A method for strengthening the internal temperature control of a computer

Abstract

The present application relates to the field of temperature control technology, and specifically to a method for controlling the internal temperature of a reinforced computer, the method comprising: collecting temperature data and utilization data of the reinforced computer CPU at each collection moment; dividing each preset time length into a temperature monitoring interval, and determining the abnormal degree of temperature change at each collection moment in each temperature monitoring interval; classifying all collection moments in each temperature monitoring interval, and determining the significant degree of temperature transient at each collection moment in each category based on the difference in the abnormal degree of temperature change between each collection moment in each category and all other collection moments, as well as the total number of collection moments in each category; obtaining the load change moment in each temperature monitoring interval; determining the temperature stability and temperature adjustment factor of each temperature monitoring interval; and controlling the temperature of the reinforced computer CPU in the next temperature monitoring interval. The present application aims to improve the accuracy and timeliness of the internal temperature control of the reinforced computer.

Description

Technical Field

[0001] The present application relates to the technical field of temperature control, and in particular to a method for strengthening the internal temperature control of a computer. Background Art

[0002] Reinforced computers, also known as computers resistant to harsh environments, are computers that take corresponding assurance measures for various factors that affect computer performance, such as system structure, electrical characteristics, and mechanical physical structure, in order to adapt to various harsh environments during the computer design process. They have strong environmental adaptability, high reliability, high maintainability, and strong real-time processing capabilities. The temperature control inside the reinforced computer is crucial to hardware performance assurance, extending equipment life, and adapting to extreme environments. It is one of the key factors to ensure the stable operation of the reinforced computer.

[0003] However, the existing temperature control methods do not take into account the fact that environmental interference may cause inaccurate temperature data collected by the sensor during the temperature control process, and do not consider the change trend of the load, which makes it difficult to accurately evaluate the heat dissipation inside the computer, thereby affecting the accuracy and timeliness of the temperature control inside the reinforced computer. Summary of the invention

[0004] In view of the above, it is necessary to provide a method for strengthening the internal temperature control of a computer, which improves the accuracy and timeliness of the internal temperature control of the computer compared with the traditional method for strengthening the internal temperature control of the computer.

[0005] A method for strengthening the internal temperature control of a computer in the present application adopts the following technical solution:

[0006] An embodiment of the present application provides a method for strengthening the internal temperature control of a computer, the method comprising the following steps:

[0007] Collect the temperature data and utilization rate data of the CPU of the reinforcement computer at each collection time;

[0008] Each preset time length is divided into a temperature monitoring interval, and the abnormal degree of temperature change at each collection moment in each temperature monitoring interval is determined based on the difference in temperature data between each collection moment and its adjacent collection moment in each temperature monitoring interval, and the difference in the abnormal degree of temperature data between each collection moment and all other collection moments;

[0009] Based on the difference in the degree of abnormality of temperature data between all acquisition moments in each temperature monitoring interval, all acquisition moments in each temperature monitoring interval are classified, and based on the difference in the degree of abnormality of temperature change between each acquisition moment in each category and all other acquisition moments, as well as the total number of acquisition moments in each category, the significance of temperature transient at each acquisition moment in each category is determined;

[0010] Based on the distribution of the significance of all temperature transients in each temperature monitoring interval, the load change moment in each temperature monitoring interval is obtained; based on the similarity between the temperature data and the utilization data at all the load change moments in each temperature monitoring interval, and the degree of fluctuation of the temperature data at all the load change moments, the temperature stability of each temperature monitoring interval is determined;

[0011] Determine the temperature adjustment factor of each temperature monitoring interval in combination with the temperature stability and the change trend of the utilization data at all the load change moments in each temperature monitoring interval;

[0012] Based on the temperature adjustment factor of the current temperature monitoring interval, the temperature data at all acquisition moments and the utilization data, the temperature of the CPU of the reinforcement computer in the next temperature monitoring interval is controlled.

[0013] In one embodiment, the process of determining the degree of abnormal temperature change is as follows:

[0014] Anomaly detection algorithm is used to obtain the local outlier factor of temperature data at each collection time in each temperature monitoring interval;

[0015] The degree of abnormality of the temperature data at each acquisition moment is obtained by the local outlier factor;

[0016] The expression of the abnormal degree of temperature change at each acquisition time in any temperature monitoring interval is:

[0017] ; In the formula, is the abnormal degree of temperature change at the i-th acquisition moment; is the absolute value of the difference in temperature data between the ith acquisition moment and its adjacent previous acquisition moment; calculate the average value of the local outlier factor of all acquisition moments in any temperature monitoring interval, is the absolute value of the difference between the local outlier factor and the average value at the i-th acquisition moment.

[0018] In one embodiment, the method for classifying all acquisition moments in each temperature monitoring interval is:

[0019] A clustering algorithm is used to obtain clusters of all acquisition moments in each temperature monitoring interval. The metric distance between any two acquisition moments in the clustering process is the difference in the local outlier factors between the any two acquisition moments.

[0020] In one embodiment, the process of determining the significance of the temperature transient is as follows:

[0021] Calculate the sum of the absolute values ​​of the differences in the degree of temperature change anomaly between each acquisition time in each cluster and all other acquisition times;

[0022] Based on the sum and the total number of acquisition moments in each cluster, the significance of temperature transient at each acquisition moment in each cluster is obtained.

[0023] In one embodiment, the temperature transient significance level is the ratio of the sum value to the total number of acquisition moments in each cluster.

[0024] In one embodiment, the method for acquiring the load change moment is: performing threshold segmentation on the temperature transient significance of all acquisition moments in each temperature monitoring interval, and taking the acquisition moment when the temperature transient significance is less than the segmentation threshold as the load change moment.

[0025] In one embodiment, the process of determining the temperature stability is:

[0026] The temperature data and utilization data of all load change moments in each temperature monitoring interval are arranged in time sequence to form a load temperature sequence and a CPU utilization sequence of each temperature monitoring interval;

[0027] The expression of temperature stability in each temperature monitoring interval is:

[0028] ; In the formula, is the temperature stability of the pth temperature monitoring interval; , are the load temperature sequence and CPU utilization sequence of the pth temperature monitoring interval respectively; Cos() is the cosine similarity function; ent() is the Shannon entropy function; μ is a preset value greater than 0.

[0029] In one embodiment, the temperature adjustment factor is determined as follows:

[0030] The time series decomposition algorithm is used to obtain the trend item of each CPU utilization sequence, and the value of the trend item at each load change moment is used as the trend item intensity at each load change moment;

[0031] Based on the temperature stability and the distribution of the intensity of the trend item at all load change moments in each temperature monitoring interval, the temperature adjustment factor of each temperature monitoring interval is determined.

[0032] In one embodiment, the expression of the temperature adjustment factor is:

[0033] ; In the formula, is the temperature adjustment factor of the pth temperature monitoring interval; It is the cumulative result of the trend item strength at all load change moments in the pth temperature monitoring interval; is the temperature stability of the pth temperature monitoring interval; norm() is the normalization function; ε is the first coefficient preset to be greater than 0.

[0034] In one embodiment, the process of controlling the temperature of the CPU of the hardened computer in the next temperature monitoring interval is:

[0035] If the temperature adjustment factor of the current temperature monitoring interval is less than the preset temperature adjustment threshold, it is determined that the internal temperature of the reinforced computer is normal at this time, and the rotation speed of the internal cooling fan of the reinforced computer is not adjusted; otherwise, it is determined that the internal temperature of the reinforced computer is abnormal at this time, and the rotation speed of the internal cooling fan of the reinforced computer is adjusted, and the temperature of the CPU of the reinforced computer in the next temperature monitoring interval is controlled according to the adjusted rotation speed;

[0036] The process of adjusting the speed of the cooling fan inside the hardened computer is as follows:

[0037] A neural network is used to obtain the temperature adjustment amount of the reinforced computer in the current temperature monitoring interval according to the temperature adjustment factor of the current temperature monitoring interval, the mean temperature data of all acquisition moments, and the mean utilization data;

[0038] A PID controller is used to adjust the rotation speed of the cooling fan inside the reinforced computer according to the temperature data mean value and the temperature adjustment amount.

[0039] This application has at least the following beneficial effects:

[0040] The present application determines the abnormal degree of temperature change at each acquisition moment according to the change characteristics of the temperature data within the temperature monitoring interval, determines the significant degree of temperature transient change at each acquisition moment according to the abnormal degree of temperature change and the characteristics of temperature transient change, comprehensively considers the abnormal temperature change conditions and temperature transient change conditions of key components inside the reinforced computer, and more accurately reflects the reasons for the temperature change inside the reinforced computer;

[0041] Furthermore, the load change moments are screened out, and the temperature adjustment factor is obtained through the temperature change stability of the reinforced computer at the load change moment and the load change trend, and the speed of the cooling fan inside the reinforced computer is controlled in combination with the temperature data and the CPU utilization data. On the basis of considering the correlation change between the temperature and the load of the reinforced computer, an analysis of the degree of load change trend is added. On the premise of determining that the internal temperature of the reinforced computer needs to be controlled, the speed of the cooling fan inside the reinforced computer is more accurately obtained, and then the temperature inside the reinforced computer is more accurately controlled, thereby improving the accuracy and timeliness of the internal temperature control of the reinforced computer. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present application or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0043] Figure 1 A flowchart of the steps of a method for strengthening the internal temperature control of a computer provided in this application;

[0044] Figure 2 This is a schematic diagram of the impact of harsh electromagnetic environment interference on temperature;

[0045] Figure 3 This is a schematic diagram showing the effect of CPU load changes on temperature;

[0046] Figure 4 It is a schematic diagram of the strength of the trend term;

[0047] Figure 5 Schematic diagram of the process of obtaining the temperature adjustment factor. DETAILED DESCRIPTION

[0048] In the description of the embodiments of the present application, words such as "exemplary", "or", "for example" and the like are used to indicate examples, illustrations or descriptions. Any embodiment or design described as "exemplary" or "for example" in the embodiments of the present application should not be interpreted as being more preferred or more advantageous than other embodiments or designs. Specifically, the use of words such as "exemplary", "or", "for example" and the like is intended to present related concepts in a concrete manner.

[0049] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art in the present application. The terms used in the present application are only for the purpose of describing specific embodiments and are not intended to limit the present application. It should be understood that, unless otherwise specified, " / " means or.

[0050] It should also be noted that the terms "first" and "second" in the present application are used to distinguish similar objects rather than to describe a specific order or sequence.

[0051] The following is a detailed description of a specific solution of a method for strengthening the internal temperature control of a computer provided by the present application in conjunction with the accompanying drawings.

[0052] An embodiment of the present application provides a method for strengthening the internal temperature control of a computer. Specifically, the following method for strengthening the internal temperature control of a computer is provided. Figure 1, the method comprises the following steps:

[0053] Step 1: Collect temperature data and utilization data of the CPU of the reinforcement computer at each collection time.

[0054] The temperature and utilization of the CPU of the hardened computer are collected in real time through the temperature sensor and the operating system task manager. The data collection time interval is T seconds. In order to avoid data loss during transmission due to interference from the external environment, the median filling method is used to fill the acquired temperature data and utilization data. The temperature data and utilization data are normalized by the Z-Score normalization method to reduce the impact of the dimension on subsequent analysis. Both the median filling method and the Z-Score normalization method are well-known technologies and will not be described in detail in this application.

[0055] In this embodiment, the value of T is 1. The value of T is preset manually and can be set by the implementer. This application does not impose any special restrictions.

[0056] Step 2: Divide each preset time length into a temperature monitoring interval, and determine the abnormal degree of temperature change at each collection moment in each temperature monitoring interval based on the difference in temperature data between each collection moment and its adjacent collection moment in each temperature monitoring interval, and the difference in the abnormal degree of temperature data between each collection moment and all other collection moments.

[0057] When the CPU of a reinforced computer processes different tasks, the load of the reinforced computer will change, and the power consumption of the reinforced computer will also change accordingly, resulting in temperature changes. When the reinforced computer faces interference from a harsh electromagnetic environment, the sensor may be interfered with, causing abnormal changes in the temperature data of the CPU of the reinforced computer.

[0058] Based on the above analysis, it is necessary to evaluate the reasons for the temperature changes inside the reinforced computer to avoid the situation where, when the temperature inside the reinforced computer is normal, the temperature sensor is interfered by the harsh external electromagnetic environment, resulting in abnormal temperature data collected, which in turn leads to the adoption of incorrect temperature control strategies and affects the operating performance of the reinforced computer. Generally, the characteristics of temperature changes caused by harsh electromagnetic environment interference and CPU load changes are different. Harsh electromagnetic environment interference causes temperature changes with a larger amplitude but a shorter duration, while CPU load changes cause temperature changes with a smaller amplitude and a longer duration. When the CPU load increases, the CPU needs to process more tasks, which usually leads to increased CPU utilization. A schematic diagram of the impact of harsh electromagnetic environment interference on temperature is shown below. Figure 2 The effect of CPU load change on temperature is shown in the figure below. Figure 3 shown.

[0059] Each V minutes is divided into a temperature monitoring interval, and the temperature data of all acquisition moments in each temperature monitoring interval is used as input. The LOF (Local Outlier Factor) outlier detection algorithm is used to obtain the local outlier factor of the temperature data at each acquisition moment. The larger the local outlier factor at a certain acquisition moment, the greater the abnormality of the temperature data at that acquisition moment. The reachable distance in the LOF outlier detection algorithm is obtained by cross-validation. The LOF outlier detection algorithm and the cross-validation method are both well-known technologies and will not be described in detail in this application.

[0060] In this embodiment, the value of V is 10. The value of V is preset manually and can be set by the implementer. This application does not impose any special restrictions.

[0061] Taking any temperature monitoring interval as an example, based on the difference in temperature data between each acquisition moment and its adjacent acquisition moment, and the difference in the local outlier factor between each acquisition moment and all acquisition moments, the abnormal degree of temperature change at each acquisition moment is determined, and the expression is:

[0062] ; In the formula, is the abnormal degree of temperature change at the i-th acquisition moment; is the absolute value of the difference between the temperature data at the ith acquisition moment and its adjacent previous acquisition moment; calculate the average value of the local outlier factor at all acquisition moments, It is the absolute value of the difference between the local outlier factor and the average value at the i-th collection moment; wherein, the temperature abnormality change intensity at the first collection moment in the temperature monitoring interval is the absolute value of the difference in temperature data between the first collection moment and the next adjacent collection moment.

[0063] It should be noted that: when the temperature change condition of the reinforcement computer CPU at the i-th collection moment is more abnormal, the difference in temperature data of the reinforcement computer between the i-th collection moment and its adjacent previous collection moment is greater, the difference in the local outlier factor between the i-th collection moment and all other collection moments is greater, then the absolute value of the difference in temperature data between the i-th collection moment and its adjacent previous collection moment is greater, the absolute value of the difference between the local outlier factor at the i-th collection moment and the average value is greater, then the degree of abnormal temperature change is greater.

[0064] Step 3, based on the difference in the degree of abnormality of temperature data between all collection moments in each temperature monitoring interval, classify all collection moments in each temperature monitoring interval, and determine the significance of temperature transients at each collection moment in each category based on the difference in the degree of abnormality of temperature changes between each collection moment in each category and all other collection moments, as well as the total number of collection moments in each category.

[0065] The abnormal degree of temperature change alone cannot reflect the significance of the instantaneous change of the temperature of the reinforced computer CPU at any collection moment. Therefore, it is necessary to perform cluster analysis on the collection moments within the temperature monitoring interval to quantify the continuity of the temperature change of the reinforced computer CPU.

[0066] All the acquisition moments in the temperature monitoring interval are used as input, and the K-Mediods clustering algorithm is used to obtain the clustering clusters of all the acquisition moments in the temperature monitoring interval, and the absolute value of the difference of the local outlier factor between any two acquisition moments is used as the metric distance between any two acquisition moments in the clustering process. The K-Mediods clustering algorithm is a well-known technology and will not be described in detail in this application.

[0067] In this embodiment, the value of the number of cluster centers K in the K-Mediods clustering algorithm is 20. The value of K is preset manually and can be set by the implementer. This application does not impose any special restrictions.

[0068] Based on the difference in the degree of abnormal temperature change between each acquisition moment in each cluster and all other acquisition moments, as well as the total number of acquisition moments in each cluster, the significance of temperature transient at each acquisition moment in each cluster is determined, and the expression is:

[0069] ; In the formula, is the significance of temperature transient at the jth acquisition moment in the kth cluster; is the sum of the absolute values ​​of the differences in the degree of temperature change anomaly between the jth acquisition time and all other acquisition times in the kth cluster; is the total number of collection moments in the kth cluster.

[0070] It should be noted that: when the possibility that the temperature change of the reinforced computer at the jth collection moment is caused by the interference of the harsh electromagnetic environment is greater, the instantaneous change degree of the temperature change condition of the reinforced computer at the jth collection moment is more significant, and the duration of the temperature change condition is shorter, then the sum of the absolute values ​​of the differences in the degree of abnormal temperature change between the jth collection moment and all other collection moments in the kth cluster is greater, and the total number of collection moments in the kth cluster is smaller, then the greater the significance of the temperature transient at the jth collection moment is.

[0071] For all the acquisition moments in the temperature monitoring interval, the temperature transient significance levels of all the acquisition moments in the temperature monitoring interval are calculated using the same calculation method as the temperature transient significance level at the ith acquisition moment in the kth cluster.

[0072] Step 4, based on the distribution of the significance of all temperature transients corresponding to each temperature monitoring interval, obtain the load change moment in each temperature monitoring interval; based on the similarity between the temperature data and utilization data at all the load change moments in each temperature monitoring interval, and the degree of fluctuation of the temperature data at all the load change moments, determine the temperature stability of each temperature monitoring interval.

[0073] The temperature transient significance of all acquisition moments within the temperature monitoring interval is used as the input of the Otsu threshold method, and the segmentation threshold is output, and the acquisition moment when the temperature transient significance is less than the segmentation threshold is used as the load change moment. The Otsu threshold method is a well-known technology and will not be described in detail in this application.

[0074] When the CPU load of a reinforced computer changes rapidly, the temperature inside the reinforced computer also changes accordingly. This may make it difficult to accurately evaluate the changing trend of the temperature of the reinforced computer, which puts the timeliness of the formulation of the temperature control strategy under test, thereby affecting the operating performance of the reinforced computer.

[0075] Based on the above analysis, the temperature data and utilization data of all load change moments in the temperature monitoring interval are arranged in time series to form the load temperature sequence and CPU utilization sequence of the temperature monitoring interval.

[0076] Based on the similarity between the load temperature sequence and the CPU utilization sequence in the temperature monitoring interval, and the degree of fluctuation of the temperature data in the load temperature sequence, the temperature stability of the temperature monitoring interval is determined, and the expression is:

[0077] ; In the formula, is the temperature stability of the pth temperature monitoring interval; , are respectively the load temperature sequence and the CPU utilization sequence of the pth temperature monitoring interval; Cos() is the cosine similarity function; ent() is the Shannon entropy function; μ is a value preset to be greater than 0, in order to avoid the denominator being 0, the value of μ is preset artificially and can be set by the implementer. In this embodiment, the value of μ is 0.1; wherein, the calculation of Shannon entropy is a well-known technology and will not be described in detail in this application.

[0078] It should be noted that: when the temperature change trend of the reinforced computer in the temperature monitoring interval is more stable, the correlation between the temperature change status at the load change time in the temperature monitoring interval and the CPU utilization change status is stronger, and the degree of confusion of the temperature data change caused by the load change of the reinforced computer in the temperature monitoring interval is less, then the cosine similarity between the load temperature sequence and the CPU utilization sequence in the temperature monitoring interval is greater, the Shannon entropy of the temperature data in the load temperature sequence is smaller, and the value of the temperature stability is greater.

[0079] Step 5: Determine the temperature adjustment factor of each temperature monitoring interval in combination with the temperature stability and the change trend of the utilization data at all the load change moments in each temperature monitoring interval.

[0080] Using only temperature stability as the basis for adjusting the internal temperature control of reinforced computers still has the following disadvantages: lack of analysis of load change trends of reinforced computers, resulting in difficulty in timely and correct assessment of whether the reinforced computer should be temperature controlled when the CPU load changes rapidly, affecting the temperature control effect of the reinforced computer.

[0081] Based on the above analysis, the CPU utilization sequence of the temperature monitoring interval is used as input, and the STL (Seasonal and Trend decomposition using Loess) time series decomposition algorithm is used to obtain the trend item of the CPU utilization sequence, and the value of the trend item at each load change moment is used as the trend item strength at each load change moment. The STL time series decomposition algorithm is a well-known technology and will not be described in detail in this application. The schematic diagram of the trend item strength is shown in FIG. Figure 4 As shown in the figure, the horizontal axis is time, the unit is s, and the vertical axis is the normalized value of the trend item strength.

[0082] Based on the temperature stability of the temperature monitoring interval and the distribution of the trend item intensity at all load change moments in the temperature monitoring interval, the temperature adjustment factor of the temperature monitoring interval is determined, and the expression is:

[0083] ; In the formula, is the temperature adjustment factor of the pth temperature monitoring interval; It is the cumulative result of the trend item strength at all load change moments in the pth temperature monitoring interval; is the temperature stability of the pth temperature monitoring interval; norm() is the normalization function; ε is the first coefficient preset to be greater than 0, the purpose of which is to avoid the denominator being 0. The value of ε is preset manually and can be set by the implementer. In this embodiment, the value of ε is 0.1.

[0084] It should be noted that: when the internal temperature condition of the reinforced computer is abnormal, the more temperature control is needed, the less significant the temperature change stability of the reinforced computer within the temperature monitoring interval, the more drastic the load change trend of the reinforced computer, the smaller the temperature stability value of the temperature monitoring interval, the greater the cumulative result of the trend item strength at all load change moments within the temperature monitoring interval, and the larger the value of the temperature adjustment factor.

[0085] For all temperature monitoring intervals, the temperature adjustment factor of each temperature monitoring interval is calculated in the same way as the temperature adjustment factor of the pth temperature monitoring interval. The schematic diagram of the temperature adjustment factor acquisition process is shown in Figure 5 shown.

[0086] Step 6: Based on the temperature adjustment factor of the current temperature monitoring interval, the temperature data at all acquisition moments, and the utilization data, the temperature of the CPU of the reinforcement computer in the next temperature monitoring interval is controlled.

[0087] Furthermore, a temperature adjustment threshold is set. If the temperature adjustment factor of the current temperature monitoring interval is less than the temperature adjustment threshold, it is determined that the internal temperature of the reinforced computer is normal at this time, and the normal operation of the reinforced computer is not affected, and the speed of the internal cooling fan of the reinforced computer is not adjusted; if the temperature adjustment factor of the current temperature monitoring interval is greater than or equal to the temperature adjustment threshold, it is determined that the internal temperature of the reinforced computer is abnormal at this time, and the speed of the internal cooling fan of the reinforced computer is adjusted. The temperature of the reinforced computer CPU in the next temperature monitoring interval is controlled by the adjusted speed of the internal cooling fan of the reinforced computer.

[0088] In this embodiment, the value of the temperature adjustment threshold is 0.7. The value of the temperature adjustment threshold is preset manually and can be set by the implementer. This application does not impose any special restrictions.

[0089] The process of adjusting the speed of the cooling fan inside the hardened computer is as follows:

[0090] The temperature adjustment factor of the current temperature monitoring interval, the average value of the temperature data at all acquisition moments in the current temperature monitoring interval, and the average value of the CPU utilization are used as the input of the recurrent neural network to obtain the temperature adjustment amount of the reinforced computer in the current temperature monitoring interval. The loss function of the recurrent neural network is the mean square error, and the optimization algorithm uses Adam.

[0091] The average value of the temperature data at all acquisition moments in the current temperature monitoring interval and the temperature adjustment amount are input into the PID controller. The PID controller calculates the difference between the average value of the temperature data and the temperature adjustment amount, outputs a control signal for the speed of the cooling fan inside the reinforced computer, and adjusts the speed of the cooling fan inside the reinforced computer. Both the recurrent neural network and the PID controller are well-known technologies and will not be described in detail in this application.

[0092] In summary, the present application determines the abnormal degree of temperature change at each acquisition moment according to the change characteristics of the temperature data within the temperature monitoring interval, determines the significant degree of temperature transient change at each acquisition moment according to the abnormal degree of temperature change and the temperature transient change characteristics, comprehensively considers the abnormal temperature change conditions and temperature transient change conditions of the key components inside the reinforced computer, and more accurately reflects the reasons for the temperature changes inside the reinforced computer;

[0093] Furthermore, the load change moments are screened out, and the temperature adjustment factor is obtained through the temperature change stability of the reinforced computer at the load change moment and the load change trend, and the speed of the cooling fan inside the reinforced computer is controlled in combination with the temperature data and the CPU utilization data. On the basis of considering the correlation change between the temperature and the load of the reinforced computer, an analysis of the degree of load change trend is added. On the premise of determining that the internal temperature of the reinforced computer needs to be controlled, the speed of the cooling fan inside the reinforced computer is more accurately obtained, and then the temperature inside the reinforced computer is more accurately controlled, thereby improving the accuracy and timeliness of the internal temperature control of the reinforced computer.

[0094] The flowchart and block diagram in the accompanying drawings show the possible architecture, function and operation of the system, method and computer program product according to the embodiment of the present disclosure. In this regard, each box in the flowchart or block diagram can represent a module, a program segment or a part of the code, and the module, the program segment or a part of the code contains one or more executable instructions for realizing the specified logical function. In some alternative implementations, the functions marked in the box can also occur in a different order from the order marked in the accompanying drawings. For example, two consecutive boxes can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, which can depend on the functions involved. In the description corresponding to the flowchart and the block diagram in the accompanying drawings, the operations or steps corresponding to different boxes can also occur in a different order from the order disclosed in the description, and sometimes there is no specific order between different operations or steps. For example, two consecutive operations or steps can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, which can depend on the functions involved. Each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, may be implemented by a dedicated hardware-based system that performs the specified functions or actions, or may be implemented by a combination of dedicated hardware and computer instructions.

[0095] It is obvious to those skilled in the art that the present application is not limited to the details of the above exemplary embodiments, and that the present application can be implemented in other specific forms without departing from the basic features of the present application. Therefore, no matter from which point of view, the above embodiments of the present application should be regarded as exemplary and non-restrictive.

Claims

1. A method for strengthening the internal temperature control of a computer, characterized in that: The method comprises the following steps: Collect the temperature data and utilization rate data of the CPU of the reinforcement computer at each collection time; Each preset time length is divided into a temperature monitoring interval, and the abnormal degree of temperature change at each collection moment in each temperature monitoring interval is determined based on the difference in temperature data between each collection moment and its adjacent collection moment in each temperature monitoring interval, and the difference in the abnormal degree of temperature data between each collection moment and all other collection moments; Based on the difference in the degree of abnormality of temperature data between all acquisition moments in each temperature monitoring interval, all acquisition moments in each temperature monitoring interval are classified, and based on the difference in the degree of abnormality of temperature change between each acquisition moment in each category and all other acquisition moments, as well as the total number of acquisition moments in each category, the significance of temperature transient at each acquisition moment in each category is determined; Based on the distribution of the significance of all temperature transients in each temperature monitoring interval, the load change moment in each temperature monitoring interval is obtained; based on the similarity between the temperature data and the utilization data at all the load change moments in each temperature monitoring interval, and the degree of fluctuation of the temperature data at all the load change moments, the temperature stability of each temperature monitoring interval is determined; Determine the temperature adjustment factor of each temperature monitoring interval in combination with the temperature stability and the change trend of the utilization data at all the load change moments in each temperature monitoring interval; Based on the temperature adjustment factor of the current temperature monitoring interval, the temperature data and utilization data at all acquisition times, the temperature of the CPU of the reinforcement computer is controlled in the next temperature monitoring interval; The process of determining the abnormal degree of temperature change is as follows: Anomaly detection algorithm is used to obtain the local outlier factor of temperature data at each collection time in each temperature monitoring interval; The degree of abnormality of the temperature data at each acquisition moment is obtained by the local outlier factor; The expression of the abnormal degree of temperature change at each acquisition time in any temperature monitoring interval is: ; In the formula, is the abnormal degree of temperature change at the i-th acquisition moment; is the absolute value of the difference in temperature data between the ith acquisition moment and its adjacent previous acquisition moment; calculate the average value of the local outlier factor of all acquisition moments in any temperature monitoring interval, is the absolute value of the difference between the local outlier factor and the average value at the i-th acquisition moment.

2. A method for strengthening the internal temperature control of a computer as claimed in claim 1, characterized in that: The method for classifying all acquisition moments in each temperature monitoring interval is as follows: A clustering algorithm is used to obtain clusters of all acquisition moments in each temperature monitoring interval. The metric distance between any two acquisition moments in the clustering process is the difference in the local outlier factors between the any two acquisition moments.

3. A method for strengthening the internal temperature control of a computer as claimed in claim 2, characterized in that: The process of determining the significance of the temperature transient is as follows: Calculate the sum of the absolute values ​​of the differences in the degree of temperature change anomaly between each acquisition time in each cluster and all other acquisition times; Based on the sum and the total number of acquisition moments in each cluster, the significance of temperature transient at each acquisition moment in each cluster is obtained.

4. A method for strengthening the internal temperature control of a computer as claimed in claim 3, characterized in that: The temperature transient significance level is the ratio of the sum value to the total number of acquisition moments in each cluster.

5. A method for strengthening the internal temperature control of a computer as claimed in claim 1, characterized in that: The method for acquiring the load change moment is: performing threshold segmentation on the temperature transient significance of all acquisition moments in each temperature monitoring interval, and taking the acquisition moment when the temperature transient significance is less than the segmentation threshold as the load change moment.

6. A method for strengthening the internal temperature control of a computer as claimed in claim 1, characterized in that: The process of determining the temperature stability is as follows: The temperature data and utilization data of all load change moments in each temperature monitoring interval are arranged in time sequence to form a load temperature sequence and a CPU utilization sequence of each temperature monitoring interval; The expression of temperature stability in each temperature monitoring interval is: ; In the formula, is the temperature stability of the pth temperature monitoring interval; , are the load temperature sequence and CPU utilization sequence of the pth temperature monitoring interval respectively; Cos() is the cosine similarity function; ent() is the Shannon entropy function; μ is a preset value greater than 0.

7. A method for strengthening the internal temperature control of a computer as claimed in claim 6, characterized in that: The determination process of the temperature adjustment factor is: The time series decomposition algorithm is used to obtain the trend item of each CPU utilization sequence, and the value of the trend item at each load change moment is used as the trend item intensity at each load change moment; The temperature adjustment factor of each temperature monitoring interval is determined based on the temperature stability and the distribution of the intensity of the trend item at all load change moments in each temperature monitoring interval.

8. A method for strengthening the internal temperature control of a computer as claimed in claim 7, characterized in that: The expression of the temperature adjustment factor is: ; In the formula, is the temperature adjustment factor of the pth temperature monitoring interval; It is the cumulative result of the trend item strength at all load change moments in the pth temperature monitoring interval; is the temperature stability of the pth temperature monitoring interval; norm() is the normalization function; ε is the first coefficient preset to be greater than 0.

9. A method for strengthening the internal temperature control of a computer as claimed in claim 1, characterized in that: The process of controlling the temperature of the CPU of the reinforcement computer in the next temperature monitoring interval is: If the temperature adjustment factor of the current temperature monitoring interval is less than the preset temperature adjustment threshold, it is determined that the internal temperature of the reinforced computer is normal at this time, and the rotation speed of the internal cooling fan of the reinforced computer is not adjusted; otherwise, it is determined that the internal temperature of the reinforced computer is abnormal at this time, and the rotation speed of the internal cooling fan of the reinforced computer is adjusted, and the temperature of the CPU of the reinforced computer in the next temperature monitoring interval is controlled according to the adjusted rotation speed; The process of adjusting the speed of the cooling fan inside the hardened computer is as follows: A neural network is used to obtain the temperature adjustment amount of the reinforced computer in the current temperature monitoring interval according to the temperature adjustment factor of the current temperature monitoring interval, the mean temperature data of all acquisition moments, and the mean utilization data; A PID controller is used to adjust the rotation speed of the cooling fan inside the reinforced computer according to the temperature data mean value and the temperature adjustment amount.

Images