Data center power environment monitoring system based on multi-sensor fusion

Abstract

The invention discloses a data center dynamic environment monitoring system based on multi-sensor fusion, and belongs to the technical field of data center intelligent monitoring, and the monitoring system comprises an infrared thermal imaging camera which is used for collecting the infrared image information of the side surface of a cabinet in a data center; the temperature analysis unit is used for calculating and acquiring the average temperature of each rectangular block on the side surface of the cabinet according to the infrared image information so as to acquire the time and space distribution of the rectangular blocks with abnormal heat dissipation in the cabinet; and the control center is used for adjusting the sampling frequency of each temperature sensor according to the time and space distribution of the rectangular blocks with abnormal heat dissipation in the cabinet. According to the invention, efficient configuration of resources is realized. Redundant data caused by fixed sampling is avoided, the storage and calculation burden is reduced, high-frequency data in the abnormal time period can be accurately captured, and the monitoring efficiency is improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of intelligent monitoring of data centers, and specifically relates to a data center power environment monitoring system based on multi-sensor fusion. Background Art

[0002] Data centers require a large number of sensors to monitor their environment and identify anomalies within the data center. Existing technologies for monitoring data centers primarily use two types of sensors: wired and wireless. Wired sensors offer a stable power supply and enable long-term, reliable detection. However, wired sensors have deployment limitations, requiring the installation of numerous data and power cables, increasing construction complexity and costs. Furthermore, deploying wired sensors within cabinets can lead to localized heat accumulation due to the extra wiring, impacting the stable operation of IT equipment. Furthermore, wired systems struggle to adapt to dynamic adjustments in data centers, such as cabinet relocations or equipment additions and subtractions.

[0003] Among them, wireless sensors have high power consumption and require frequent battery replacement or reliance on external power supply. They are not suitable for long-term deployment in confined spaces such as cabinets.

[0004] In order to solve the above problems, the present invention provides the following technical solutions. Summary of the Invention

[0005] The purpose of the present invention is to provide a data center power environment monitoring system based on multi-sensor fusion to solve the problems in the prior art that laying lines for wired sensors is difficult and costly and affects heat dissipation, and wireless sensors have high power consumption and require frequent battery replacement or external power supply, making them unsuitable for long-term deployment in confined spaces such as inside cabinets.

[0006] The purpose of the present invention can be achieved through the following technical solutions:

[0007] The data center power environment monitoring system based on multi-sensor fusion includes:

[0008] Infrared thermal imaging camera, used to collect infrared image information of the side of the cabinet in the data center;

[0009] The temperature analysis unit is used to calculate and obtain the average temperature of each rectangular block on the side of the cabinet based on the infrared image information, thereby obtaining the time and spatial distribution of the rectangular blocks in the cabinet where heat dissipation abnormalities occur;

[0010] The control center adjusts the sampling frequency of each temperature sensor according to the time and space distribution of the rectangular blocks with abnormal heat dissipation in the cabinet.

[0011] As a further solution of the present invention, a method for the control center to adjust the layout position of the temperature sensors and the sampling frequency of each temperature sensor includes the following steps:

[0012] Step 1: Use an infrared thermal imaging camera to capture thermal imaging frames of the cabinet surface at preset intervals of time t1;

[0013] Step 2: For a cabinet, divide its detection surface into several rectangular blocks;

[0014] Step 3: Obtain the abnormally stable period. Within the same abnormally stable period in different cycles, the distribution of rectangular blocks with heat dissipation anomalies is similar.

[0015] Step 4: For a rectangular block, obtain the frequency H1 of being marked as heat dissipation abnormality during each abnormal stable period;

[0016] The data acquisition frequency H2 of the temperature sensor associated with the rectangular block is adjusted according to the frequency H1 corresponding to the rectangular block, which is manifested as the greater the frequency H1, the greater the frequency H2.

[0017] As a further solution of the present invention, the frequency H3 of the rectangular block being marked as heat dissipation abnormality in a time period that does not belong to the abnormally stable period is calculated, and the data acquisition frequency H2 of the temperature sensor associated with the rectangular area within the time range is adjusted according to H3, which is expressed as the larger H3 is, the larger H2 is.

[0018] As a further solution of the present invention, a method for obtaining an abnormally stable period is:

[0019] Step 31. Establish a plane rectangular coordinate system with time as the horizontal coordinate and the sequence number of the rectangular block as the vertical coordinate;

[0020] Step 32: When analyzing the thermal imaging frame, if the temperature of a rectangular block is greater than a preset threshold, it is considered that there is heat dissipation anomaly in the corresponding rectangular block when the thermal imaging frame is acquired. At this time, the rectangular block is marked in the plane rectangular coordinate system;

[0021] Step 33, obtaining a heat dissipation anomaly mark map of each rectangular block in a cycle;

[0022] Step 34, obtaining a heat dissipation anomaly mark corresponding to m cycles collected within a preset time period in the past;

[0023] Then, the time of one cycle is evenly divided into k comparative analysis periods, and each comparative analysis period is marked as Tij, where i ranges from 1 to m and j ranges from 1 to k;

[0024] The m comparative analysis periods with different i and the same j value are regarded as a comparative analysis group;

[0025] Step 35: Perform similarity comparison on the m comparative analysis periods in a comparative analysis group. When the similarity is greater than a preset threshold, it is considered that the comparative analysis periods in the comparative group are consistent.

[0026] The distribution within a complete cycle is divided into several abnormally stable periods within a complete cycle. In an abnormally stable period, the proportion of the sum of the corresponding durations of all consistent comparative analysis groups to the total duration of the abnormally stable period should be greater than the preset threshold β.

[0027] As a further solution of the present invention, the threshold β is set to 0.6.

[0028] As a further solution of the present invention, when a temperature sensor corresponds to two or more rectangular blocks, when adjusting the sampling frequency of the temperature sensor, the frequency value of the rectangular block with the highest frequency of being marked as having abnormal heat dissipation shall prevail.

[0029] Beneficial effects of the present invention:

[0030] 1. This invention achieves efficient resource allocation by dynamically adjusting sensor sampling frequency through collaborative analysis of infrared thermal imaging and temperature sensors. This avoids redundant data caused by fixed sampling, reduces storage and computing burdens, and accurately captures high-frequency data during abnormal periods, improving monitoring efficiency.

[0031] 2. This invention uses similarity comparisons across multiple periods of historical data to identify the spatial distribution and temporal patterns of heat dissipation anomalies, reducing accidental misjudgments. Combining the temperature distribution of infrared images with point temperature sensor data, it cross-validates abnormal areas and improves detection reliability.

[0032] 3. The present invention adjusts the sampling frequency according to the abnormal frequency, obtains more monitoring resources in the high-frequency abnormal area, reduces redundant collection in the low-frequency area, and optimizes system energy consumption. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] The present invention will be further described below with reference to the accompanying drawings.

[0034] Figure 1 It is a flow chart of the method for the control center to adjust the sampling frequency of each temperature sensor in the present invention. DETAILED DESCRIPTION

[0035] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making any creative efforts shall fall within the scope of protection of the present invention.

[0036] The data center power environment monitoring system based on multi-sensor fusion is characterized by including:

[0037] Infrared thermal imaging camera, used to collect infrared image information of the side of the cabinet in the data center;

[0038] The temperature analysis unit is used to calculate and obtain the average temperature of each rectangular block on the side of the cabinet based on the infrared image information, thereby obtaining the time and spatial distribution of the rectangular blocks in the cabinet where heat dissipation abnormalities occur;

[0039] The control center adjusts the sampling frequency of each temperature sensor according to the time and space distribution of the rectangular blocks with abnormal heat dissipation in the cabinet.

[0040] like Figure 1 As shown, the method for the control center to adjust the sampling frequency of each temperature sensor includes the following steps:

[0041] Step 1: Use an infrared thermal imaging camera to capture the cabinet surface image;

[0042] Specifically, the infrared thermal imaging camera collects an image of the cabinet surface every preset time t1 to obtain a number of thermal imaging image frames;

[0043] Step 2: For a cabinet, divide its detection surface (generally refers to the side photographed by the infrared thermal imaging camera) into several rectangular blocks.

[0044] When dividing the rectangular blocks, one rectangular block may cover an integer number of servers, specifically one or more servers;

[0045] Step 3: Obtain the abnormally stable period. In the same abnormally stable period of different cycles, the distribution of rectangular blocks with heat dissipation anomalies is consistent and similar.

[0046] Specifically:

[0047] Step 31. Establish a plane rectangular coordinate system with time as the horizontal coordinate and the sequence number of the rectangular block as the vertical coordinate. Specifically, the starting point of the horizontal coordinate of the plane rectangular coordinate system is the starting time of a monitoring cycle, and the end point of the horizontal coordinate is the end time of a monitoring cycle;

[0048] Among them, a monitoring cycle is generally one day;

[0049] Step 32: When analyzing a thermal imaging frame, if the temperature of a rectangular block is greater than a preset threshold, it is considered that there is a heat dissipation anomaly in the corresponding rectangular block when the thermal imaging frame is acquired. In this case, the rectangular block is marked in the plane rectangular coordinate system;

[0050] Step 33, according to the method in step 32, obtain the heat dissipation abnormality mark map of each rectangular block in a cycle;

[0051] Step 34, obtaining a heat dissipation anomaly mark corresponding to m cycles collected within a preset time period in the past;

[0052] Then, the time of one cycle is evenly divided into k comparative analysis periods, and each comparative analysis period is marked as Tij, where i ranges from 1 to m, j ranges from 1 to k, and Tij represents the jth comparative analysis period in the heat dissipation anomaly marking diagram corresponding to the i-th cycle;

[0053] The m comparative analysis periods with different i and the same j value are regarded as a comparative analysis group;

[0054] Step 35: Perform similarity comparison on the m comparative analysis periods in a comparative analysis group. When the similarity is greater than a preset threshold, it is considered that the comparative analysis periods in the comparative group are consistent.

[0055] Whether the comparative analysis periods in other comparative analysis groups are consistent is judged in turn, thereby obtaining all consistent comparative analysis groups;

[0056] Obtain the time ranges corresponding to each consistent comparative analysis group within a complete cycle, and then divide the complete cycle into several abnormally stable periods based on the distribution of these time ranges within the complete cycle. In an abnormally stable period, the proportion of the sum of the corresponding durations of all consistent comparative analysis groups to the total duration of the abnormally stable period should be greater than the preset threshold β;

[0057] The threshold β is preferably set to 0.6;

[0058] In the same abnormal stable period of different cycles, the distribution of rectangular blocks with heat dissipation anomalies is regular and consistent, that is, the distribution positions are similar.

[0059] Step 4: For a rectangular block, obtain the frequency H1 of being marked as heat dissipation abnormality during each abnormal stable period;

[0060] According to the frequency H1 corresponding to the rectangular block, the data acquisition frequency H2 of the temperature sensor associated with the rectangular block is adjusted. Specifically, the larger the frequency H1, the larger the frequency H2;

[0061] The frequency H3 of the rectangular block being marked as heat dissipation abnormality in the time period that is not an abnormally stable period is calculated, and the data collection frequency H2 of the temperature sensor associated with the rectangular area within the time range is adjusted according to H3. Similarly, the larger H3 is, the larger H2 is.

[0062] It should be noted that when a temperature sensor corresponds to two or more rectangular blocks, when adjusting the sampling frequency of the temperature sensor, the frequency value of the rectangular block with the highest frequency of being marked as heat dissipation abnormality shall prevail.

[0063] In addition, the temperature sensor mentioned here refers to a wireless temperature sensor. Through the implementation of the above solution, compared with the method of setting a fixed temperature sampling frequency in traditional technology, the sampling frequency of the wireless temperature sensor can be reasonably adjusted within different time ranges, reducing its sampling frequency during low-frequency abnormal periods and increasing its sampling frequency during high-frequency abnormal periods, thereby reducing the sampling energy consumption of the wireless temperature sensor while ensuring the timeliness of abnormality detection.

[0064] This achieves efficient resource allocation, avoiding redundant data caused by fixed sampling, reducing storage and computing burdens, while accurately capturing high-frequency data during abnormal periods and improving monitoring efficiency.

[0065] The above contents are merely examples and explanations of the present invention. Those skilled in the art may make various modifications or additions to the described specific embodiments or replace them in similar ways. As long as they do not deviate from the invention or exceed the scope defined by the claims, they should all fall within the scope of protection of the present invention.

Claims

1. The data center power environment monitoring system based on multi-sensor fusion is characterized by: include: Infrared thermal imaging camera, used to collect infrared image information of the side of the cabinet in the data center; The temperature analysis unit is used to calculate and obtain the average temperature of each rectangular block on the side of the cabinet based on the infrared image information, thereby obtaining the time and spatial distribution of the rectangular blocks in the cabinet where heat dissipation abnormalities occur; The control center adjusts the sampling frequency of each temperature sensor according to the time and space distribution of the rectangular blocks with abnormal heat dissipation in the cabinet.

2. The data center power environment monitoring system based on multi-sensor fusion according to claim 1 is characterized in that: The method for the control center to adjust the sampling frequency of each temperature sensor includes the following steps: Step 1: Use an infrared thermal imaging camera to capture thermal imaging frames of the cabinet surface at preset intervals of time t1; Step 2: For a cabinet, divide its detection surface into several rectangular blocks; Step 3: Obtain the abnormally stable period. Within the same abnormally stable period in different cycles, the distribution of rectangular blocks with heat dissipation anomalies is similar. Step 4: For a rectangular block, obtain the frequency H1 of being marked as heat dissipation abnormality during each abnormal stable period; The data acquisition frequency H2 of the temperature sensor associated with the rectangular block is adjusted according to the frequency H1 corresponding to the rectangular block, which is manifested as the greater the frequency H1, the greater the frequency H2.

3. The data center power environment monitoring system based on multi-sensor fusion according to claim 2 is characterized in that: The frequency H3 of the rectangular block being marked as heat dissipation abnormality in the time period that is not an abnormally stable period is calculated, and the data collection frequency H2 of the temperature sensor associated with the rectangular area within the time range is adjusted according to H3. The larger H3 is, the larger H2 is.

4. The data center power environment monitoring system based on multi-sensor fusion according to claim 2 is characterized in that: The method for obtaining the abnormal stable period is: Step 31. Establish a plane rectangular coordinate system with time as the horizontal coordinate and the sequence number of the rectangular block as the vertical coordinate; Step 32: When analyzing the thermal imaging frame, if the temperature of a rectangular block is greater than a preset threshold, it is considered that there is heat dissipation anomaly in the corresponding rectangular block when the thermal imaging frame is acquired. At this time, the rectangular block is marked in the plane rectangular coordinate system; Step 33, obtaining a heat dissipation anomaly mark map of each rectangular block in a cycle; Step 34, obtaining a heat dissipation anomaly mark corresponding to m cycles collected within a preset time period in the past; Then, the time of one cycle is evenly divided into k comparative analysis periods, and each comparative analysis period is marked as Tij, where i ranges from 1 to m and j ranges from 1 to k; The m comparative analysis periods with different i and the same j value are regarded as a comparative analysis group; Step 35: Perform similarity comparison on the m comparative analysis periods in a comparative analysis group. When the similarity is greater than a preset threshold, it is considered that the comparative analysis periods in the comparative group are consistent. The distribution within a complete cycle is divided into several abnormally stable periods within a complete cycle. In an abnormally stable period, the proportion of the sum of the corresponding durations of all consistent comparative analysis groups to the total duration of the abnormally stable period should be greater than the preset threshold β.

5. The data center power environment monitoring system based on multi-sensor fusion according to claim 4 is characterized in that: The threshold β is set to 0.

6.

6. The data center power environment monitoring system based on multi-sensor fusion according to any one of claims 2 or 3, characterized in that: When a temperature sensor corresponds to two or more rectangular blocks, when adjusting the sampling frequency of the temperature sensor, the frequency value of the rectangular block with the highest frequency of being marked as having abnormal heat dissipation shall prevail.

Images