipment room; The correlation analysis unit is used to build a three-dimensional correlation model based on data center data to obtain an IoT event topology map; The event attribution unit is used to build a logical rule engine based on the IoT event topology map for event attribution. The tiered early warning unit is used to issue tiered early warnings based on the results of event attribution.

10. A storage medium, characterized by The device stores instructions that, when executed on a computer, cause the computer to perform the IoT-based integrated monitoring method for communication equipment rooms as described in any one of claims 1-8.

Images