A Machine Learning-Based Method and System for Patrol Inspection of Computer Room Operation and Maintenance

Abstract

The present invention relates to the technical field of computer room operation and maintenance inspection, and specifically discloses a computer room operation and maintenance inspection method and system based on machine learning. The system includes: a temperature abnormal device judgment module, a cooling fan speed adjustment module, a temperature continuously abnormal device judgment module, a heat dissipation abnormal operation and maintenance plan providing module, a database, a fault device judgment module, and a patrol route optimization module; The present invention conducts a secondary inspection on each abnormal inspection point in the target computer room to determine whether there are devices with continuously abnormal temperatures. If so, it confirms the cause of heat dissipation abnormality and provides an operation and maintenance plan for heat dissipation abnormality. At the same time, in combination with the actual status of the devices in the target computer room during the current period, it optimizes the preset patrol route of the target computer room, improves the timeliness of discovering and handling device temperature abnormalities, reduces the risk of device damage, and at the same time avoids causing the patrol robot to ignore key areas and repeatedly check normal areas, saving valuable patrol time.

Description

Technical Field

[0001] The present invention belongs to the technical field of computer room operation and maintenance inspection, and relates to a computer room operation and maintenance inspection method and system based on machine learning. Background Art

[0002] With the in-depth digital transformation, the computer room, as a key facility for enterprise IT services, its operation and maintenance inspection has become increasingly important. However, the traditional manual inspection method has problems such as low inspection efficiency, easy omission and misjudgment of inspections, which not only increases labor costs, but also may bring safety and economic risks. Therefore, it is particularly important to develop an efficient, accurate and automated computer room inspection method and system.

[0003] There are also the following problems in the existing methods for operation and maintenance inspection of computer rooms: 1. The equipment with abnormal temperature still existing in the computer room during the first inspection is not inspected for the second time, and the cause of the abnormal temperature of the equipment still existing is not analyzed in detail during the second inspection, which reduces the timeliness of discovering and handling equipment temperature anomalies, increases the risk of equipment damage, and thus increases the operation and maintenance costs of the enterprise.

[0004] 2. The inspection route of the inspection robot in the current computer room is set and unchanged, and the inspection route of the inspection robot in the computer room is not dynamically adjusted in combination with the actual state of the equipment in the current cycle of the computer room, which may cause the inspection robot to ignore key areas and repeatedly check normal areas, wasting valuable inspection time and resources, unable to meet the new inspection requirements in the computer room, and affecting the overall operation and maintenance level of the computer room. Summary of the Invention

[0005] In view of this, in order to solve the problems raised in the above background art, a computer room operation and maintenance inspection method and system based on machine learning are proposed.

[0006] The object of the present invention can be achieved by the following technical solutions: In the first aspect of the present invention, a computer room operation and maintenance inspection method based on machine learning is provided, including: S1. Judgment of equipment with abnormal temperature: Input the preset inspection route of the target computer room into the background of the target inspection robot. When the target inspection robot conducts inspections according to the preset inspection route, collect the thermal images of each device within the scope of each inspection point in the inspection route, and judge whether there is any equipment with abnormal temperature within the scope of each inspection point in the inspection route. If there is, execute step S2. If not, it indicates that the equipment has normal heat dissipation.

[0007] S2. Adjustment of the rotation speed of the cooling fan: Confirm each abnormal inspection point in the inspection route, extract the current rotation speed of the cooling fan corresponding to the scope of each abnormal inspection point, adjust the rotation speed of the cooling fan corresponding to the scope of each abnormal inspection point, and execute step S3.

[0008] S3. Judgment of equipment with continuously abnormal temperature: After the first inspection by the target inspection robot is completed, the positions of each abnormal inspection point are sent to the background of the target inspection robot again for secondary inspection. The thermal images of each temperature-abnormal equipment within the scope of each abnormal inspection point are collected during the secondary inspection to determine whether there is any equipment with continuously abnormal temperature within the scope of each abnormal inspection point. If there is, step S4 is executed; if not, it indicates that the equipment temperature has returned to normal.

[0009] S4. Provision of operation and maintenance plan for abnormal heat dissipation: Mark the equipment with continuously abnormal temperature as the target equipment, collect the heat dissipation grid images of each target equipment within the scope of each abnormal inspection point and the images of the corresponding heat dissipation fans within the scope of each abnormal inspection point. At the same time, collect the ambient temperature of each monitoring point corresponding to the target computer room, the air flow speed at the heat dissipation outlet, and the rotation speed of the heat dissipation fan, confirm the reasons for the abnormal heat dissipation of each target equipment, and provide an operation and maintenance plan for the abnormal heat dissipation.

[0010] S5. Judgment of faulty equipment: Collect the colors of the fault indicator lights of each equipment within the scope of each inspection point in the inspection route, and judge whether there is any faulty equipment within the scope of each inspection point. If there is, feedback on the faulty equipment is carried out.

[0011] S6. Optimization of inspection route: Count the number of temperature-abnormal equipment, the number of target equipment, and the number of faulty equipment that appear within the scope of each inspection point in the inspection route corresponding to each inspection day during the current inspection cycle, and optimize the preset inspection route of the target computer room.

[0012] In the second aspect of the present invention, a computer room operation and maintenance inspection system based on machine learning is provided, including: a temperature-abnormal equipment judgment module, which is used to input the preset inspection route of the target computer room into the background of the target inspection robot. When the target inspection robot conducts inspections according to the preset inspection route, collect the thermal images of each equipment within the scope of each inspection point in the inspection route, and judge whether there is any temperature-abnormal equipment within the scope of each inspection point in the inspection route.

[0013] A heat dissipation fan rotation speed adjustment module, which is used to confirm each abnormal inspection point in the inspection route, extract the current rotation speed of the heat dissipation fan corresponding to the scope of each abnormal inspection point, and adjust the rotation speed of the heat dissipation fan corresponding to the scope of each abnormal inspection point.

[0014] A temperature continuously abnormal equipment judgment module, which is used to, after the first inspection by the target inspection robot is completed, send the positions of each abnormal inspection point to the background of the target inspection robot again for secondary inspection, collect the thermal images of each temperature-abnormal equipment within the scope of each abnormal inspection point during the secondary inspection, and judge whether there is any temperature continuously abnormal equipment within the scope of each abnormal inspection point.

[0015] The heat dissipation anomaly operation and maintenance solution providing module is used to mark the devices with continuously abnormal temperatures as target devices, collect the heat dissipation grid images of each target device within the scope of each abnormal inspection point and the heat dissipation fan images corresponding to the scope of each abnormal inspection point, and at the same time collect the ambient temperature of each monitoring point corresponding to the target computer room, the air flow velocity of the heat dissipation ports, and the rotation speed of the heat dissipation fans, confirm the reasons for the heat dissipation anomalies of each target device, and provide operation and maintenance solutions for the heat dissipation anomalies.

[0016] The database is used to store the appropriate rotation speeds of the heat dissipation fans corresponding to each temperature anomaly index, and store the compensation rotation speeds required for the heat dissipation fans corresponding to the heat dissipation anomalies of the unit computer room.

[0017] The faulty device judgment module is used to collect the colors of the fault indicator lights of each device within the scope of each inspection point in the inspection route, and judge whether there are faulty devices within the scope of each inspection point.

[0018] The inspection route optimization module is used to count the number of devices with abnormal temperatures, the number of target devices, and the number of faulty devices that appear within the scope of each inspection point in the inspection routes corresponding to each inspection day during the current inspection period, and optimize the preset inspection routes of the target computer room.

[0019] Compared with the prior art, the embodiments of the present invention have at least the following advantages or beneficial effects:

[0020] (1) By performing secondary inspections on each abnormal inspection point in the target computer room, the present invention judges whether there are devices with continuously abnormal temperatures within the scope of each abnormal inspection point. If so, it confirms the reasons for the heat dissipation anomalies and provides operation and maintenance solutions for the heat dissipation anomalies, improving the timeliness of discovering and handling device temperature anomalies, reducing the risk of device damage, and thus reducing the operation and maintenance costs of the enterprise.

[0021] (2) By counting the number of devices with abnormal temperatures, the number of target devices, and the number of faulty devices that appear within the scope of each inspection point in the inspection routes corresponding to each inspection day during the current inspection period, the present invention optimizes the preset inspection routes of the target computer room, avoiding the inspection robot from ignoring key areas and repeatedly checking normal areas, saving valuable inspection time and resources, meeting the new inspection requirements in the computer room, and improving the overall operation and maintenance level of the computer room. Description of the Drawings

[0022] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0023] Figure 1Schematic diagram of the method steps of the present invention.

[0024] Figure 2 Schematic diagram of the system structure connection of the present invention.

[0025] Figure 3 Flow chart of abnormal temperature inspection of equipment of the present invention. Detailed implementation manners

[0026] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0027] Please refer to Figure 1 As shown, a machine room operation and maintenance inspection method based on machine learning is provided in the first aspect of the present invention, including: S1. Judgment of temperature-abnormal equipment: Input the preset inspection route of the target machine room into the background of the target inspection robot. When the target inspection robot conducts inspections according to the preset inspection route, collect the thermal images of each device within the scope of each inspection point in the inspection route, and judge whether there are temperature-abnormal devices within the scope of each inspection point in the inspection route. If so, execute step S2; if not, it indicates that the equipment has normal heat dissipation.

[0028] It should be noted that the thermal images of each device within the scope of each inspection point in the inspection route are collected by a thermal imager installed on the target inspection robot.

[0029] In a specific embodiment of the present invention, the specific process of judging whether there are temperature-abnormal devices within the scope of each inspection point in the inspection route is as follows: A1. Locate the temperature values corresponding to each temperature distribution area from the thermal images of each device within the scope of each inspection point in the inspection route, calculate their mean values, and obtain the average temperature of each device within the scope of each inspection point, denoted as where i represents the number of the inspection point, i = 1, 2,..., n, and j represents the number of the device, j = 1, 2,..., m.

[0030] A2. Extract the maximum value from the temperature values corresponding to each temperature distribution area of each device within the scope of each inspection point in the inspection route, and denote it as

[0031] A3. Calculate the temperature anomaly index β of each device within the scope of each inspection point ij , where W′ represents the set reference device temperature, and e represents the natural constant.

[0032] A4. Compare the temperature anomaly indices of the devices within the scope of each inspection point with the set reference temperature anomaly index. If the temperature anomaly index of a device within the scope of a certain inspection point is greater than or equal to the set reference temperature anomaly index, then mark this device as a temperature anomaly device, and thus determine whether there are temperature anomaly devices within the scope of each inspection point in the inspection route.

[0033] S2. Adjust the rotation speed of the cooling fan: Confirm each abnormal inspection point in the inspection route, extract the current rotation speed of the cooling fan corresponding to the scope of each abnormal inspection point, adjust the rotation speed of the cooling fan corresponding to the scope of each abnormal inspection point, and execute step S3.

[0034] It should be noted that the method for confirming each abnormal inspection point in the inspection route is: when there is any temperature anomaly device within the scope of a certain inspection point in the inspection route, mark this inspection point as an abnormal inspection point.

[0035] In a specific embodiment of the present invention, the specific process of adjusting the rotation speed of the cooling fan corresponding to the scope of each abnormal inspection point is: B1. Denote the current rotation speed of the cooling fan corresponding to the scope of each abnormal inspection point as v g , where g represents the number of the abnormal inspection point, g = 1, 2,..., p.

[0036] B2. Extract the maximum value from the temperature anomaly indices of the temperature anomaly devices within the scope of each abnormal inspection point, and compare it with the appropriate cooling fan rotation speed corresponding to each temperature anomaly index stored in the database to obtain the appropriate cooling fan rotation speed corresponding to the scope of each abnormal inspection point, and denote it as

[0037] B3. Set the value of the rotation speed that needs to be increased for the cooling fan corresponding to the scope of each abnormal inspection point

[0038]

[0039] S3. Judge the temperature continuously abnormal devices: After the first inspection of the target inspection robot is completed, send the positions of each abnormal inspection point to the background of the target inspection robot for secondary inspection again, collect the thermal images of the temperature abnormal devices within the scope of each abnormal inspection point during the secondary inspection, and judge whether there are temperature continuously abnormal devices within the scope of each abnormal inspection point. If there are, execute step S4. If not, it indicates that the device temperature has returned to normal.

[0040] It should be noted that the thermal images of the temperature abnormal devices within the scope of each abnormal inspection point during the secondary inspection are also collected by the thermal imager installed on the target inspection robot.

[0041] In a specific embodiment of the present invention, the method for determining whether there are temperature continuously abnormal devices within the scope of each abnormal inspection point is as follows: Based on the thermal images of each temperature abnormal device within the scope of each abnormal inspection point in the secondary inspection, the method for determining whether there are temperature abnormal devices within the scope of each inspection point in the inspection route is used to similarly determine whether there are temperature continuously abnormal devices within the scope of each abnormal inspection point.

[0042] S4. Providing an operation and maintenance plan for abnormal heat dissipation: Mark the temperature continuously abnormal devices as target devices, collect the heat dissipation grid images of each target device within the scope of each abnormal inspection point and the heat dissipation fan images corresponding to the scope of each abnormal inspection point, and at the same time collect the ambient temperature of each monitoring point corresponding to the target computer room, the air flow velocity of the heat dissipation port, and the rotation speed of the heat dissipation fan, confirm the reasons for the abnormal heat dissipation of each target device, and provide an operation and maintenance plan for the abnormal heat dissipation.

[0043] It should be noted that the heat dissipation grid images of each target device within the scope of each abnormal inspection point and the heat dissipation fan images corresponding to the scope of each abnormal inspection point are all collected by the installed high-definition cameras, the ambient temperature of each monitoring point corresponding to the target computer room and the air flow velocity of the heat dissipation port are respectively collected by temperature sensors and anemometers, and the rotation speed of the heat dissipation fan is extracted from the display screen of the heat dissipation fan.

[0044] In a specific embodiment of the present invention, the specific process of confirming the reasons for the abnormal heat dissipation of each target device is as follows: C1. Based on the heat dissipation grid images of each target device within the scope of each abnormal inspection point, calculate the heat dissipation grid abnormal index χ of each target device within the scope of each abnormal inspection point gf , where f represents the number of the target device, f = 1, 2,..., z.

[0045] It should be noted that the specific process of calculating the heat dissipation grid abnormal index of each target device within the scope of each abnormal inspection point is as follows: Mark the leftmost heat dissipation grid in the heat dissipation grid image of each target device within the scope of each abnormal inspection point as the target heat dissipation grid, and mark the other heat dissipation grids as each reference heat dissipation grid.

[0046] Arrange each detection point in turn from top to bottom in the target heat dissipation grid of each target device within the scope of each abnormal inspection point, and mark it as each target detection point. Map each target detection point to each reference heat dissipation grid to obtain the corresponding mapping points of each target detection point in each reference heat dissipation grid, and then obtain the distance between each target detection point in the target heat dissipation grid of each target device within the scope of each abnormal inspection point and the corresponding mapping points of each target detection point in each reference heat dissipation grid, and mark it as where q represents the number of the reference heat dissipation grid, q = 1, 2,..., k, and u represents the number of the mapping point, u = 1, 2,..., x.

[0047] Calculate the abnormal index χ of the heat dissipation grid of each target device within the range of each abnormal inspection point gf , where ΔL′ represents the horizontal spacing deviation between the heat dissipation grids of the set reference, and q + 1 represents the (q + 1)-th reference heat dissipation grid

[0048] C2. Based on the heat dissipation fan images corresponding to the ranges of each abnormal inspection point, calculate the abnormal index δ of the heat dissipation fan corresponding to the range of each abnormal inspection point g .

[0049] It should be noted that the specific process of calculating the abnormal index of the heat dissipation fan corresponding to the range of each abnormal inspection point is as follows: Locate the number of dirt areas and the dirt areas corresponding to each dirt area in the heat dissipation fan images corresponding to the ranges of each abnormal inspection point

[0050] Record the number of dirt areas of the heat dissipation fans corresponding to the ranges of each abnormal inspection point as σ g .

[0051] Accumulate the dirt areas corresponding to each dirt area of the heat dissipation fans corresponding to the ranges of each abnormal inspection point to obtain the total dirt area of the heat dissipation fans corresponding to the ranges of each abnormal inspection point, and record it as S g .

[0052] Calculate the abnormal index δ of the heat dissipation fan corresponding to the range of each abnormal inspection point g , where σ′ and S′ respectively represent the number of dirt areas and the dirt area of the set reference

[0053] C3. Based on the ambient temperature of each monitoring point corresponding to the target computer room and the air flow velocity of the heat dissipation outlets, calculate the heat dissipation abnormal index ω of the target computer room

[0054] It should be noted that the specific process of calculating the heat dissipation abnormal index of the target computer room is as follows: Calculate the average value of the ambient temperatures of each monitoring point corresponding to the target computer room to obtain the ambient temperature corresponding to the target computer room, and record it as W

[0055] Record the air flow velocity of the heat dissipation outlets corresponding to the target computer room as v

[0056] Calculate the heat dissipation abnormal index of the target computer room where W″ and v′ respectively represent the ambient temperature and air flow velocity of the set reference

[0057] C4. When the abnormal index of the heat dissipation grid of a target device within the scope of a certain abnormal inspection point is greater than the set reference abnormal index of the heat dissipation grid, the reason for the abnormal heat dissipation of the target device is recorded as the reason of its own heat dissipation grid. If the abnormal index of the heat dissipation fan corresponding to the scope of the abnormal inspection point is greater than the set reference abnormal index of the heat dissipation fan, the reason for the abnormal heat dissipation of the target device is recorded as the reason of the heat dissipation fan. If the abnormal index of the heat dissipation of the target computer room is greater than the set reference abnormal index of the heat dissipation, the reason for the abnormal heat dissipation of the target device is recorded as the reason of the computer room heat dissipation.

[0058] In a specific embodiment of the present invention, the specific process of providing an operation and maintenance solution for abnormal heat dissipation is as follows: D1. If the reason for the abnormal heat dissipation of a target device is the reason of its own heat dissipation grid, an abnormal feedback of the heat dissipation grid is performed.

[0059] D2. If the reason for the abnormal heat dissipation of a target device is the reason of the heat dissipation fan, an abnormal feedback of the heat dissipation fan is performed.

[0060] D3. If the reason for the abnormal heat dissipation of a target device is the reason of the computer room heat dissipation, the rotation speed of the heat dissipation fan corresponding to the target computer room is recorded as v 目 , the required compensation rotation speed of the heat dissipation fan corresponding to the unit computer room heat dissipation deviation is extracted from the database and recorded as v0, and the appropriate rotation speed v of the heat dissipation fan corresponding to the target computer room is set 适 , and the rotation speed of the heat dissipation fan corresponding to the target computer room is adjusted to the appropriate rotation speed, where represents the set reference abnormal index of the heat dissipation.

[0061] In the embodiment of the present invention, by performing a secondary inspection on each abnormal inspection point in the target computer room, it is judged whether there are devices with continuously abnormal temperatures within the scope of each abnormal inspection point. If so, the reason for the abnormal heat dissipation is confirmed, and an operation and maintenance solution is provided for the abnormal heat dissipation, which improves the timeliness of discovering and handling the abnormal device temperature, reduces the risk of device damage, and thus reduces the operation and maintenance cost of the enterprise.

[0062] S5. Fault device judgment: Collect the colors of the fault indicator lights of each device within the scope of each inspection point in the inspection route, and judge whether there are fault devices within the scope of each inspection point. If so, a fault device feedback is performed.

[0063] It should be noted that the colors of the fault indicator lights of each device within the scope of each inspection point in the inspection route are collected by a camera installed on the target inspection robot to capture the corresponding fault indicator light images, and the colors of the fault indicator lights are identified from the fault indicator light images.

[0064] In a specific embodiment of the present invention, the specific method for determining whether there are faulty devices within the scope of each inspection point is as follows: Compare the colors of the fault indicator lights of each device within the scope of each inspection point in the inspection route with the fault colors displayed by the indicator lights stored in the database. If the color of the fault indicator light of a certain device is the same as the fault color displayed by the indicator light, then mark this device as a faulty device.

[0065] S6. Inspection route optimization: Count the number of temperature-abnormal devices, the number of target devices, and the number of faulty devices that appear within the scope of each inspection point in the inspection route corresponding to each inspection day in the current inspection cycle, and optimize the preset inspection route for the target computer room.

[0066] In a specific embodiment of the present invention, the specific process of optimizing the preset inspection route for the target computer room is as follows: E1. Based on the number of temperature-abnormal devices, the number of target devices, and the number of faulty devices that appear within the scope of each inspection point in the inspection route corresponding to each inspection day in the current inspection cycle, calculate the frequency of abnormal occurrences of each inspection point in the inspection route during the current inspection cycle.

[0067] In a specific embodiment of the present invention, the specific process of calculating the frequency of abnormal occurrences of each inspection point in the inspection route during the current inspection cycle is as follows: F1. Denote the number of temperature-abnormal devices, the number of target devices, and the number of faulty devices that appear within the scope of each inspection point in the inspection route corresponding to each inspection day in the current inspection cycle as and where r represents the number of the inspection day, and r = 1, 2,..., w.

[0068] F2. Calculate the frequency of abnormal occurrences of each inspection point in the inspection route during the current inspection cycle where K1, K2, and K3 respectively represent the set reference ratios of temperature-abnormal devices, target devices, and faulty devices, m represents the number of devices, z represents the number of target devices, and w represents the number of inspection days.

[0069] E2. Sort the frequencies of abnormal occurrences of each inspection point in the inspection route during the current inspection cycle from largest to smallest to obtain the order of frequencies of abnormal occurrences of the inspection points, and connect the inspection points in sequence according to the order of frequencies of abnormal occurrences to obtain the inspection route corresponding to the target computer room in the next inspection cycle.

[0070] In an embodiment of the present invention, by counting the number of temperature - abnormal devices, the number of target devices, and the number of faulty devices that appear within the scope of each inspection point in the inspection route corresponding to each inspection day during the current inspection cycle, the preset inspection route of the target computer room is optimized, avoiding the inspection robot from ignoring key areas and repeatedly inspecting normal areas, saving valuable inspection time and resources, meeting the new inspection requirements in the computer room, and improving the overall operation and maintenance level of the computer room.

[0071] Referring to Figure 2 As shown, in a second aspect of the present invention, a computer - room operation and maintenance inspection system based on machine learning is provided, including: a temperature - abnormal device judgment module, a cooling - fan speed adjustment module, a continuously - temperature - abnormal device judgment module, a heat - dissipation - abnormal operation and maintenance plan providing module, a database, a faulty device judgment module, and an inspection - route optimization module.

[0072] The temperature - abnormal device judgment module is connected to the cooling - fan speed adjustment module, the continuously - temperature - abnormal device judgment module is connected to the heat - dissipation - abnormal operation and maintenance plan providing module, both the cooling - fan speed adjustment module and the heat - dissipation - abnormal operation and maintenance plan providing module are connected to the database, and the temperature - abnormal device judgment module, the continuously - temperature - abnormal device judgment module, and the faulty device judgment module are all connected to the inspection - route optimization module.

[0073] The temperature - abnormal device judgment module is used to input the preset inspection route of the target computer room into the background of the target inspection robot. When the target inspection robot conducts inspections according to the preset inspection route, it collects the thermal images of each device within the scope of each inspection point in the inspection route, and judges whether there are temperature - abnormal devices within the scope of each inspection point in the inspection route.

[0074] The cooling - fan speed adjustment module is used to confirm each abnormal inspection point in the inspection route, extract the current speed of the cooling fan corresponding to the scope of each abnormal inspection point, and adjust the speed of the cooling fan corresponding to the scope of each abnormal inspection point.

[0075] The continuously - temperature - abnormal device judgment module is used to, after the target inspection robot finishes the first inspection, send the positions of each abnormal inspection point to the background of the target inspection robot again for a second inspection, collect the thermal images of each temperature - abnormal device within the scope of each abnormal inspection point during the second inspection, and judge whether there are continuously - temperature - abnormal devices within the scope of each abnormal inspection point.

[0076] The heat dissipation anomaly operation and maintenance solution providing module is used to mark the devices with continuously abnormal temperatures as target devices, collect the heat dissipation grid images of each target device within the scope of each abnormal inspection point and the heat dissipation fan images corresponding to the scope of each abnormal inspection point. At the same time, collect the ambient temperature of each monitoring point corresponding to the target computer room, the air circulation speed of the heat dissipation outlet, and the rotation speed of the heat dissipation fan, confirm the reasons for the heat dissipation anomalies of each target device, and provide operation and maintenance solutions for the heat dissipation anomalies.

[0077] The database is used to store the appropriate heat dissipation fan rotation speeds corresponding to each temperature anomaly index, and store the compensation rotation speeds required for the heat dissipation fans corresponding to the heat dissipation anomalies of the unit computer room.

[0078] Table 1 shows the data sources in the database of this embodiment.

[0079]

[0080] The faulty device judgment module is used to collect the colors of the fault indicator lights of each device within the scope of each inspection point in the inspection route, and judge whether there are faulty devices within the scope of each inspection point.

[0081] The inspection route optimization module is used to count the number of devices with temperature anomalies, the number of target devices, and the number of faulty devices that appear within the scope of each inspection point in the inspection route corresponding to each inspection day during the current inspection period, and optimize the preset inspection route of the target computer room.

[0082] The above content is only an example and explanation of the concept of the present invention. Those skilled in the art of this technology can make various modifications or supplements to the described specific embodiments or use similar methods for substitution, as long as they do not deviate from the concept of the invention or exceed the scope defined by the present invention, they should fall within the protection scope of the present invention.

Claims

1. A machine room operation and maintenance inspection method based on machine learning, characterized in that, Including: S1. Judgment of temperature-abnormal devices: Input the preset inspection route of the target computer room into the background of the target inspection robot. When the target inspection robot conducts inspections according to the preset inspection route, collect the thermal images of each device within the scope of each inspection point in the inspection route, and then judge whether there are temperature-abnormal devices within the scope of each inspection point in the inspection route. If there are, execute step S2; if not, it indicates that the device heat dissipation is normal. S2. Adjustment of the rotation speed of the cooling fan: Confirm each abnormal inspection point in the inspection route, extract the current rotation speed of the cooling fan corresponding to the scope of each abnormal inspection point, adjust the rotation speed of the cooling fan corresponding to the scope of each abnormal inspection point, and execute step S3. S3. Judgment of temperature-persistently abnormal devices: After the first inspection of the target inspection robot is completed, send the positions of each abnormal inspection point to the background of the target inspection robot again for a second inspection, collect the thermal images of each temperature-abnormal device within the scope of each abnormal inspection point during the second inspection, and judge whether there are temperature-persistently abnormal devices within the scope of each abnormal inspection point. If there are, execute step S4; if not, it indicates that the device temperature has returned to normal. S4. Provide an operation and maintenance plan for abnormal heat dissipation: Record the temperature-persistently abnormal devices as target devices, collect the heat dissipation grid images of each target device within the scope of each abnormal inspection point and the cooling fan images corresponding to the scope of each abnormal inspection point, and at the same time collect the ambient temperature of each monitoring point corresponding to the target computer room, the air flow speed of the heat dissipation port, and the rotation speed of the heat dissipation fan, confirm the reasons for the abnormal heat dissipation of each target device, and provide an operation and maintenance plan for the abnormal heat dissipation. S5. Judgment of faulty devices: Collect the colors of the fault indicator lights of each device within the scope of each inspection point in the inspection route, and judge whether there are faulty devices within the scope of each inspection point. If there are, give feedback on the faulty devices. S6. Optimization of the inspection route: Count the number of temperature-abnormal devices, target devices, and faulty devices that appear within the scope of each inspection point in the inspection route corresponding to each inspection day during the current inspection cycle, and optimize the preset inspection route of the target computer room.

2. The method for machine room operation and maintenance inspection based on machine learning according to claim 1, characterized in that: The specific process of judging whether there are temperature-abnormal devices within the scope of each inspection point in the inspection route is as follows: A1. Locate the temperature values corresponding to each temperature distribution region from the thermal images of each device within the scope of each inspection point on the inspection route, calculate their average values, and obtain the average temperature of each device within the scope of each inspection point, denoted as where i represents the number of the inspection point, i = 1, 2,..., n, and j represents the number of the device, j = 1, 2,..., m; A2. Extract the maximum value from the temperature values corresponding to each temperature distribution region of each device within the scope of each inspection point on the inspection route, and denote it as A3. Calculate the temperature anomaly index β of each device within the scope of each inspection point ij , where W′ represents the temperature of the device set as a reference, and e represents the natural constant; A4. Compare the temperature-abnormal index of each device within the scope of each inspection point with the set reference temperature-abnormal index. If the temperature-abnormal index of a certain device within the scope of a certain inspection point is greater than or equal to the set reference temperature-abnormal index, record this device as a temperature-abnormal device, and thus judge whether there are temperature-abnormal devices within the scope of each inspection point in the inspection route.

3. The method for machine room operation and maintenance inspection based on machine learning according to claim 2, wherein: The specific process of adjusting the rotation speed of the cooling fan corresponding to the scope of each abnormal inspection point is as follows: B1. Denote the current rotational speed of the cooling fan corresponding to the range where each abnormal inspection point is located as v g , where g represents the number of the abnormal inspection point, and g = 1, 2,..., p; B2. Extract the maximum value from the temperature anomaly indices of each temperature-anomalous device within the scope of each abnormal inspection point, and compare it with the appropriate heat dissipation fan speed corresponding to each temperature anomaly index stored in the database to obtain the appropriate heat dissipation fan speed corresponding to the scope of each abnormal inspection point, and record it as B3. Set the increased rotational speed value required for the cooling fan corresponding to the range to which each abnormal inspection point belongs 4. The method for machine room operation and maintenance inspection based on machine learning according to claim 2, characterized in that: The method of judging whether there are temperature-persistently abnormal devices within the scope of each abnormal inspection point is: Based on the thermal images of each temperature-abnormal device within the scope of each abnormal inspection point during the second inspection, judge whether there are temperature-persistently abnormal devices within the scope of each abnormal inspection point in the same way as the method of judging whether there are temperature-abnormal devices within the scope of each inspection point in the inspection route.

5. The method for machine room operation and maintenance inspection based on machine learning according to claim 3, characterized in that: The specific process of confirming the reasons for abnormal heat dissipation of each target device is as follows: C1. Calculate the heat dissipation grid anomaly index χ of each target device within the range of each abnormal inspection point based on the heat dissipation grid images of each target device within the range of each abnormal inspection point gf , where f represents the number of the target device, f = 1, 2,..., z; C2. Calculate the abnormal index δ of the cooling fan corresponding to the range where each abnormal inspection point is located based on the image of the cooling fan corresponding to the range where each abnormal inspection point is located g ; C3. Calculate the heat dissipation anomaly index of the target computer room based on the ambient temperature of each monitoring point corresponding to the target computer room and the air flow velocity at the heat dissipation outlet C4. When the abnormal index of the heat dissipation grid of a target device within the range of a certain abnormal inspection point is greater than the set reference abnormal index of the heat dissipation grid, the reason for the abnormal heat dissipation of the target device is recorded as its own heat dissipation grid reason. If the abnormal index of the heat dissipation fan corresponding to the range of the abnormal inspection point is greater than the set reference abnormal index of the heat dissipation fan, the reason for the abnormal heat dissipation of the target device is recorded as the heat dissipation fan reason. If the abnormal index of the heat dissipation of the target computer room is greater than the set reference abnormal index of the heat dissipation, the reason for the abnormal heat dissipation of the target device is recorded as the computer room heat dissipation reason.

6. The method for machine room operation and maintenance inspection based on machine learning according to claim 5, wherein: The specific process of providing an operation and maintenance plan for abnormal heat dissipation is as follows: D1. If the reason for the abnormal heat dissipation of a target device is its own heat dissipation grid reason, an abnormal feedback of the heat dissipation grid is performed. D2. If the reason for the abnormal heat dissipation of a target device is the heat dissipation fan reason, an abnormal feedback of the heat dissipation fan is performed. D3. If the reason for the abnormal heat dissipation of a target device is the heat dissipation in the computer room, record the rotation speed of the cooling fan corresponding to the target computer room as v 目 , extract the required compensation rotation speed of the cooling fan corresponding to the unit computer room heat dissipation abnormality deviation from the database, and record it as v0, and set the appropriate rotation speed v of the cooling fan corresponding to the target computer room 适 , and adjust the rotation speed of the cooling fan corresponding to the target computer room to the appropriate rotation speed, where represents the heat dissipation abnormality index set for reference.

7. The machine room operation and maintenance inspection method based on machine learning according to claim 5, characterized in that: The specific method for judging whether there are faulty devices within the range of each inspection point is as follows: Compare the color of the fault indicator of each device within the range of each inspection point in the inspection route with the fault color displayed by the indicator stored in the database. If the color of the fault indicator of a device is the same as the fault color displayed by the indicator, the device is recorded as a faulty device.

8. A machine room operation and maintenance inspection method based on machine learning according to claim 7, characterized in that: The specific process of optimizing the preset inspection route of the target computer room is as follows: E1. Calculate the anomaly occurrence frequency of each inspection point in the inspection route during the current inspection cycle based on the number of temperature anomaly devices, target devices, and faulty devices within the scope of each inspection point in the inspection route corresponding to each inspection day during the current inspection cycle E2. Sort the abnormal occurrence frequencies of each inspection point in the inspection route during the current inspection cycle from large to small to obtain the order of the abnormal occurrence frequencies of the inspection points, and connect the inspection points in sequence according to the order of the abnormal occurrence frequencies to obtain the inspection route corresponding to the target computer room in the next inspection cycle.

9. The method for machine room operation and maintenance inspection based on machine learning according to claim 8, wherein: The specific process of calculating the abnormal occurrence frequencies of each inspection point in the inspection route during the current inspection cycle is as follows: F1. Denote the number of temperature - abnormal devices, the number of target devices, and the number of faulty devices that appear within the scope of each inspection point in the inspection route corresponding to each inspection day within the current inspection cycle as and where r represents the serial number of the inspection day, and r = 1, 2,..., w; F2. Calculate the frequency of anomalies occurring at each inspection point in the inspection route during the current inspection cycle Among them, K1, K2, and K3 respectively represent the proportion of temperature anomaly devices, target devices, and faulty devices set for reference. m represents the number of devices, z represents the number of target devices, and w represents the number of inspection days.

10. A machine room operation and maintenance inspection system based on machine learning, characterized in that, Including: A temperature abnormal device judgment module, which is used to input the preset inspection route of the target computer room into the background of the target inspection robot. When the target inspection robot conducts inspections according to the preset inspection route, collect the thermal images of each device within the range of each inspection point in the inspection route, and judge whether there are temperature abnormal devices within the range of each inspection point in the inspection route. A heat dissipation fan speed adjustment module, which is used to confirm each abnormal inspection point in the inspection route, extract the current speed of the heat dissipation fan corresponding to the range of each abnormal inspection point, and adjust the speed of the heat dissipation fan corresponding to the range of each abnormal inspection point. A temperature continuously abnormal device judgment module, which is used to, after the target inspection robot finishes the first inspection, send the positions of each abnormal inspection point to the background of the target inspection robot again for a second inspection, collect the thermal images of each temperature abnormal device within the range of each abnormal inspection point during the second inspection, and judge whether there are temperature continuously abnormal devices within the range of each abnormal inspection point. The heat dissipation anomaly operation and maintenance solution providing module is used to record the devices with continuous temperature anomalies as target devices, collect the heat dissipation grid images of each target device within the scope of each abnormal inspection point and the heat dissipation fan images corresponding to the scope of each abnormal inspection point, and at the same time collect the ambient temperature of each monitoring point corresponding to the target computer room, the air flow velocity of the heat dissipation outlet, and the rotation speed of the heat dissipation fan, confirm the reasons for the heat dissipation anomalies of each target device, and provide operation and maintenance solutions for heat dissipation anomalies; The database is used to store the appropriate heat dissipation fan rotation speeds corresponding to each temperature anomaly index, and store the compensation rotation speeds required for the heat dissipation fans corresponding to the heat dissipation anomalies of the unit computer room; The faulty device judgment module is used to collect the colors of the fault indicator lights of each device within the scope of each inspection point in the inspection route, and judge whether there are faulty devices within the scope of each inspection point; The inspection route optimization module is used to count the number of temperature anomaly devices, the number of target devices, and the number of faulty devices that appear within the scope of each inspection point in the inspection route corresponding to each inspection day in the current inspection cycle, and optimize the preset inspection route of the target computer room.

Images