Temperature control method of sheet metal cabinet with intelligent temperature control system

Abstract

The invention relates to the technical field of cabinet temperature control, in particular to a temperature control method of a sheet metal cabinet with an intelligent temperature control system. According to extreme value fluctuation characteristics and change frequencies of temperature values of all monitoring positions in a historical analysis period of the current moment, temperature value differences of different monitoring positions at the current moment are adjusted, and the heating unevenness of the cabinet is obtained; according to the fluctuation degree of the dominant heat source quantity in the local time period of each moment and the dominant heat source position change condition of the adjacent moment, the heat source comprehensive change degree is obtained, and the multi-heat-source temperature change possibility degree is obtained in combination with the correlation degree of the temperature values of different monitoring positions in the historical analysis time period of the current moment and the dominant heat source quantity; and performing real-time temperature regulation and control on the sheet metal cabinet according to the heating unevenness of the cabinet and the multi-heat-source temperature change possibility. According to the method, uneven heating in the cabinet and the root thereof under the complex thermal working condition are considered, and the internal temperature regulation and control precision of the cabinet is improved.

Description

Technical Field

[0001] This invention relates to the field of cabinet temperature control technology, and specifically to a temperature control method for a sheet metal cabinet with an intelligent temperature control system. Background Technology

[0002] Sheet metal cabinets are the core mounting carriers for electrical and electronic equipment. They not only provide physical protection for the internal equipment, effectively blocking dust, moisture, and external impacts, but also have the characteristic of inhibiting internal heat loss. The movement of equipment inside the sheet metal cabinet generates a large amount of heat. To ensure the stable operation of the internal equipment, sheet metal cabinets are usually equipped with intelligent temperature control systems to regulate the internal temperature.

[0003] Existing technology, patent document CN119212331A, discloses a module temperature control method for a server rack. Specifically, it acquires the highest and lowest operating temperatures of the modules after the rack has stabilized, determines whether the highest temperature falls within the module's upper tolerance temperature range, and whether the difference between the highest and lowest temperatures is less than the module's maximum allowable temperature difference. It then dynamically adjusts the axial fan airflow to implement internal temperature control. In practical applications of sheet metal server racks, the addition or removal of internal operating equipment and sudden changes in core loads such as servers cause real-time changes in heat generation power and heat source distribution, resulting in uneven temperature distribution within the rack and multiple dynamically changing locations of abnormal temperature increases. Using only the highest and lowest operating temperatures of the modules is insufficient to accurately reflect the temperature distribution of the rack under complex thermal dynamics, making temperature control prone to deviations and leading to poor heat dissipation. Summary of the Invention

[0004] To address the technical problem of insufficient temperature control accuracy caused by uneven temperature distribution inside the cabinet under complex dynamic thermal conditions, the present invention aims to provide a temperature control method for sheet metal cabinets with an intelligent temperature control system. The specific technical solution adopted is as follows: One embodiment of the present invention provides a temperature control method for a sheet metal cabinet with an intelligent temperature control system, the method comprising: Real-time acquisition of temperature values ​​at each monitoring location within the sheet metal cabinet at every moment; Based on the extreme fluctuation characteristics and change frequency of temperature values ​​at each monitoring location during the historical analysis period at the current moment, the temperature value differences at different monitoring locations at the current moment are adjusted to obtain the cabinet heat unevenness at the current moment. Obtain the dominant heat source quantity at each moment; based on the fluctuation degree of the dominant heat source quantity within a local time period at each moment and the change of the dominant heat source position at adjacent moments, obtain the comprehensive heat source change degree at the current moment; based on the correlation degree of temperature values ​​of different monitoring locations within the historical analysis period at the current moment, the comprehensive heat source change degree and the dominant heat source quantity, obtain the multi-heat source temperature change probability at the current moment. Based on the uneven heating of the cabinet and the probability of temperature changes from multiple heat sources, the sheet metal cabinet is subjected to real-time temperature control.

[0005] Furthermore, obtaining the current rack heat unevenness includes: Based on the extreme value fluctuation characteristics and change frequency of the temperature value of each monitoring location within the historical analysis period at the current moment, the attention level of each monitoring location at the current moment is obtained; The ratio of the absolute difference between the temperature values ​​of any two monitoring locations at the current moment to the distance between the two monitoring locations is taken as the local heating non-uniformity of the two monitoring locations at the current moment. Based on the average of the attention levels at any two monitoring locations at the current time, the local heat unevenness between all monitoring locations is weighted and summed to obtain the rack heat unevenness at the current time.

[0006] Furthermore, obtaining the attention level of each monitoring location at the current moment includes: The temperature values ​​of each monitoring location at all times within the historical analysis period at the current moment are arranged in chronological order to obtain a temperature sequence; global extrema and local extrema are extracted from the temperature sequence and recorded as characteristic extrema. For each eigenvalue, calculate the ratio of the absolute difference between any two adjacent eigenvalues ​​to the mean of the time intervals between the corresponding times of the two eigenvalues ​​and the current time, and denot it as the fluctuation impact degree; average the fluctuation impact degrees of all adjacent pairs of eigenvalues ​​to obtain the overall impact degree; The average of the global extreme value and the local extreme value is used as the temperature change performance factor of each monitoring location at the current moment. The proportion of local extrema in the temperature sequence is denoted as the temperature change frequency. The percentage of temperature values ​​in the temperature sequence that are greater than a preset temperature threshold is denoted as the high temperature intensity. Based on the temperature change performance factor, the temperature change frequency, and the high temperature intensity, the attention level of each monitoring location at the current moment is obtained.

[0007] Furthermore, obtaining the dominant heat source quantity at each moment includes: By performing a negative correlation mapping on the correlation coefficients of the temperature sequences of any two monitoring locations at each time step, the temperature distance between the two monitoring locations at each time step can be obtained. Based on the temperature distance, all monitoring locations are clustered, and the number of clusters at each time moment is recorded as the dominant heat source at each time moment.

[0008] Furthermore, obtaining the comprehensive change rate of the heat source at the current moment includes: For any two adjacent time points, calculate the ratio of the number of monitoring locations commonly contained in each cluster at the previous time point and each cluster at the next time point to the total number of all non-overlapping monitoring locations in the two corresponding clusters. This ratio is denoted as the location similarity of the corresponding clusters. The maximum value of the positional similarity between each cluster at the previous time step and all clusters at the next time step is selected, and the average of the maximum values ​​corresponding to all clusters at the previous time step is calculated to obtain the local positional invariance between two adjacent time steps. Calculate the average of the local position invariance between every two adjacent time intervals within the current time interval, and use it as the heat source position invariance at the current time. The variance of the dominant heat source quantity at all times within a local time period at the current time is denoted as the heat source quantity variation degree. Based on the invariance of the heat source location and the change in the number of heat sources, the overall change in the heat sources at the current moment is obtained.

[0009] Furthermore, obtaining the probability of temperature variation from multiple heat sources at the current moment includes: Calculate the ratio of the dominant heat source quantity to the total number of monitored locations at the current moment, and use it as the heat source distribution dispersion. The average correlation coefficient of the temperature sequence of every two monitoring locations at the current time is used as the neighborhood temperature similarity. Based on the comprehensive change degree of the heat source, the dispersion degree of the heat source distribution, and the similarity of the neighborhood temperature, the probability of temperature change of multiple heat sources at the current moment is obtained.

[0010] Furthermore, the real-time temperature control of the sheet metal cabinet based on the uneven heating of the cabinet and the probability of temperature changes from multiple heat sources includes: The product of the uneven heating of the cabinet and the probability of temperature change of the multiple heat sources is normalized, and the sum of the normalization result and the constant 1 is used as the temperature adjustment weight at the current moment. The sheet metal cabinet includes heat dissipation equipment; based on the temperature values ​​of all monitoring locations at the current moment and the temperature adjustment weights, the operating speed of the heat dissipation equipment at the current moment is obtained using a fuzzy control algorithm.

[0011] Furthermore, the overall variability of the heat source and the dispersion of the heat source distribution are both positively correlated with the probability of temperature change of the multiple heat sources, while the neighborhood temperature similarity is negatively correlated with the probability of temperature change of the multiple heat sources.

[0012] Furthermore, the K-means clustering algorithm was used to cluster all monitoring locations.

[0013] Furthermore, the correlation coefficient is the Pearson correlation coefficient.

[0014] The present invention has the following beneficial effects: In this embodiment of the invention, changes in the operating status of equipment inside the cabinet will cause temperature changes. By analyzing the extreme value fluctuation characteristics and frequency of change of temperature values ​​at each monitoring location during historical analysis periods, the required level of attention for each monitoring location at the current moment is determined. Since the temperature changes at different monitoring locations are more complex, adjusting the temperature differences at different monitoring locations using the level of attention can keenly capture local hot spots with high fluctuation risks that are masked by the overall temperature. This can accurately reflect the uneven heating of the cabinet under complex thermal conditions, obtaining the cabinet heating unevenness. By comprehensively considering the fluctuation degree and location changes of the dominant heat source, the internal heat sources of the cabinet can be analyzed. The dynamic changes in the layout yield a comprehensive change in heat sources. Further analysis, combined with the correlation of temperature values ​​at different monitoring locations over historical periods, generates the probability of temperature variations from multiple heat sources. This effectively identifies the root causes of temperature unevenness, enabling the control system to accurately recognize the more uncertain heat dissipation challenges caused by asynchronous small fluctuations from multiple heat sources. This provides a crucial basis for adopting the correct control strategy. By integrating the unevenness of heat generation within the cabinet with the probability of temperature variations from multiple heat sources, the system adaptively controls the cabinet temperature, adopting a more proactive global heat dissipation strategy. This improves the accuracy of controlling uneven temperature distribution within the cabinet under complex thermal conditions, thereby enhancing the cabinet's heat dissipation performance. Attached Figure Description

[0015] To more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0016] Figure 1 A flowchart illustrating the steps of a temperature control method for a sheet metal cabinet with an intelligent temperature control system, provided in an embodiment of the present invention. Figure 2 A flowchart illustrating a method for obtaining attention according to an embodiment of the present invention; Figure 3 This is a schematic diagram of a computer device for temperature control of a sheet metal cabinet with an intelligent temperature control system, provided as an embodiment of the present invention. Detailed Implementation

[0017] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of a temperature control method for a sheet metal cabinet with an intelligent temperature control system proposed according to the present invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0018] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0019] The following description, in conjunction with the accompanying drawings, details a specific scheme for a temperature control method for a sheet metal cabinet with an intelligent temperature control system provided by the present invention.

[0020] Example 1: This invention proposes a temperature control method for sheet metal cabinets with an intelligent temperature control system. Please refer to [link / reference]. Figure 1 The diagram illustrates a flowchart of a temperature control method for a sheet metal cabinet with an intelligent temperature control system according to an embodiment of the present invention. The method includes: Step S1: Real-time acquisition of temperature values ​​at each monitoring location within the sheet metal cabinet at each moment.

[0021] Sheet metal cabinets are used to provide physical protection for internal equipment and suppress internal heat leakage. For example, the heat dissipation cabinet disclosed in CN218888866U, entitled "A Heat Dissipation Cabinet", can be a sheet metal cabinet, and its heat dissipation device is a fan.

[0022] At least four temperature sensors are installed in key areas such as densely populated equipment areas, airflow channels (e.g., air inlet and outlet groups), and near fans within the sheet metal cabinet. Implementers can arrange the location and number of temperature sensors according to specific circumstances. Each temperature sensor is installed at a monitoring location to monitor the temperature value at each monitoring location at any given time. The data is then transmitted to the control unit, which determines the temperature and makes corresponding control decisions. Finally, the cooling system executes specific operations based on the control commands to achieve precise control of the cabinet temperature.

[0023] In this embodiment of the invention, the data acquisition frequency of the temperature sensor is set to 6 times per minute, which can be set by the implementer according to specific circumstances. All temperature sensors have the same data acquisition frequency, and data acquisition needs to be synchronized.

[0024] Step S2: Based on the extreme fluctuation characteristics and change frequency of the temperature values ​​of each monitoring location during the historical analysis period at the current moment, adjust the temperature value differences of different monitoring locations at the current moment to obtain the cabinet heating unevenness at the current moment.

[0025] When the network communication equipment such as servers, switches and routers inside the cabinet are running, the power loss of components such as CPU, memory and hard drive will be converted into heat. Power devices such as energy storage batteries and frequency converters will generate significant heat due to their internal resistance when operating under high current. Power lines and network cables will generate heat due to line loss when current passes through them. In addition, the ambient temperature of the cabinet fluctuates greatly, resulting in complex and different temperature changes at different monitoring locations inside the cabinet.

[0026] Equipment start-up and shutdown, as well as sudden load changes, can cause temperature fluctuations. Monitoring locations with larger temperature fluctuations have a more urgent need for heat dissipation. Frequent equipment start-up and shutdown, and fluctuations in computing power such as server task scheduling fluctuations, can lead to frequent temperature fluctuations, while short-term high-load tasks such as data backup and batch processing can cause occasional temperature fluctuations. Frequent fluctuations accelerate equipment aging and increase the risk of downtime, requiring close monitoring. By analyzing the extreme value fluctuations and frequency of changes in temperature values ​​at a monitoring location over a historical period, the required level of attention for the equipment at that location can be determined. Since fans dissipate heat throughout the entire rack area, the temperature differences between different monitoring locations at the current moment generally indicate uneven heating in localized areas. Because the complexity of temperature changes varies across different monitoring locations, adjusting for uneven heating in localized areas using the aforementioned level of attention can accurately reflect the uneven heating of the rack under complex operating conditions, thus obtaining the rack's heat unevenness.

[0027] In this embodiment of the invention, the 10 minutes preceding each moment are taken as the historical analysis period.

[0028] Step S3: Obtain the dominant heat source quantity at each moment; based on the fluctuation of the dominant heat source quantity within a local time period at each moment and the change in the dominant heat source position at adjacent moments, obtain the comprehensive heat source change degree at the current moment; based on the correlation of temperature values ​​at different monitoring locations within the historical analysis period at the current moment, the comprehensive heat source change degree, and the dominant heat source quantity, obtain the probability of multi-heat source temperature change at the current moment.

[0029] When a heat source is present at the monitoring location, the temperature change at that location is unique to that heat source. When no heat source is present, the temperature change at the monitoring location is influenced by the temperature changes of nearby heat sources. The start-up and shutdown times, heating and cooling rates, and other temperature change rhythms of the dominant heat source with large temperature fluctuations become the benchmark rhythm of the temperature field inside the cabinet through heat conduction and air convection. Even if other heat sources have stable self-heating, the temperature of their respective areas will be superimposed with the conducted heat from the dominant heat source, causing the temperature of these areas to change in accordance with the temperature fluctuations of the dominant heat source. The temperature change in areas far from the dominant heat source is determined by the diffusion process of the dominant heat source, exhibiting a similar trend of "rising and falling in tandem" with the dominant heat source, with only a decrease in amplitude due to distance and heat loss.

[0030] Because server racks typically contain multiple heat sources that may appear or disappear over time, uneven heating can be caused by several factors. These factors might include simultaneous small temperature fluctuations from multiple devices. While individual device temperature fluctuations may be small, their combined effect leads to a chaotic temperature distribution within the rack. Small temperature differences in different areas overlap, resulting in a lack of clear hotspots and an overall disordered temperature distribution within the rack, meaning the similarity of temperature change trends across different monitoring locations is low. Alternatively, a single high-heat-generating device might experience a large temperature fluctuation, with its temperature change trend showing a high degree of similarity to other locations. Since the uncertainty caused by the superposition of small temperature fluctuations from multiple heat sources makes temperature control more difficult, it is necessary to analyze whether the uneven heating within the rack is caused by temperature changes from multiple heat sources to determine the probability of such temperature changes.

[0031] Because the number and location of heat sources within the server rack change over time—for example, when multiple devices start up and enter working mode simultaneously, the number of heat sources increases rapidly in a short period; when devices are in sleep or powered off, the corresponding heat sources disappear, and this change is normal—the more drastic these changes are, the more severe the uneven heating within the server rack becomes, potentially introducing more sources of temperature fluctuation. Therefore, by combining the fluctuation degree of the dominant heat source quantity with the changes in the location of the dominant heat source within a local time period, and analyzing the changes in the location and quantity of the dominant heat source, we can obtain the overall degree of heat source variation.

[0032] The characteristics of multi-heat-source temperature variation are: no clear concentrated hot spot (i.e., few dominant heat sources), and chaotic overall temperature inside the cabinet (i.e., low similarity in temperature change trends across different monitoring locations). The overall heat source variation reflects the risk of introducing temperature fluctuations into the cabinet. By combining the correlation of temperature values ​​at different monitoring locations over historical analysis periods with the amount of dominant heat sources, the probability of meeting the characteristics of multi-heat-source temperature variation at the current moment is analyzed, thus obtaining the probability of multi-heat-source temperature variation.

[0033] In this embodiment of the invention, the 30 seconds preceding each moment are taken as its local time period.

[0034] Step S4: Based on the uneven heating of the cabinet and the possibility of temperature changes from multiple heat sources, perform real-time temperature control on the sheet metal cabinet.

[0035] Uneven heat distribution within the server rack reflects the degree of temperature inconsistency within the rack, while the potential for temperature variations from heat sources indicates the risk of multiple dominant heat sources within the rack. Combining both factors to control the rack's temperature can prevent uneven temperature distribution from escalating into overall high temperatures and improve the effectiveness of temperature control.

[0036] Preferably, in some possible implementations of the embodiments of the present invention, the method for obtaining attention is described in [reference needed]. Figure 2 The diagram illustrates a flowchart of a method for obtaining attention according to an embodiment of the present invention, the method comprising: Step S210: Based on the extreme value fluctuation characteristics and change frequency of the temperature value of each monitoring location during the historical analysis period at the current moment, obtain the attention level of each monitoring location at the current moment.

[0037] Preferably, in some possible implementations of the embodiments of the present invention, the method for obtaining attention includes: arranging the temperature values ​​of all times within the historical analysis period of each monitoring location at the current time in chronological order to obtain a temperature sequence; extracting global extreme values ​​and local extreme values ​​from the temperature sequence, and recording them as characteristic extreme values; for each characteristic extreme value, calculating the ratio of the absolute difference between each pair of adjacent characteristic extreme values ​​to the average of the time intervals between the corresponding times of the two characteristic extreme values ​​and the current time, and recording it as the fluctuation impact degree; averaging the fluctuation impact degrees of all pairs of adjacent characteristic extreme values ​​to obtain the overall impact degree; using the average of the overall impact degree of the global extreme values ​​and the overall impact degree of the local extreme values ​​as the temperature change performance factor of each monitoring location at the current time; recording the proportion of the number of local extreme values ​​in the temperature sequence as the temperature change frequency; recording the proportion of the number of temperature values ​​in the temperature sequence that are greater than a preset temperature threshold as the high temperature intensity; and obtaining the attention degree of each monitoring location at the current time based on the temperature change performance factor, the temperature change frequency, and the high temperature intensity.

[0038] It should be noted that global extreme values ​​reflect extreme temperature fluctuations within the historical analysis period, while local extreme values ​​reflect short-term temperature fluctuations. A greater fluctuation impact indicates a larger amplitude of fluctuations in the two extreme values ​​and a more direct impact on current control, thus a more significant impact of the fluctuation on current heat dissipation needs. The overall impact reflects the overall intensity and urgency of temperature changes at the monitoring location within the historical analysis period. A greater overall impact indicates that the monitoring location experiences long-term, severe, and nearby temperature fluctuations, requiring more attention to the equipment at that location. By calculating the average of the overall impact of global and local extreme values, and integrating global and local temperature fluctuation information, a temperature change performance factor is obtained. A higher temperature change frequency indicates more frequent temperature fluctuations within the historical analysis period, possibly caused by high-frequency equipment start-ups and shutdowns, computing power fluctuations, or system instability. To ensure equipment operational reliability, the equipment at the monitoring location requires more attention. Equipment with higher temperatures may be under high load and requires more efficient heat dissipation. Temperatures exceeding a preset temperature threshold can be considered high temperatures. A greater intensity of high temperatures indicates that the monitoring location is more likely to be in a high-temperature state for a long period during the historical analysis period, potentially leading to performance degradation or damage to the equipment at that location, thus requiring more attention to heat dissipation.

[0039] Therefore, the temperature change performance factor, temperature change frequency, and high temperature intensity are all positively correlated with the level of attention. In this embodiment of the invention, the temperature change performance factor is normalized, and the normalized result of the temperature change performance factor at each detection location at the current moment, along with the arithmetic mean of the temperature change frequency and the high temperature intensity, is used as the level of attention.

[0040] In this embodiment of the invention, based on the temperature change performance factor of each time step before each time step, the temperature change performance factor of each time step is normalized using the minimax normalization method, or other normalization methods such as the Z-score normalization method, which are not limited here.

[0041] In this embodiment of the invention, global extrema refer to the maximum and minimum temperature values ​​in the temperature sequence, while local extrema refer to the maximum and minimum temperature values ​​in the temperature sequence obtained through an extremum detection algorithm, similar to the values ​​corresponding to extreme points on a curve. The extremum detection algorithm can be the finite difference method or the first derivative method, etc. It is important to note that when local extrema in the temperature sequence refer to global extrema, the temperature change inside the cabinet is mainly dominated by slow background heat changes or the conduction effect of stable heat sources, rather than being driven by frequent local thermal events. This means that this monitoring location can be assigned less attention, and in this embodiment, the overall influence of the local extrema is directly set to 0.

[0042] In one implementation of this invention, the preset temperature threshold is the upper limit of the safe operating temperature of the cabinet design standard.

[0043] Step S220: The ratio of the absolute difference between the temperature values ​​of any two monitoring locations at the current moment to the distance between the two monitoring locations is taken as the local heat unevenness of the two monitoring locations at the current moment; based on the average attention of any two monitoring locations at the current moment, the local heat unevenness between all monitoring locations is weighted and summed to obtain the rack heat unevenness at the current moment.

[0044] In one specific implementation of this invention, the formula for uneven heat distribution in the server rack is expressed as: In the formula, G represents the heat unevenness of the rack at the current moment; N represents the total number of monitoring locations; This represents the average level of attention given to the nth and mth monitoring locations at the current moment. This represents the absolute difference between the temperature values ​​at the nth monitoring location and the mth monitoring location at the current moment. This represents the distance between the nth monitoring location and the mth monitoring location. This represents the local heating unevenness between the nth and mth monitoring locations at the current moment.

[0045] It should be noted that localized heating unevenness is similar to a temperature gradient in thermodynamics, reflecting the rate of temperature change between two monitoring locations; if localized heating unevenness... The larger the value, the more significant the temperature difference per unit distance between the nth and mth monitoring locations, and the more severe the uneven heating between the two corresponding monitoring locations. Since the value of the attention threshold ranges from 0 to 1, then... As The weights of the average attention at the nth and mth monitoring positions. The larger the value, the greater the contribution of the uneven heating between the two monitoring locations to the overall temperature unevenness inside the cabinet, making the cabinet heating unevenness, which reflects the overall unevenness of the temperature distribution inside the entire cabinet, more accurate.

[0046] Preferably, in some possible implementations of the embodiments of the present invention, the method for obtaining the dominant heat source quantity includes: performing negative correlation mapping on the correlation coefficient of the temperature sequences of any two monitoring locations at each time to obtain the temperature distance between the two monitoring locations at each time; clustering all monitoring locations based on the temperature distance, and recording the number of clusters obtained at each time as the dominant heat source quantity at each time.

[0047] It should be noted that if the temperature distance between two monitoring locations is smaller and the temperature change trends of the temperature sequences of the two monitoring locations are more similar, then the two monitoring locations are more likely to be affected by the same dominant heat source, that is, more likely to be classified into the same cluster.

[0048] In one implementation of this invention, the K-means clustering algorithm is used to cluster the monitoring locations, wherein the K value is determined by the elbow method.

[0049] In one implementation of this invention, the correlation coefficient is the Pearson correlation coefficient, but it can also be the Spearman correlation coefficient or the canonical correlation coefficient, etc.

[0050] In this embodiment of the invention, since the correlation coefficient ranges from -1 to 1, the difference between the constant 1 and the correlation coefficient is used to achieve a negative correlation mapping of the correlation coefficient. The constant 1 can also be replaced by other constants greater than 1.

[0051] Preferably, in some possible implementations of the embodiments of the present invention, the method for obtaining the comprehensive change degree of heat sources includes: for any two adjacent time points, calculating the ratio of the number of monitoring locations commonly contained in each cluster at the previous time point and each cluster at the next time point to the total number of all non-overlapping monitoring locations in the corresponding two clusters, and recording it as the position similarity of the corresponding clusters; selecting the maximum value among the position similarities of each cluster at the previous time point and all clusters at the next time point, averaging the maximum values ​​corresponding to all clusters at the previous time point to obtain the local position invariance of two adjacent time points; calculating the mean of the local position invariance of every two adjacent time points within the local time period of the current time point, as the heat source position invariance of the current time point; recording the variance of the dominant heat source quantity at all time points within the local time period of the current time point as the heat source quantity change degree; and obtaining the comprehensive change degree of heat sources at the current time point based on the heat source position invariance and the heat source quantity change degree.

[0052] It should be noted that the clustering results of the monitoring locations may differ at different times. If the maximum value of the similarity score between each cluster at a previous time and the cluster at a later time is larger, it indicates a greater overlap of monitoring locations within the clusters corresponding to the maximum value. This means the location of the dominant heat source in the two clusters is more stable between adjacent times, and the thermal influence domain of the dominant heat source has not changed drastically. When the local location invariance is greater, the influence area of ​​the dominant heat source between adjacent times is more stable, and the probability of the dominant heat source location remaining unchanged between adjacent times is greater. Heat source location invariance reflects the overall invariance of the dominant heat source location within a local time period. A smaller heat source location invariance indicates rapid changes in the heat source layout, resulting in an extremely unstable thermal environment inside the cabinet and a high risk of uneven heating. A larger change in the number of heat sources indicates more drastic changes in the amount of dominant heat sources within a local time period, more frequent changes in the equipment load within the cabinet, and a greater likelihood of uneven heating within the cabinet. Therefore, the invariance of heat source location is negatively correlated with the overall change in heat source quantity, while the change in the number of heat sources is positively correlated with the overall change in heat source quantity. In this embodiment of the invention, the change in the number of heat sources is normalized by weighting and summing the difference between the constant 1 and the invariance of heat source location at the current moment and the normalized result of the change in the number of heat sources, to obtain the overall change in heat source quantity at the current moment.

[0053] In this embodiment of the invention, based on the change in the number of heat sources at previous times, the change in the number of heat sources at the current time is normalized using the min-max normalization method. Alternatively, Z-score normalization or other normalization methods can be used, and no limitation is made here. From the historical operating data of the cabinet, the invariance of heat source location and the change in the number of heat sources under typical operating conditions are extracted. Key indicators recording the actual control difficulty during the same period, such as the variance of fan speed fluctuations, are used as training targets. Through regression analysis, machine learning, and other algorithms, the optimal weighting coefficients for the two indicators—the difference between the weighted summation time constant 1 and the invariance of heat source location at the current time, and the normalized result of the change in the number of heat sources—are fitted.

[0054] Preferably, in some possible implementations of the embodiments of the present invention, the method for obtaining the probability of temperature change of multiple heat sources includes: calculating the ratio of the amount of the dominant heat source at the current moment to the total number of monitoring locations, as the heat source distribution dispersion; taking the average of the correlation coefficients of the temperature sequences of every two monitoring locations at the current moment as the neighborhood temperature similarity; and obtaining the probability of temperature change of multiple heat sources at the current moment based on the comprehensive change of heat sources, the heat source distribution dispersion, and the neighborhood temperature similarity.

[0055] It should be noted that the greater the overall variability of heat sources, the more drastic the changes in the location and number of dominant heat sources in the historical time period prior to the current moment, i.e., the more frequent the heat source changes (such as frequent equipment start-ups and shutdowns), the more temperature fluctuation sources the cabinet may introduce, and the greater the likelihood of multi-heat source temperature changes in the cabinet at the current moment. A greater dispersion of heat source distribution means that the cabinet has more dominant heat sources at the current moment, i.e., a lower probability of concentrated hotspots. A greater similarity in neighboring temperatures indicates that the temperature change trends of different monitoring locations are more similar during the historical analysis period at the current moment, resulting in fewer dominant heat sources in the cabinet; conversely, a smaller similarity indicates more dominant heat sources in the cabinet. Therefore, both the overall variability of heat sources and the dispersion of heat source distribution are positively correlated with the likelihood of multi-heat source temperature changes, while the similarity of neighboring temperatures is negatively correlated with the likelihood of multi-heat source temperature changes.

[0056] In this embodiment of the invention, the neighborhood temperature similarity is negatively correlated and normalized. The arithmetic mean of the processed neighborhood temperature similarity, the comprehensive change degree of heat sources, and the dispersion degree of heat source distribution at the current moment is calculated as the probability of multi-heat source temperature change at the current moment. Since the neighborhood temperature similarity ranges from -1 to 1, half the difference between the constant 1 and the neighborhood temperature similarity is calculated to achieve negative correlation and normalization of the neighborhood temperature similarity.

[0057] Preferably, in some possible implementations of the embodiments of the present invention, the temperature control method includes: normalizing the product of the uneven heating of the cabinet and the probability of temperature change of multiple heat sources, and using the sum of the normalization result and the constant 1 as the temperature adjustment weight at the current moment; the sheet metal cabinet includes heat dissipation equipment; and using a fuzzy control algorithm to obtain the operating speed of the heat dissipation equipment at the current moment based on the temperature values ​​of all monitoring locations at the current moment and the temperature adjustment weight.

[0058] It's important to note that the greater the unevenness of heat generation within the server rack, the more significant the temperature distribution imbalance inside the rack. To eliminate this unevenness, the control system must respond more forcefully, i.e., increase fan speed and enhance heat dissipation capacity, thus placing a greater emphasis on temperature adjustment. Conversely, a higher probability of temperature fluctuations from multiple heat sources means more dominant heat sources exist within the rack at any given moment. The start-up, shutdown, and power fluctuations of these heat sources lead to diverse temperature change patterns, increasing the risk of uneven temperature distribution. In such cases, increasing fan speed is necessary to suppress temperature fluctuations, further emphasizing temperature adjustment to prevent localized overheating from spreading to overall high temperatures. Therefore, server rack heat unevenness is positively correlated with both the probability of temperature fluctuations from multiple heat sources and the emphasis on temperature adjustment.

[0059] It should be noted that the temperature adjustment weight is set to 1 for each moment during the first 10 minutes of temperature control of the sheet metal cabinet, and the temperature adjustment weight is obtained using the above method for all subsequent moments.

[0060] In this embodiment of the invention, the temperature values ​​of all monitoring locations of the sheet metal cabinet at the current moment, along with the temperature adjustment weights at the current moment, are input into a fuzzy control algorithm. A membership function is used to convert the input into a fuzzy set. A rule base is constructed based on expert experience, and a production rule format is used to map the input fuzzy set to the output fuzzy set. The temperature control weights are used as condition variables in the rules, combined with the temperature data, to determine the aggressiveness of the fan speed. The output fuzzy sets of all rules, i.e., the fuzzy values ​​of the fan speed, are merged into a single overall output fuzzy set and converted into the fan speed value at the current moment. In this solution, the heat dissipation device refers to the fan, and the fan speed value refers to the operating speed of the aforementioned heat dissipation device.

[0061] This invention is now complete.

[0062] Example 2: This invention also proposes a computer device for temperature control of a sheet metal cabinet with an intelligent temperature control system; please refer to [link to relevant documentation]. Figure 3 The computer device includes a memory 501, a processor 502, and a computer program 503 stored in the memory 501 and running on the processor 502. When the processor 502 executes the computer program 503, the computer device can execute any of the temperature control methods for sheet metal cabinets with intelligent temperature control systems described above.

[0063] Furthermore, this application also protects an apparatus that may include a memory and a processor, wherein the memory stores executable program code, and the processor is used to call and execute the executable program code to perform a temperature control method for a sheet metal cabinet with an intelligent temperature control system provided in this application.

[0064] This embodiment can divide the device into functional modules based on the above method example. For example, each module can correspond to a separate function, or two or more functions can be integrated into one processing module. The integrated module can be implemented in hardware. It should be noted that the module division in this embodiment is illustrative and only represents one logical functional division. In actual implementation, there may be other division methods.

[0065] When each module is divided according to its function, the device may also include a communication module, a signal analysis module, a complexity analysis module, and a positioning module. It should be noted that all relevant content of each step involved in the above method embodiments can be referenced from the functional descriptions of the corresponding functional modules, and will not be repeated here.

[0066] It should be understood that the device provided in this embodiment is used to execute the temperature control method of a sheet metal cabinet with an intelligent temperature control system described above, and therefore can achieve the same effect as the above implementation method.

[0067] When using integrated units, the device may include a processing module and a storage module. When applied to a workpiece, the processing module can be used to control and manage the workpiece's operations. The storage module can be used to support the execution of program code by the workpiece.

[0068] The processing module may be a processor or a controller, which can implement or execute various exemplary logic blocks, modules, and circuits contained in conjunction with the disclosure of this application. The processor may also be a combination of functions that implement computing capabilities, such as a combination of one or more microprocessors, a combination of digital signal processing (DSP) and a microprocessor, etc., and the storage module may be a memory.

[0069] Example 3: This embodiment also provides a computer-readable storage medium storing computer program code. When the computer program code is run on a computer, the computer executes the above-mentioned method steps to implement the temperature control method for a sheet metal cabinet with an intelligent temperature control system provided in the above embodiment.

[0070] Example 4: This embodiment also provides a computer program product. When the computer program product is run on a computer, it causes the computer to perform the above-mentioned related steps to realize the temperature control method of a sheet metal cabinet with an intelligent temperature control system provided in the above embodiment.

[0071] In this embodiment, the device, computer-readable storage medium, computer program product, or chip are all used to execute the corresponding methods provided above. Therefore, the beneficial effects they can achieve can be referred to the beneficial effects in the corresponding methods provided above, and will not be repeated here.

[0072] In the embodiments provided in this application, it should be understood that the disclosed apparatus and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of modules or units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another device, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0073] It should be noted that the order of the above embodiments of the present invention is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0074] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

Claims

1. A method for temperature control of a sheet metal cabinet with an intelligent temperature control system, characterized in that, The method includes: Real-time acquisition of temperature values ​​at each monitoring location within the sheet metal cabinet at every moment; Based on the extreme fluctuation characteristics and change frequency of temperature values ​​at each monitoring location during the historical analysis period at the current moment, the temperature value differences at different monitoring locations at the current moment are adjusted to obtain the cabinet heating unevenness at the current moment. Obtain the dominant heat source quantity at each moment; based on the fluctuation degree of the dominant heat source quantity within a local time period at each moment and the change of the dominant heat source position at adjacent moments, obtain the comprehensive heat source change degree at the current moment; based on the correlation degree of temperature values ​​of different monitoring locations within the historical analysis period at the current moment, the comprehensive heat source change degree and the dominant heat source quantity, obtain the multi-heat source temperature change probability at the current moment. Based on the uneven heating of the cabinet and the potential temperature change of the multiple heat sources, the sheet metal cabinet is subjected to real-time temperature control.

2. The temperature control method for a sheet metal cabinet with an intelligent temperature control system according to claim 1, characterized in that, The process of obtaining the current rack heat unevenness includes: Based on the extreme value fluctuation characteristics and change frequency of the temperature value of each monitoring location within the historical analysis period at the current moment, the attention level of each monitoring location at the current moment is obtained; The ratio of the absolute difference between the temperature values ​​of any two monitoring locations at the current moment to the distance between the two monitoring locations is taken as the local heating non-uniformity of the two monitoring locations at the current moment. Based on the average of the attention levels at any two monitoring locations at the current moment, the local heat unevenness between all monitoring locations is weighted and summed to obtain the rack heat unevenness at the current moment.

3. The temperature control method for a sheet metal cabinet with an intelligent temperature control system according to claim 2, characterized in that, The process of obtaining the attention level of each monitoring location at the current moment includes: The temperature values ​​of each monitoring location at all times within the historical analysis period at the current moment are arranged in chronological order to obtain a temperature sequence; global extrema and local extrema are extracted from the temperature sequence and recorded as characteristic extrema. For each eigenvalue, calculate the ratio of the absolute difference between any two adjacent eigenvalues ​​to the mean of the time intervals between the corresponding times of the two eigenvalues ​​and the current time, and denot it as the fluctuation impact degree; average the fluctuation impact degrees of all adjacent pairs of eigenvalues ​​to obtain the overall impact degree; The average of the global extreme value and the local extreme value is used as the temperature change performance factor of each monitoring location at the current moment. The proportion of local extrema in the temperature sequence is denoted as the temperature change frequency. The percentage of temperature values ​​in the temperature sequence that are greater than a preset temperature threshold is denoted as the high temperature intensity. Based on the temperature change performance factor, the temperature change frequency, and the high temperature intensity, the attention level of each monitoring location at the current moment is obtained.

4. The temperature control method for a sheet metal cabinet with an intelligent temperature control system according to claim 3, characterized in that, The acquisition of the dominant heat source quantity at each moment includes: By performing a negative correlation mapping on the correlation coefficients of the temperature sequences of any two monitoring locations at each time step, the temperature distance between the two monitoring locations at each time step can be obtained. Based on the temperature distance, all monitoring locations are clustered, and the number of clusters at each time moment is recorded as the dominant heat source at each time moment.

5. The temperature control method for a sheet metal cabinet with an intelligent temperature control system according to claim 1, characterized in that, The process of obtaining the comprehensive change rate of the heat source at the current moment includes: For any two adjacent time points, calculate the ratio of the number of monitoring locations commonly contained in each cluster at the previous time point and each cluster at the next time point to the total number of all non-overlapping monitoring locations in the two corresponding clusters. This ratio is denoted as the location similarity of the corresponding clusters. The maximum value of the positional similarity between each cluster at the previous time step and all clusters at the next time step is selected, and the average of the maximum values ​​corresponding to all clusters at the previous time step is calculated to obtain the local positional invariance between two adjacent time steps. Calculate the average of the local position invariance between every two adjacent time intervals within the current time interval, and use it as the heat source position invariance at the current time. The variance of the dominant heat source quantity at all times within a local time period at the current time is denoted as the heat source quantity variation degree. Based on the invariance of the heat source location and the change in the number of heat sources, the overall change in the heat sources at the current moment is obtained.

6. The temperature control method for a sheet metal cabinet with an intelligent temperature control system according to claim 1, characterized in that, The process of obtaining the probability of temperature change from multiple heat sources at the current moment includes: Calculate the ratio of the dominant heat source quantity to the total number of monitored locations at the current moment, and use it as the heat source distribution dispersion. The average correlation coefficient of the temperature sequence of every two monitoring locations at the current time is used as the neighborhood temperature similarity. Based on the comprehensive change degree of the heat source, the dispersion degree of the heat source distribution, and the similarity of the neighborhood temperature, the probability of temperature change of multiple heat sources at the current moment is obtained.

7. The temperature control method for a sheet metal cabinet with an intelligent temperature control system according to claim 1, characterized in that, The real-time temperature control of the sheet metal cabinet based on the uneven heating of the cabinet and the probability of temperature changes from multiple heat sources includes: The product of the uneven heating of the cabinet and the probability of temperature change of the multiple heat sources is normalized, and the sum of the normalization result and the constant 1 is used as the temperature adjustment weight at the current moment. The sheet metal cabinet includes heat dissipation equipment; based on the temperature values ​​of all monitoring locations at the current moment and the temperature adjustment weights, the operating speed of the heat dissipation equipment at the current moment is obtained using a fuzzy control algorithm.

8. The temperature control method for a sheet metal cabinet with an intelligent temperature control system according to claim 6, characterized in that, The overall variability of the heat source and the dispersion of the heat source distribution are both positively correlated with the probability of temperature change of the multiple heat sources, while the neighborhood temperature similarity is negatively correlated with the probability of temperature change of the multiple heat sources.

9. A temperature control method for a sheet metal cabinet with an intelligent temperature control system according to claim 4, characterized in that, The K-means clustering algorithm was used to cluster all monitoring locations.

10. A temperature control method for a sheet metal cabinet with an intelligent temperature control system according to claim 4 or 6, characterized in that, The correlation coefficient mentioned is the Pearson correlation coefficient.

Images