A programmable logic control method and system for data center equipment room

Abstract

The application discloses a programmable logic control method and system for data center equipment, relates to the field of programmable logic control of data center equipment, and is used for improving the operation stability of the data center room and comprises the following steps: acquiring temperature values of temperature sensors in the room, relative positions of the temperature sensors and cabinets, thermal response characteristics of the temperature sensors and power change rates of the cabinets; evaluating the credibility scores of the temperature sensors according to the relative positions, the thermal response characteristics and the power change rates; compensating temperature values of abnormal temperature sensors according to temperature rising accelerations of the abnormal temperature sensors, temperature values of normal temperature sensors in adjacent areas and the power change rates; generating a three-dimensional temperature field distribution of the room according to the temperature values of the compensated abnormal temperature sensors and the temperature values of the normal temperature sensors; and generating and issuing adjusting instructions matched with the action range of cooling end equipment according to the three-dimensional temperature field distribution.

Description

Technical Field

[0001] This invention relates to the field of programmable logic control for data center equipment, and more particularly to a programmable logic control method and system for data center equipment. Background Technology

[0002] In the daily operation of data center server rooms, ambient temperature sensors typically collect spatial temperature field data. Control logic periodically acquires this data to determine if localized hotspots have appeared, and then sends adjustment commands to the cooling terminals to maintain stable server room temperatures. However, in actual deployments, temperature sensors are usually installed above the rack aisles or in fixed locations, with a certain spatial distance between them and the rack air vents. Furthermore, the sensors themselves have thermal capacitance characteristics, leading to a response lag. In addition, the sampling times of multiple sensor nodes are often not completely synchronized.

[0003] When a high-density server rack experiences a sudden surge in computing load, causing its internal servers to rapidly reach full capacity, the temperature at its exhaust vents will rise dramatically within a very short time, creating a localized hotspot. However, due to spatial offset and thermal inertia, the temperature rise curve collected by the nearest sensor is extremely flat and the peak value is low. Furthermore, due to the misalignment of sampling times by the various sensors, the control logic, when piecing together the current spatial temperature field, cannot accurately reconstruct the true transient temperature distribution of this localized hotspot area.

[0004] Existing conventional control methods typically rely directly on the currently acquired temperature vector, treating it as the true environmental state, lacking compensation mechanisms for spatial sampling blind spots and perception lag. Therefore, in data center server rooms, when high-density server racks experience a sudden surge in exhaust temperature due to unexpected computing loads, the spatial temperature field state sequence acquired by the control logic lags significantly behind the evolution of the actual local hotspot due to limitations in the spatial coverage of ambient temperature sensors, their own thermal response lag, and asynchronous sampling times of multiple sensors. This results in a failure to trigger timely enhanced cooling commands for the hotspot, leading to localized overheating spreading to adjacent areas and causing server frequency reduction or shutdown. Summary of the Invention

[0005] This application discloses a programmable logic control method and system for data center equipment, aiming to solve the technical problem that in existing data center equipment, the ambient temperature sensor is limited by spatial coverage blind spots, its own thermal response lags, and the sampling time of multiple sensors is not synchronized. This causes the spatial temperature field state sequence obtained by the control logic to lag far behind the evolution of the actual local hot spots, and fails to trigger the enhanced cooling command in time, which in turn causes local overheating to spread and causes the server to reduce frequency or shut down.

[0006] In a first aspect, this application discloses a programmable logic control method for data center equipment, comprising the following steps: Acquire the temperature values ​​of each temperature sensor in the computer room, the relative position of each temperature sensor to the cabinet, the thermal response characteristics of the temperature sensors, and the power change rate of the cabinet; The reliability score of each temperature sensor is evaluated based on its relative position, thermal response characteristics, and power change rate. The temperature value of the abnormal temperature sensor is compensated based on the temperature rise acceleration of the abnormal temperature sensor, the temperature value of the normal temperature sensor in the adjacent area, and the power change rate. The abnormal temperature sensor is the temperature sensor with a confidence score less than or equal to the scoring threshold; the normal temperature sensor is the temperature sensor with a confidence score greater than the scoring threshold. Based on the temperature values ​​of the compensated abnormal temperature sensor and the temperature values ​​of the normal temperature sensor, a three-dimensional temperature field distribution of the computer room is generated. Based on the three-dimensional temperature field distribution, adjustment instructions are generated and issued that match the effective range of the cooling terminal equipment. The adjustment instructions include the adjustment object, adjustment priority, and adjustment intensity.

[0007] Optionally, the reliability score of each temperature sensor is evaluated based on its relative position, thermal response characteristics, and power change rate, including: Based on their relative positions, calculate the Euclidean spatial distance between each temperature sensor and the center point of the air outlet of the cabinet it is responsible for monitoring; Based on Euclidean spatial distance and exponential decay function, the spatial representativeness score of the temperature sensor is calculated; the spatial representativeness score is inversely proportional to the Euclidean spatial distance. The reliability score of each temperature sensor is evaluated by combining spatial representativeness score, thermal response characteristics, and power change rate.

[0008] Optionally, the temperature value of the abnormal temperature sensor is compensated based on the temperature rise acceleration of the abnormal temperature sensor, the temperature value of the normal temperature sensor in the adjacent area, and the power change rate, including: The spatial temperature gradient of adjacent areas is determined based on the temperature values ​​of normal temperature sensors in adjacent areas. The dynamic compensation item is determined based on the compensation formula, which is: F = 1 / 2 × V × T 2 +A+B×W1×W2; Where F is the dynamic compensation term, V is the temperature rise acceleration of the abnormal temperature sensor, T is the preset step time, A is the projection increment of the spatial temperature gradient in the corresponding direction, B is the power change rate, W1 is the preset thermoelectric conversion coefficient, and W2 is the thermal inertia compensation gain. The temperature value of the abnormal temperature sensor is compensated based on the dynamic compensation term.

[0009] Optionally, based on the three-dimensional temperature field distribution, adjustment commands matching the effective range of the cooling terminal equipment are generated and issued, including: Based on the three-dimensional temperature field distribution, the location of hotspots is determined; Analyze the temperature field change sequence of hotspots over multiple consecutive periods to determine the evolution stage of the hotspots; the evolution stages include the initial stage, the diffusion stage, and the overheating stage. Based on the location and evolution stage of the hotspot, adjustment commands are generated and issued that match the effective range of the cooling terminal equipment.

[0010] Optionally, the temperature field change sequence of the hotspot over multiple consecutive periods can be analyzed to determine the evolutionary stage of the hotspot, including: If the first preset condition is met, the evolution stage of the hotspot is determined to be the overheating stage; the first preset condition is that the temperature within a preset range from the hotspot is greater than a preset temperature threshold. If the second preset condition is met, the evolution stage of the hotspot is determined to be the initial stage; the second preset condition is that the temperature of some areas within a preset range from the hotspot is less than or equal to a preset temperature threshold, and the temperature rise rate of the hotspot is greater than the speed threshold, and the temperature rise rate within the preset range from the hotspot is less than the speed threshold. If the first and second preset conditions are not met, the hotspot is determined to be in the diffusion stage of its evolution.

[0011] Optionally, based on the location and evolution stage of the hotspot, adjustment commands matching the effective range of the cooling terminal equipment are generated and issued, including: If the hot spot is in the initial stage of evolution, increase the fan speed of the cooling terminal device closest to the hot spot. If the hot spot is in the diffusion stage of its evolution, then reduce the air supply temperature of the cooling terminal equipment within a preset range from the hot spot. If the hot spot is in an overheating stage of evolution, then reduce the supply air temperature of all cooling terminal equipment.

[0012] Optionally, the method also includes: After the adjustment command is issued, temperature data in the computer room is continuously acquired. Obtain the parameters of the issued adjustment command, the spatial distance from the cooling terminal device to the temperature sensor, and the current air supply speed; Adjust the control commands based on temperature data, issued control command parameters, spatial distance, and current air supply speed.

[0013] Optionally, the adjustment command can be adjusted based on temperature data, issued adjustment command parameters, spatial distance, and current supply air velocity, including: Based on the temperature data, determine whether the cooling intervention of the adjustment command has taken effect on the hot spot; If it has taken effect, the theoretical cooling effect of the adjustment command on the hot spot is calculated based on the adjustment command parameters, spatial distance and current air supply speed. Subtracting the theoretical cooling rate from the temperature in the three-dimensional temperature field distribution yields the controlled expected temperature. The temperature data was compared with the controlled expected temperature to obtain the reconstruction deviation; Adjust the control instructions based on the reconstruction deviation.

[0014] Optionally, based on the three-dimensional temperature field distribution, the location of hotspots can be determined, including: Based on the three-dimensional temperature field distribution, the coordinates of the locations where the temperature value is greater than the identification threshold and the spatial temperature gradient is greater than the gradient threshold are determined. Based on location coordinates, locate the hotspot's position.

[0015] Secondly, this application also discloses a programmable logic control system for data center equipment, the system comprising: The information acquisition module is used to acquire the temperature values ​​of each temperature sensor in the computer room, the relative position of each temperature sensor to the cabinet, the thermal response characteristics of the temperature sensors, and the power change rate of the cabinet. The scoring module is used to evaluate the reliability score of each temperature sensor based on its relative position, thermal response characteristics, and power change rate. The compensation module is used to compensate the temperature value of the abnormal temperature sensor based on the temperature rise acceleration of the abnormal temperature sensor, the temperature value of the normal temperature sensor in the adjacent area, and the power change rate. The abnormal temperature sensor is a temperature sensor with a confidence score less than or equal to the scoring threshold; the normal temperature sensor is a temperature sensor with a confidence score greater than the scoring threshold. The temperature field generation module is used to generate a three-dimensional temperature field distribution in the computer room based on the temperature values ​​of the compensated abnormal temperature sensors and the temperature values ​​of the normal temperature sensors. The instruction generation and issuance module is used to generate and issue adjustment instructions that match the range of action of the cooling terminal equipment based on the three-dimensional temperature field distribution. The adjustment instructions include the adjustment object, adjustment priority, and adjustment intensity.

[0016] Beneficial effects This application provides a programmable logic control method for data center equipment. It acquires the temperature values, relative positions, thermal response characteristics, and power change rates of various temperature sensors within the data center, and evaluates the reliability score of each temperature sensor based on this information. For abnormal temperature sensors with low reliability scores, this application dynamically compensates for them based on their temperature rise acceleration, the temperature values ​​of normal temperature sensors in adjacent areas, and the power change rate. This effectively solves the distortion problem of sensor data caused by spatial offset, thermal inertia, and asynchronous sampling in existing technologies. Furthermore, this application uses the compensated temperature data to generate a precise three-dimensional temperature field distribution within the data center, and generates adjustment commands matching the effective range of the cooling terminal equipment based on this distribution. These commands include the adjustment target, adjustment priority, and adjustment intensity.

[0017] Through the above technical solution, this application overcomes the shortcomings of conventional control methods in the prior art, which directly rely on the acquired temperature vector and lack compensation mechanisms for spatial sampling blind spots and perception lag. When the outlet temperature of a high-density server rack rises sharply due to a sudden increase in computing load, this application can accurately restore the true transient temperature distribution of the local hotspot area by assessing and compensating for the reliability of sensor data. This avoids the problem that the spatial temperature field state sequence obtained by the control logic lags significantly behind the actual evolution of the local hotspot. Therefore, this application can promptly trigger enhanced cooling commands for the hotspot, effectively preventing local overheating from spreading to adjacent areas, avoiding server frequency reduction or shutdown, and significantly improving the operational stability and energy efficiency of the data center. Attached Figure Description

[0018] Figure 1 This is a schematic flowchart of a programmable logic control method for data center equipment provided in an embodiment of the present invention; Figure 2 This is a schematic flowchart of another programmable logic control method for data center equipment provided in an embodiment of the present invention; Figure 3 This is a schematic diagram of a programmable logic control system for a data center equipment according to an embodiment of the present invention. Detailed Implementation

[0019] The technical solutions of this application will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are merely some embodiments of this application, and not all embodiments. The components of this application described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely to illustrate selected embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0020] It should be noted that similar reference numerals and letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. Furthermore, in the description of this application, terms such as "first," "second," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0021] To better understand the technical solution proposed in this application, some key terms involved will be explained first.

[0022] Temperature sensors within a data center refer to various temperature-sensing elements, such as thermistors, thermocouples, or infrared sensors, deployed inside the data center to monitor ambient temperature in real time. These sensors are typically installed at air inlets, outlets, passageways, or specific areas within server racks to obtain temperature data from different locations.

[0023] The relative position of the cabinet refers to the spatial coordinate relationship between the temperature sensor and the cabinet it monitors (especially the cabinet air outlet), which is crucial for assessing the spatial representativeness of the sensor data.

[0024] The thermal response characteristics of a temperature sensor refer to the speed and accuracy of its response to temperature changes, and are usually determined by its thermal inertia. Sensors with poor thermal response characteristics will exhibit lag in their readings when the temperature changes rapidly.

[0025] The power change rate of a server rack refers to the rate at which the power consumption of the equipment inside the rack changes over time. It directly reflects the changing trend of the rack's heat generation and is an important indicator for predicting the formation and evolution of local hotspots.

[0026] The reliability score is a quantitative assessment of the reliability of each temperature sensor's data, taking into account factors such as the distance between the sensor and the heat source, response speed, and changes in the heat source's power. A higher score indicates more reliable sensor data.

[0027] An abnormal temperature sensor is a sensor whose reliability score is lower than a preset threshold. Its data may have a large deviation or lag, and compensation is required.

[0028] A normal temperature sensor is one whose reliability score is higher than a preset threshold, and its data is considered relatively reliable.

[0029] Temperature rise acceleration refers to the rate at which the temperature rises within a short period of time by an abnormal temperature sensor, reflecting the urgency of the formation of local hot spots.

[0030] A normal temperature sensor in the adjacent area refers to a sensor located near an abnormal temperature sensor, whose data is considered reliable. Data from these sensors can be used to infer the true temperature of the abnormal area.

[0031] Three-dimensional temperature field distribution is a three-dimensional visualization model of the temperature inside the computer room. Through interpolation, fitting and other algorithms, discrete sensor data is transformed into a continuous temperature field, which can intuitively show the location, range and intensity of hot spots.

[0032] Cooling terminal equipment refers to equipment used to provide cooling capacity in data center computer rooms, such as computer room air conditioners (CRAC), computer room air processors (CRAH), fan walls, cold plates, or liquid cooling units.

[0033] Adjustment commands are operational instructions issued by the control system to the cooling terminal equipment to adjust its operating parameters in response to changes in the machine room temperature. Adjustment commands typically include the object of adjustment (such as a specific cooling device), the adjustment priority (such as the urgency level), and the adjustment intensity (such as fan speed, supply air temperature, etc.).

[0034] The following specific embodiments will provide a detailed description and explanation of a programmable logic control method for data center equipment provided in this application.

[0035] Reference Figure 1 This invention provides a programmable logic control method for data center equipment, comprising the following steps: S1 acquires the temperature values ​​of each temperature sensor in the computer room, the relative position of each temperature sensor to the cabinet, the thermal response characteristics of the temperature sensors, and the power change rate of the cabinet.

[0036] Temperature values ​​can be directly acquired through sensors; for example, current temperature data can be read from various temperature sensors using a periodic polling method. Relative location information can be determined through measurement or modeling during the initial deployment of the data center and stored in a database. Thermal response characteristics can be obtained through sensor factory parameters or actual testing; for example, the sensor's response time constant can be measured through a step response experiment. The power change rate of the server rack can be obtained by monitoring the power consumption data of the servers inside the rack and calculating its change over a period of time. For example, monitoring data from the rack's PDU (Power Distribution Unit) can be integrated to obtain the rack's instantaneous power in real time and calculate its rate of change.

[0037] S2. Evaluate the reliability score of each temperature sensor based on its relative position, thermal response characteristics, and power change rate.

[0038] For example, a scoring model can be set up where the closer the sensor is to the air vent of the cabinet, the higher its spatial representativeness score; the shorter the sensor's thermal response time, the higher its thermal response score; and the greater the cabinet power change rate, the higher the dynamic response score if the sensor can promptly reflect this change. Finally, these scores are weighted and averaged to obtain the sensor's overall reliability score. Alternatively, a machine learning model can be used, trained on historical data, to learn the weights of different factors influencing sensor reliability, thereby more accurately assessing the reliability score.

[0039] S3. Based on the temperature rise acceleration of the abnormal temperature sensor, the temperature value of the normal temperature sensor in the adjacent area, and the power change rate, the temperature value of the abnormal temperature sensor is compensated.

[0040] Among them, abnormal temperature sensors are temperature sensors with a confidence score less than or equal to the scoring threshold; normal temperature sensors are temperature sensors with a confidence score greater than the scoring threshold.

[0041] For example, when the reliability score of a sensor falls below a preset threshold, the system marks it as abnormal. At this point, the system acquires the temperature rise acceleration of the abnormal sensor, for example, by calculating its temperature change rate over the most recent sampling periods. Simultaneously, the system acquires the temperature values ​​of normal temperature sensors in the adjacent area; the data from these normal sensors are considered reliable. Furthermore, the system acquires the power change rate of the server rack. Based on this information, a compensation model can be constructed. For example, using algorithms such as Kalman filtering or neural networks, combined with the historical trend of the abnormal sensor, the actual temperature of the adjacent area, and the power change of the server rack, the temperature value of the abnormal sensor can be dynamically corrected to compensate for errors caused by spatial offset, thermal inertia, or sampling asynchrony.

[0042] S4. Based on the temperature values ​​of the compensated abnormal temperature sensor and the temperature values ​​of the normal temperature sensor, generate the three-dimensional temperature field distribution of the computer room.

[0043] For example, interpolation algorithms (such as Kriging interpolation and inverse distance weighted interpolation) or computational fluid dynamics (CFD) simulations based on physical models can be used to transform discrete sensor data (including compensated abnormal and normal sensor data) into a continuous three-dimensional temperature field. This can provide temperature information at any location within the computer room, thus reflecting the temperature conditions of the computer room more comprehensively and accurately.

[0044] S5. Based on the three-dimensional temperature field distribution, generate and issue adjustment commands that match the operating range of the cooling terminal equipment.

[0045] The adjustment command includes the adjustment target, adjustment priority, and adjustment intensity.

[0046] For example, the system can identify hotspot areas based on the three-dimensional temperature field distribution, and intelligently select the most suitable cooling terminal equipment for intervention based on the location, intensity, and evolution trend of the hotspots. The adjustment command can specify a single cooling terminal equipment or a group of cooling terminal equipment (the adjustment object), set the urgency level of its adjustment (adjustment priority), and provide specific adjustment parameters (adjustment intensity), such as increasing the fan speed or decreasing the supply air temperature.

[0047] After acquiring the raw data, this application does not blindly trust it but introduces a credibility assessment mechanism. By comprehensively analyzing the spatial representativeness of the sensors (relative position to the rack), time response capability (thermal response characteristics), and dynamic changes in the heat source (rack power change rate), the credibility score of the data from each temperature sensor is calculated. This step effectively identifies abnormal sensors that may have biases or lags, avoiding misjudgments caused by unreliable data.

[0048] For temperature sensors identified as abnormal, the system activates an intelligent compensation mechanism. This mechanism relies not only on the abnormal sensor's own data but also combines the abnormal sensor's temperature rise acceleration, real-time data from normal temperature sensors in adjacent areas, and the power change rate of the cabinet. Through this multi-source information fusion, the system dynamically compensates for the abnormal sensor's temperature value, thereby restoring, to a certain extent, the true temperature of local hotspots and solving the problems of sensor data lag and inaccuracy in traditional methods. The compensated abnormal sensor data, together with the normal sensor data, constitutes a more accurate and comprehensive temperature dataset.

[0049] Ultimately, based on the generated three-dimensional temperature field distribution, the system intelligently generates and issues adjustment commands that match the effective range of the cooling terminal equipment. This means that the system no longer simply performs uniform adjustments to the entire computer room, but can precisely control the corresponding cooling terminal equipment for specific hotspot areas. For example, it can increase the fan speed of the cooling terminal equipment closest to the hotspot, or reduce the supply air temperature in a specific area. This refined control strategy allows cooling resources to be utilized more effectively, avoiding problems of over-cooling or under-cooling.

[0050] This application effectively solves the problems of spatial sampling blind spots, sensing lag, and asynchronous sampling by multiple sensors in traditional data center control, thereby enabling timely detection and intervention of local hotspots, preventing their spread, and ensuring the stable operation of the data center. This refined control enables precise intervention of local hotspots, avoids waste of cooling resources, and improves the timeliness and effectiveness of hotspot handling.

[0051] like Figure 2 As shown, the reliability score of each temperature sensor is evaluated based on its relative position, thermal response characteristics, and power change rate, including: S101. Based on the relative position, calculate the Euclidean spatial distance between each temperature sensor and the center point of the air outlet of the cabinet it is responsible for monitoring.

[0052] The calculation of the Euclidean spatial distance between each temperature sensor and the center point of the air vent of the rack it is responsible for monitoring refers to quantifying the straight-line distance between each temperature sensor and the heat dissipation area of ​​the rack it is primarily responsible for monitoring (i.e., the center point of the rack's air vent) in the three-dimensional space of the data center. This Euclidean spatial distance is used to characterize the physical proximity of the sensor to monitor the temperature of a specific rack.

[0053] S102. Calculate the spatial representativeness score of the temperature sensor based on Euclidean spatial distance and exponential decay function.

[0054] The aim is to transform physical distance into an indicator that reflects the effectiveness of sensor monitoring. The exponential decay function simulates the decay characteristics of the temperature field in space; that is, the farther the sensor is from the monitored target, the weaker its representativeness, and this decay is non-linear. The spatial representativeness score is inversely proportional to the Euclidean spatial distance, meaning that the closer the distance, the higher the spatial representativeness score, and vice versa, thus ensuring a reasonable influence of spatial location on the score.

[0055] S103. The reliability score of each temperature sensor is evaluated by combining spatial representativeness score, thermal response characteristics and power change rate.

[0056] Thermal response characteristics can be understood as the speed and accuracy of a temperature sensor's response to changes in ambient temperature; sensors with faster response times are generally considered to have higher reliability. Power change rate reflects the dynamics of the cabinet's thermal load; drastic changes in cabinet power can lead to rapid fluctuations in local temperature, placing higher demands on the sensor's real-time monitoring capabilities. By weighting or fusing information from these three dimensions, a comprehensive reliability score can be obtained, which more comprehensively reflects the reliability of each temperature sensor reading.

[0057] This application's solution employs a multi-dimensional and refined evaluation of temperature sensor reliability by introducing spatial representativeness scores, thermal response characteristics, and power change rate. Specifically, by calculating the Euclidean spatial distance between the temperature sensor and the center point of the air outlet of the cabinet it is monitoring, and converting this distance into a spatial representativeness score based on an exponential decay function, the monitoring effectiveness of the sensor in its spatial location can be objectively quantified. The closer the distance, the stronger the spatial representativeness and the higher the score. Simultaneously, considering the thermal response characteristics of the temperature sensor, its ability to capture temperature changes can be reflected; sensors with faster response times generally have higher reliability. Furthermore, taking into account the power change rate of the cabinet reflects the dynamics of the local heat load in the data center; areas with large power change rates experience drastic temperature fluctuations, requiring higher sensor accuracy. By integrating these factors, the reliability of each temperature sensor can be evaluated more comprehensively and accurately, thus providing a more reliable data foundation for subsequent temperature compensation and three-dimensional temperature field generation.

[0058] In some of the above embodiments, a scheme was proposed to compensate for the temperature value of the abnormal temperature sensor based on the temperature rise acceleration of the abnormal temperature sensor, the temperature value of the normal temperature sensor in the adjacent area, and the power change rate. However, in the actual data center server room environment, the dynamic changes of the temperature field are complex and rapid. If only simple compensation is performed, it may not be able to fully capture the rapid heating trend of hot spots, the diffusion of heat in space, and the impact of rack power changes on local temperature. This will lead to a deviation between the compensated temperature value and the actual situation, affecting the accuracy of the subsequent three-dimensional temperature field distribution and the effectiveness of adjustment commands.

[0059] In this regard, this application further proposes the following steps for compensating the temperature value of the abnormal temperature sensor: S201. Determine the spatial temperature gradient of adjacent areas based on the temperature values ​​of normal temperature sensors in adjacent areas.

[0060] Specifically, determining the spatial temperature gradient of an adjacent region refers to calculating the rate of temperature change in that region by analyzing the temperature values ​​measured by normal temperature sensors around the abnormal temperature sensor. For example, the finite difference method or interpolation method can be used to construct a local temperature field model based on the location and temperature values ​​of adjacent normal temperature sensors, and then derive the spatial temperature gradient at the location of the abnormal temperature sensor. This spatial temperature gradient reflects the non-uniformity of heat distribution in space.

[0061] S202. Determine the dynamic compensation item according to the compensation formula.

[0062] The compensation formula is: F = 1 / 2 × V × T 2 +A+B×W1×W2; Where F is the dynamic compensation term, V is the temperature rise acceleration of the abnormal temperature sensor, T is the preset step time, A is the projection increment of the spatial temperature gradient in the corresponding direction, B is the power change rate, W1 is the preset thermoelectric conversion coefficient, and W2 is the thermal inertia compensation gain.

[0063] The dynamic compensation term F is a key parameter used to correct the original temperature value of the abnormal temperature sensor, aiming to more accurately reflect the actual temperature of the area. This compensation term comprehensively considers various dynamic thermal factors. V is the temperature rise acceleration of the abnormal temperature sensor, representing the rate of change of temperature rise, which can capture the trend of rapid temperature rise in hot spots. T is the preset step time, which refers to the time interval set when performing temperature compensation calculations, such as a data acquisition cycle or a control cycle. A is the projection increment of the spatial temperature gradient in the corresponding direction, which quantifies the influence of the spatial temperature gradient on the location of the abnormal sensor, reflecting the heat conduction effect from high-temperature areas to low-temperature areas. B is the power change rate, which refers to the amount of power change of the cabinet per unit time, directly related to the increase or decrease of the cabinet's heat generation, and is an important factor affecting the local temperature of the computer room. W1 is the preset thermoelectric conversion coefficient, used to convert the cabinet power change rate B into an impact on temperature, reflecting the efficiency of electrical energy conversion into heat energy. W2 is the thermal inertia compensation gain, used to characterize the hysteresis response of the computer room environment to temperature changes. That is, after a power change or heat input, the temperature does not immediately reach stability, but there is a certain delay and inertia.

[0064] S203. Compensate the temperature value of the abnormal temperature sensor according to the dynamic compensation item.

[0065] In some preferred embodiments, it is assumed that a temperature sensor is identified as abnormal in a certain area of ​​a data center server room. The temperature rise acceleration V of this abnormal temperature sensor in the most recent preset step time T (e.g., 10 seconds) is 0.05℃ / s². By analyzing the normal temperature sensor data in its adjacent area, the projected increment A of the spatial temperature gradient at the location of the abnormal sensor in the corresponding direction is calculated to be 0.2℃. Simultaneously, the power change rate B of the rack monitored by the abnormal sensor is detected to be 100W / s. The preset thermoelectric conversion coefficient W1 is known to be 0.001℃ / (W·s), and the thermal inertia compensation gain W2 is 0.5.

[0066] According to the compensation formula F=1 / 2×V×T2+A+B×W1×W2, the dynamic compensation term F can be calculated: F = 1 / 2 × 0.05 × (10)² + 0.2 + 100 × 0.001 × 0.5 F = 1 / 2 × 0.05 × 100 + 0.2 + 0.05 F = 2.5 + 0.2 + 0.05 F = 2.75 ℃ If the original temperature value measured by the abnormal temperature sensor is 35℃, the compensated temperature value will be 35℃ + 2.75℃ = 37.75℃. In this way, even if the sensor itself is abnormal, its temperature value can be dynamically and accurately corrected, thus providing reliable data support for generating an accurate three-dimensional temperature field distribution.

[0067] The proposed solution introduces a dynamic compensation term, enabling more comprehensive and accurate compensation for the temperature values ​​of abnormal temperature sensors. Specifically, the product of the temperature rise acceleration V and the preset step time T (1 / 2 × V × T²) effectively predicts the temperature change trend of the abnormal area within a short period, particularly suitable for capturing rapid temperature rises in hotspot areas. The projected increment A of the spatial temperature gradient in the corresponding direction considers the conduction and diffusion effects of heat in space, ensuring that the compensated temperature value better aligns with the temperature distribution of the surrounding normal area. Furthermore, the product of the power change rate B, the preset thermoelectric conversion coefficient W1, and the thermal inertia compensation gain W2 (B × W1 × W2) directly relates to the dynamic changes in the cabinet's heat generation and its impact on local temperature, while also considering the thermal inertia of the environment, thus enabling timely compensation to respond to heat fluctuations caused by changes in cabinet load. By integrating these dynamic factors, this compensation mechanism overcomes the lag and inaccuracy that may exist in traditional compensation methods, resulting in a compensated temperature value that is closer to the actual physical state.

[0068] In some embodiments described above, this application proposes generating and issuing adjustment commands based on a three-dimensional temperature field distribution, matching the effective range of the cooling terminal equipment. However, in its implementation, if the command generation is based solely on a static three-dimensional temperature field distribution, it may fail to effectively identify and address dynamic hotspots within the data center, leading to inaccurate cooling resource allocation and potentially missing the optimal intervention window, thus affecting the stable operation of data center equipment. To address this, this application further proposes a more refined scheme for generating and issuing adjustment commands. By locating hotspots and analyzing their evolution stages, more targeted adjustment commands can be generated.

[0069] In this regard, this application further proposes the above-mentioned method of generating and issuing adjustment commands based on the three-dimensional temperature field distribution, which are matched to the effective range of the cooling terminal equipment, including: S301. Based on the three-dimensional temperature field distribution, locate the hot spot.

[0070] Specifically, locating hotspots refers to identifying areas in the generated three-dimensional temperature field that show abnormally high temperatures or may cause equipment overheating. For example, based on the three-dimensional temperature field distribution, the coordinates of locations where the temperature value exceeds a threshold and the spatial temperature gradient exceeds a threshold can be determined, thus locating the hotspot. The threshold for identification and the gradient threshold can be set according to the specific operating requirements of the data center and the temperature tolerance of the equipment.

[0071] S302. Analyze the temperature field change sequence of the hotspot over multiple consecutive cycles to determine the evolution stage of the hotspot.

[0072] The evolutionary stages include the initial stage, the diffusion stage, and the overheating stage. The initial stage usually refers to the hotspot just forming, with a small area and a rapid temperature rise, but not yet reaching a dangerous level. The diffusion stage refers to the hotspot gradually expanding, affecting the surrounding area, and the temperature continuing to rise. The overheating stage indicates that the hotspot temperature has reached or exceeded the dangerous threshold, posing a serious threat to equipment operation.

[0073] Specifically, if the first preset condition is met, the evolution stage of the hotspot is determined to be the overheating stage; the first preset condition is that the temperature within a preset range from the hotspot is greater than a preset temperature threshold. If the second preset condition is met, the evolution stage of the hotspot is determined to be the initial stage; the second preset condition is that the temperature of some areas within a preset range from the hotspot is less than or equal to a preset temperature threshold, and the temperature rise rate of the hotspot is greater than the speed threshold, and the temperature rise rate within the preset range from the hotspot is less than the speed threshold. If the first and second preset conditions are not met, the hotspot is determined to be in the diffusion stage of its evolution.

[0074] The first preset condition aims to identify situations where a hotspot has developed to a severe level. Specifically, when the temperature in all areas within a preset range around the hotspot exceeds a preset temperature threshold, it indicates that heat has accumulated extensively and localized cooling measures may no longer be sufficient to control it effectively. At this point, the hotspot is determined to be in an overheating stage. The preset temperature threshold can be set based on factors such as the data center's operating specifications, equipment's temperature tolerance, and safety margins; for example, it can be set to 30°C or higher.

[0075] The second preset condition is used to identify situations where a hotspot is just beginning to form and has a rapid development trend. Specifically, when the temperature in some areas within a preset range around the hotspot has not yet reached a preset temperature threshold, but the hotspot's own temperature rise rate has exceeded a speed threshold, while the temperature rise rate in its surrounding areas is relatively low, this indicates that the hotspot is forming rapidly locally but has not yet spread widely. The speed threshold can be determined based on historical data, empirical values, or through experiments; for example, it can be set to 0.5℃ / minute.

[0076] If neither the first nor the second pre-defined condition is met, it indicates that the hotspot has neither reached the severity of overheating nor is it in the initial rapid formation stage, but rather is in an intermediate, gradually spreading state. At this point, the hotspot is determined to be in the diffusion stage, meaning that heat is slowly spreading from the hotspot area to the surrounding environment, and corresponding intervention measures are needed to prevent further deterioration.

[0077] S303. Based on the location and evolution stage of the hotspot, generate and issue adjustment commands that match the effective range of the cooling terminal equipment.

[0078] The generation of adjustment commands takes into account the specific location of the hotspot, its current evolutionary stage, and the actual range and capability of the cooling terminal equipment to ensure the accuracy and effectiveness of cooling intervention. For example, different cooling strategies can be adopted for hotspots at different evolutionary stages. In the initial stage, only localized enhanced cooling may be needed, while in the overheating stage, strong global cooling intervention may be required.

[0079] Specifically, if the hot spot is in the initial stage of evolution, the fan speed of the cooling terminal device closest to the hot spot should be increased. If the hot spot is in the diffusion stage of its evolution, then reduce the air supply temperature of the cooling terminal equipment within a preset range from the hot spot. If the hot spot is in an overheating stage of evolution, then reduce the supply air temperature of all cooling terminal equipment.

[0080] In some preferred embodiments, assuming that a temperature sensor in a certain area of ​​the data center server room detects a continuous rise in temperature, the system locates a hotspot in that area based on the three-dimensional temperature field distribution.

[0081] Specifically, if the system analyzes the temperature field change sequence of the hot spot over several consecutive periods and determines that it is in its initial stage—for example, the temperature in the hot spot area has just exceeded a preset threshold, and the temperature change in the surrounding area is not significant—the system will immediately identify the cooling terminal device closest to the hot spot and issue an adjustment command to it, increasing the fan speed of the cooling terminal device from the default 800 RPM to 1200 RPM to increase the local supply of cool air and quickly remove heat.

[0082] If the hotspot develops further, the system determines that it has entered the diffusion stage. For example, the temperature in the hotspot area continues to rise, and the temperature in some areas within a preset range around it also begins to rise. At this time, the system will issue adjustment commands to all cooling terminal devices within a preset range from the hotspot location, reducing the air supply temperature of these devices from 22°C to 20°C to enhance the overall cooling capacity of the area and effectively suppress the spread of heat to a wider area.

[0083] If a hotspot cannot be effectively controlled, the system determines that it has entered an overheating stage. For example, if the temperature of all items within a preset range of the hotspot exceeds a preset temperature threshold, it indicates that the hotspot has severely affected a large area. At this point, the system will issue adjustment commands to all cooling terminal devices in the computer room, uniformly reducing the air supply temperature of all devices to 18°C ​​to achieve rapid cooling of the entire computer room and ensure safe equipment operation.

[0084] This application's solution addresses the potential blindness and lag issues in traditional methods for generating control commands by introducing hotspot location and evolution stage analysis. Specifically, by locating hotspots based on three-dimensional temperature field distribution, cooling interventions can be precisely applied to areas requiring cooling, avoiding overcooling of non-hotspot areas and thus improving cooling efficiency. Simultaneously, by analyzing the temperature field change sequence of hotspots over multiple consecutive cycles, the formation, development, and deterioration processes of hotspots can be dynamically monitored, enabling early warning and intervention in the early stages of hotspot development. This effectively prevents hotspots from spreading to overheating stages and nipps potential risks in the bud. This command generation mechanism based on the dynamic evolution stage of hotspots allows cooling strategies to adaptively adjust according to actual needs, ensuring the stable operation of data center equipment and the efficient use of energy.

[0085] In some embodiments, the method further includes: S401. After the adjustment command is issued, continuously acquire temperature data from the computer room.

[0086] Temperature data acquisition reflects the actual temperature changes after the execution of control commands and is a key basis for evaluating the control effect. Among them, hot spots refer to areas in the computer room where the temperature rises abnormally, and their location is usually determined by the three-dimensional temperature field distribution.

[0087] S402: Obtain the parameters of the issued adjustment command, the spatial distance from the cooling terminal device to the temperature sensor, and the current air supply speed.

[0088] S403. Adjust the control command based on the temperature data, the issued control command parameters, the spatial distance, and the current air supply speed.

[0089] The adjustment process may include modifications to the adjustment intensity, the object of adjustment, or the adjustment priority to more precisely match the cooling effect to actual needs. For example, if the actual temperature drop is not as expected, the adjustment intensity may need to be increased; if the temperature drops too quickly, the adjustment intensity may need to be appropriately reduced to avoid overcooling.

[0090] In some preferred embodiments, a specific example is given below. Suppose a hotspot is detected in an area of ​​a data center server room, and its temperature has exceeded a preset threshold. Based on the three-dimensional temperature field distribution, the system generates and issues an adjustment command, for example, instructing the cooling terminal device closest to the hotspot to reduce its air supply temperature by 2 degrees Celsius. After the command is issued and executed for a period of time, the system continuously acquires temperature data from the hotspot area. If the actual temperature drop is less than expected, or the rate of drop is too slow, the system performs a comprehensive analysis based on the issued command parameter of "reducing the air supply temperature by 2 degrees Celsius," the spatial distance between the cooling terminal device and the hotspot sensor, and the current air supply speed. Based on this analysis, the system may determine that further cooling intervention is needed, for example, adjusting the adjustment command to "reducing the air supply temperature by another 1 degree Celsius," or simultaneously increasing the fan speed of the cooling terminal device. Conversely, if the actual temperature drops too quickly, the system may also adjust the command to appropriately increase the air supply temperature to avoid over-cooling and energy waste. Through this dynamic feedback and adjustment, the hotspot can be eliminated accurately and efficiently.

[0091] This application's solution effectively addresses the lag or inaccuracy issues that may exist in traditional adjustment commands by introducing a closed-loop feedback mechanism. Specifically, after an adjustment command is issued, the system no longer simply waits for its effect but continuously acquires temperature data from the computer room to monitor the cooling effect in real time. Simultaneously, by combining the issued adjustment command parameters, the spatial distance between the cooling terminal equipment and the temperature sensor, and key operating parameters such as the current airflow velocity, the system can comprehensively evaluate the effectiveness of the current cooling strategy. For example, spatial distance and airflow velocity help quantify the time and intensity required for the cooling airflow to reach the hotspot area, while the adjustment command parameters provide detailed information on the current intervention measures. By comparing this real-time data with the expected cooling effect, the system can identify the deviation between the adjustment command and the actual demand. Therefore, based on the magnitude and direction of the deviation, the adjustment command can be dynamically adjusted, such as increasing or decreasing fan speed, raising or lowering airflow temperature, thereby ensuring that cooling resources are accurately allocated and utilized, enabling hotspots to be eliminated promptly and effectively.

[0092] In some of the embodiments described above in this application, a scheme is proposed to adjust the adjustment command based on temperature data, the parameters of the issued adjustment command, the spatial distance between the cooling terminal equipment and the temperature sensor, and the current airflow velocity after the adjustment command is issued. However, in practical applications, adjusting based solely on these parameters may not accurately assess the difference between the actual and expected effects of the cooling intervention, resulting in inaccurate adjustments to the adjustment command. This may prevent the timely and effective elimination of hot spots, or even cause over-cooling or under-cooling, affecting the overall energy efficiency of the data center and the operational stability of the equipment.

[0093] In this regard, this application further proposes that the steps of adjusting the control command based on temperature data, issued control command parameters, spatial distance, and current supply air velocity include: S501. Based on the temperature data, determine whether the cooling intervention of the adjustment command has taken effect on the hot spot.

[0094] Specifically, determining whether the cooling intervention of the adjustment command has taken effect on the hot spot can be based on a comparison between temperature data collected and a preset effective threshold or trend. For example, if the actual temperature in the hot spot area shows a continuous downward trend for a period of time after the adjustment command is issued, and the decrease reaches or exceeds the preset minimum temperature reduction, then the cooling intervention can be considered effective. This determination aims to confirm that the cooling system has begun to affect the hot spot area, avoiding unnecessary or erroneous adjustments before the cooling effect is apparent.

[0095] S502. If it has taken effect, calculate the theoretical cooling effect of the adjustment command on the hot spot based on the adjustment command parameters, spatial distance and current air supply speed.

[0096] If cooling intervention has taken effect, the theoretical temperature reduction of the hotspot is calculated based on the adjustment command parameters, spatial distance, and current airflow velocity. Adjustment command parameters may include airflow temperature and fan speed; spatial distance refers to the distance between the cooling terminal equipment and the hotspot location; and current airflow velocity refers to the actual output airflow velocity of the cooling terminal equipment. The theoretical temperature reduction can be calculated using a pre-established physical model, empirical formula, or machine learning model. This model comprehensively considers factors such as the performance of the cooling terminal equipment, the hotspot location, airflow organization, and the heat load of the data center to predict the temperature change in the hotspot area under a given adjustment command.

[0097] S503. Subtract the theoretical cooling rate from the temperature in the three-dimensional temperature field distribution to obtain the controlled expected temperature.

[0098] The three-dimensional temperature field distribution comprehensively reflects the current temperature status of the computer room. The theoretical cooling range represents the amount of temperature reduction that the adjustment command should bring under ideal conditions. Therefore, the controlled expected temperature refers to the temperature value that the hotspot area should reach under ideal conditions after the adjustment command takes effect.

[0099] S504. Compare the temperature recovery data with the controlled expected temperature to obtain the reconstruction deviation.

[0100] Temperature data collected back represents the actual monitored temperature of the hotspot area, while the controlled expected temperature is the ideal temperature predicted by the model. By comparing the two, the difference between the actual cooling effect and the theoretical prediction can be quantified, i.e., the reconstruction bias. This bias reflects the inconsistency between the model prediction and the actual situation, which may be caused by factors such as model errors, environmental disturbances, or incomplete response of the cooling system.

[0101] S505. Adjust the control command according to the reconstruction deviation.

[0102] If the reconfiguration deviation is positive (i.e., the actual temperature is higher than the controlled expected temperature), it indicates insufficient cooling, requiring further cooling intervention, such as lowering the supply air temperature or increasing the fan speed. If the reconfiguration deviation is negative (i.e., the actual temperature is lower than the controlled expected temperature), it indicates potential overcooling, and cooling intervention can be appropriately reduced to save energy. This deviation-based feedback adjustment mechanism enables precise and dynamic control of hotspots.

[0103] In some preferred embodiments, a specific example is given below. Suppose a hotspot is detected in an area of ​​a data center server room, with a temperature of 35°C, exceeding a preset temperature threshold. Based on the three-dimensional temperature field distribution, the system generates and issues an adjustment command, for example, reducing the air supply temperature of the cooling terminal device closest to the hotspot from 20°C to 18°C ​​and increasing its fan speed by 10%. After the command is issued and a period of time has elapsed, the system continuously acquires temperature data from within the server room.

[0104] First, the system determines whether the cooling intervention has taken effect based on the temperature data collected. For example, if the hot spot temperature has dropped from 35°C to 34°C and the downward trend is stable, it is determined that the intervention has taken effect.

[0105] Next, based on the issued adjustment command parameters (air supply temperature 18℃, fan speed increased by 10%), the spatial distance between the cooling terminal equipment and the hot spot, and the current air supply speed, the system calculates, through a preset thermodynamic model, that the theoretical temperature reduction of the hot spot caused by the adjustment command is 2℃.

[0106] Then, the system subtracts the theoretical cooling rate of 2℃ from the temperature at the hot spot location in the original three-dimensional temperature field distribution (assumed to be 35℃) to obtain the controlled expected temperature of 33℃.

[0107] At this point, the system obtains the latest temperature data, for example, the actual temperature of the hotspot is 33.5℃. This temperature data of 33.5℃ is compared with the controlled expected temperature of 33℃, resulting in a reconstruction deviation of 0.5℃ (33.5℃ - 33℃).

[0108] Since the reconstruction deviation is positive, it indicates that the actual cooling effect is slightly lower than expected. The system will fine-tune the adjustment command based on this 0.5℃ deviation. For example, the supply air temperature of the cooling terminal device can be further reduced by 0.5℃, or the fan speed can be increased by 5%, in order to more accurately control the hot spot temperature at the expected level. Through this iterative and refined adjustment, it is ensured that the hot spot can be eliminated quickly and stably.

[0109] This application's solution addresses the potential blindness and inaccuracy of existing solutions in adjusting control commands by introducing a mechanism that compares theoretical predictions of cooling intervention effects with actual monitoring. Specifically, after the control command is issued and confirmed to be effective, the system no longer relies solely on simple changes in actual temperature. Instead, it first calculates the theoretical cooling effect on the hotspot area based on known control command parameters, the spatial distance between the cooling terminal equipment and the temperature sensor, and the current airflow velocity, combined with a preset thermodynamic or empirical model. This calculation of the theoretical cooling effect provides a benchmark for subsequent precise adjustments. Then, the theoretical cooling effect is subtracted from the current temperature in the three-dimensional temperature field distribution to obtain a controlled expected temperature, representing the target temperature that the hotspot area should ideally reach under the current control command. By precisely comparing the actual acquired temperature data with this controlled expected temperature, a reconstruction deviation can be quantified. This reconstruction deviation directly reflects the gap between the actual cooling effect and the theoretical expected effect. Based on this quantified deviation, the system can adjust the control commands in a targeted and intelligent manner. For example, when the actual temperature is higher than expected, the cooling intervention is enhanced; when the actual temperature is lower than expected, the cooling intervention is appropriately weakened. This closed-loop adjustment mechanism, based on model prediction and actual feedback, makes cooling control more refined and intelligent.

[0110] Specifically, the location of hotspots based on the three-dimensional temperature field distribution includes: S601. Based on the three-dimensional temperature field distribution, determine the location coordinates where the temperature value is greater than the identification threshold and the spatial temperature gradient is greater than the gradient threshold.

[0111] Specifically, in the three-dimensional temperature field distribution of a data center, hotspots typically manifest as localized areas with significantly higher temperatures than the surrounding environment and large temperature gradients. Therefore, to accurately identify and locate these hotspots, it is necessary to comprehensively consider both the absolute value of the temperature and its spatial rate of change. "Temperature value greater than the identification threshold" means that the temperature value at a certain location in the three-dimensional temperature field distribution exceeds a pre-set upper temperature limit, which is used to initially screen areas that may have thermal anomalies. "Spatial temperature gradient greater than the gradient threshold" means that at that location, the rate of change of temperature in space (i.e., the temperature gradient) exceeds a pre-set upper gradient limit. The spatial temperature gradient can be understood as the intensity of temperature change in different directions; the larger the value, the more drastic the temperature change in that direction. By simultaneously satisfying these two conditions, areas with high but uniform temperatures, as well as areas with drastic temperature changes but low absolute temperatures, can be effectively excluded, thus more accurately pinpointing the true hotspot areas. The identification threshold and gradient threshold can be flexibly configured and adjusted according to the specific operating conditions of the data center, the type of equipment, and the requirements for hotspot sensitivity.

[0112] S602. Locate the hotspot based on location coordinates.

[0113] This application's solution, by simultaneously considering both temperature values ​​and spatial temperature gradients, enables more accurate identification and location of hotspots within data center server rooms. Traditional hotspot location methods may rely solely on temperature thresholds, potentially leading to misclassification of large-scale temperature rise areas as hotspots, resulting in inefficient or inaccurate allocation of cooling resources. By introducing spatial temperature gradients as a criterion, this application's solution ensures that identified hotspots not only possess high temperatures but also exhibit significant localized concentration, meaning a substantial temperature increase over a short distance. This dual-discrimination mechanism effectively distinguishes between locally overheated areas and generalized temperature rise areas, avoiding misjudgments and improving the accuracy and reliability of hotspot location.

[0114] like Figure 3 As shown in the figure, this embodiment of the invention also provides a programmable logic control system for data center equipment. The system includes: The information acquisition module is used to acquire the temperature values ​​of each temperature sensor in the computer room, the relative position of each temperature sensor to the cabinet, the thermal response characteristics of the temperature sensors, and the power change rate of the cabinet. The scoring module is used to evaluate the reliability score of each temperature sensor based on its relative position, thermal response characteristics, and power change rate. The compensation module is used to compensate the temperature value of the abnormal temperature sensor based on the temperature rise acceleration of the abnormal temperature sensor, the temperature value of the normal temperature sensor in the adjacent area, and the power change rate. The abnormal temperature sensor is a temperature sensor with a confidence score less than or equal to the scoring threshold; the normal temperature sensor is a temperature sensor with a confidence score greater than the scoring threshold. The temperature field generation module is used to generate a three-dimensional temperature field distribution in the computer room based on the temperature values ​​of the compensated abnormal temperature sensors and the temperature values ​​of the normal temperature sensors. The instruction generation and issuance module is used to generate and issue adjustment instructions that match the range of action of the cooling terminal equipment based on the three-dimensional temperature field distribution. The adjustment instructions include the adjustment object, adjustment priority, and adjustment intensity.

[0115] This application also provides a computer-readable storage medium. All or part of the processes in the above method embodiments can be executed by a computer program instructing related hardware. This program can be stored in the computer-readable storage medium, and when executed, it can include the processes of the above method embodiments. The computer-readable storage medium can be an internal storage unit of the task execution device (including a data sending end and / or a data receiving end) of any of the foregoing embodiments, such as the hard disk or memory of the task execution device. The computer-readable storage medium can also be an external storage device of the terminal device, such as a plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, etc., equipped on the terminal device. Further, the computer-readable storage medium can include both the internal storage unit of the task execution device and an external storage device. The computer-readable storage medium is used to store the computer program and other programs and data required by the task execution device. The computer-readable storage medium can also be used to temporarily store data that has been output or will be output.

[0116] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0117] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a readable storage medium. Based on this understanding, the technical solutions of the embodiments of this application, essentially, or the parts that contribute to the prior art, or all or part of the technical solutions, can be embodied in the form of a software product. This software product is stored in a storage medium and includes several instructions to cause a device (which may be a microcontroller, chip, etc.) or processor to execute all or part of the steps of the methods of the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, ROM, RAM, magnetic disks, or optical disks.

[0118] The above are merely specific embodiments of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions within the technical scope disclosed in this application should be covered within the scope of protection of this application.

Claims

1. A programmable logic control method for data center equipment, characterized in that, include: Acquire the temperature values ​​of each temperature sensor in the computer room, the relative position of each temperature sensor to the cabinet, the thermal response characteristics of the temperature sensors, and the power change rate of the cabinet; The reliability score of each temperature sensor is evaluated based on the relative position, the thermal response characteristics, and the power change rate. The temperature value of the abnormal temperature sensor is compensated based on the temperature rise acceleration of the abnormal temperature sensor, the temperature value of the normal temperature sensor in the adjacent area, and the power change rate; the abnormal temperature sensor is a temperature sensor with a confidence score less than or equal to the scoring threshold; the normal temperature sensor is a temperature sensor with a confidence score greater than the scoring threshold. Based on the temperature values ​​of the compensated abnormal temperature sensor and the temperature values ​​of the normal temperature sensor, a three-dimensional temperature field distribution of the computer room is generated. Based on the three-dimensional temperature field distribution, adjustment instructions are generated and issued that match the effective range of the cooling terminal equipment. The adjustment instructions include the adjustment object, adjustment priority, and adjustment intensity.

2. The programmable logic control method for data center equipment according to claim 1, characterized in that, The reliability score of each temperature sensor is evaluated based on the relative position, the thermal response characteristics, and the power change rate, including: Based on the relative positions, calculate the Euclidean spatial distance between each temperature sensor and the center point of the air outlet of the cabinet it is responsible for monitoring; The spatial representativeness score of the temperature sensor is calculated based on Euclidean spatial distance and an exponential decay function; the spatial representativeness score is inversely proportional to the Euclidean spatial distance. The reliability score of each temperature sensor is evaluated by combining the spatial representativeness score, the thermal response characteristics, and the power change rate.

3. The programmable logic control method for data center equipment according to claim 1, characterized in that, The compensation for the temperature value of the abnormal temperature sensor based on the temperature rise acceleration of the abnormal temperature sensor, the temperature value of the normal temperature sensor in the adjacent area, and the power change rate includes: The spatial temperature gradient of the adjacent areas is determined based on the temperature values ​​of the normal temperature sensors in the adjacent areas. The dynamic compensation item is determined according to the compensation formula, which is: F=1 / 2×V×T 2 +A+B×W1×W2; Wherein, F is the dynamic compensation term, V is the temperature rise acceleration of the abnormal temperature sensor, T is the preset step time, A is the projection increment of the spatial temperature gradient in the corresponding direction; B is the power change rate, W1 is the preset thermoelectric conversion coefficient, and W2 is the thermal inertia compensation gain. The temperature value of the abnormal temperature sensor is compensated according to the dynamic compensation term.

4. The programmable logic control method for data center equipment according to claim 1, characterized in that, The process of generating and issuing adjustment commands that match the effective range of the cooling terminal equipment based on the three-dimensional temperature field distribution includes: Based on the three-dimensional temperature field distribution, the location of the hotspots is determined; The temperature field change sequence of the hotspot over multiple consecutive periods is analyzed to determine the evolution stage of the hotspot; the evolution stage includes the initial stage, the diffusion stage, and the overheating stage; Based on the location and evolution stage of the hotspot, adjustment commands are generated and issued that match the effective range of the cooling terminal equipment.

5. The programmable logic control method for data center equipment according to claim 4, characterized in that, The analysis of the temperature field change sequence of the hotspot over multiple consecutive periods to determine the evolution stage of the hotspot includes: If the first preset condition is met, the evolution stage of the hotspot is determined to be the overheating stage; the first preset condition is that the temperature within a preset range from the hotspot is greater than a preset temperature threshold. If the second preset condition is met, the evolution stage of the hotspot is determined to be the initial stage; the second preset condition is that the temperature of some areas within a preset range from the hotspot is less than or equal to the preset temperature threshold, and the temperature rise rate of the hotspot is greater than the speed threshold, and the temperature rise rate within the preset range from the hotspot is less than the speed threshold. If the first preset condition and the second preset condition are not met, then the evolution stage of the hotspot is determined to be the diffusion stage.

6. The programmable logic control method for data center equipment according to claim 4, characterized in that, The step of generating and issuing adjustment commands that match the effective range of the cooling terminal equipment based on the location and evolution stage of the hotspot includes: If the hot spot is in the initial stage of evolution, then increase the fan speed of the cooling terminal device closest to the hot spot. If the hot spot is in the diffusion stage of its evolution, then reduce the air supply temperature of the cooling terminal equipment within a preset range from the hot spot. If the hot spot is in an overheating stage of evolution, then reduce the supply air temperature of all cooling terminal equipment.

7. The programmable logic control method for data center equipment according to claim 1, characterized in that, The method further includes: After the adjustment command is issued, temperature data in the computer room is continuously acquired. Obtain the issued adjustment command parameters, the spatial distance from the cooling terminal device to the temperature sensor, and the current air supply speed; The adjustment command is adjusted based on the temperature data, the issued adjustment command parameters, the spatial distance, and the current air supply speed.

8. A programmable logic control method for data center equipment according to claim 7, characterized in that, The step of adjusting the adjustment command based on the temperature data, the issued adjustment command parameters, the spatial distance, and the current air supply speed includes: Based on the temperature data collected, it is determined whether the cooling intervention of the adjustment command has taken effect on the hot spot; If it has taken effect, then the theoretical cooling effect of the adjustment command on the hot spot is calculated based on the adjustment command parameters, the spatial distance and the current air supply speed; Subtracting the theoretical cooling rate from the temperature in the three-dimensional temperature field distribution yields the controlled expected temperature. The temperature data is compared with the controlled expected temperature to obtain the reconstruction deviation; The adjustment command is adjusted based on the reconstruction deviation.

9. A programmable logic control method for data center equipment according to claim 4, characterized in that, The process of locating the hotspots based on the three-dimensional temperature field distribution includes: Based on the three-dimensional temperature field distribution, the coordinates of the locations where the temperature value is greater than the identification threshold and the spatial temperature gradient is greater than the gradient threshold are determined. Based on the location coordinates, the location of the hotspot is determined.

10. A programmable logic control system for data center equipment, characterized in that, The system includes: The information acquisition module is used to acquire the temperature values ​​of each temperature sensor in the computer room, the relative position of each temperature sensor to the cabinet, the thermal response characteristics of the temperature sensors, and the power change rate of the cabinet. The scoring module is used to evaluate the reliability score of each temperature sensor based on the relative position, the thermal response characteristics, and the power change rate. The compensation module is used to compensate the temperature value of the abnormal temperature sensor based on the temperature rise acceleration of the abnormal temperature sensor, the temperature value of the normal temperature sensor in the adjacent area, and the power change rate; the abnormal temperature sensor is a temperature sensor with a confidence score less than or equal to a scoring threshold; the normal temperature sensor is a temperature sensor with a confidence score greater than a scoring threshold. The temperature field generation module is used to generate a three-dimensional temperature field distribution in the computer room based on the temperature values ​​of the compensated abnormal temperature sensors and the temperature values ​​of the normal temperature sensors. The instruction generation and issuance module is used to generate and issue adjustment instructions that match the range of action of the cooling terminal equipment based on the three-dimensional temperature field distribution. The adjustment instructions include the adjustment object, adjustment priority, and adjustment intensity.

Images