Computer heat dissipation regulation and control system and method based on temperature prediction

Abstract

The invention discloses a computer heat dissipation regulation and control system and method based on temperature prediction, and the method comprises the following steps: obtaining the historical temperature data of a computer, and building a temperature prediction model according to the historical temperature data of the computer; acquiring a computer temperature curve by using the temperature prediction model; a temperature peak value is obtained, a temperature adjusting mode is determined according to the temperature peak value, temperature adjusting strategies are made, and the temperature adjusting strategies comprise a fan temperature adjusting strategy, a water-cooling temperature adjusting strategy and a progress temperature adjusting strategy; obtaining an actual temperature change curve, obtaining performance change data of the processor through the actual temperature change curve, and estimating the residual life of the processor; the loss function is improved, the temperature prediction precision is improved, and an accurate and effective temperature adjustment strategy can be formulated; a process heat dissipation mode is added, so that heat dissipation can be timely and effectively performed when the computer is at high temperature; the correlativity between the temperature and the processor performance is obtained through the computer temperature data and the processor performance data, and the influence of the temperature on computer hardware can be known.

Description

Technical Field

[0001] The present invention relates to the technical field of computer data processing, and in particular to a computer heat dissipation control system and method based on temperature prediction. Background Art

[0002] In recent years, with the rapid development of new technologies such as big data and cloud computing, the performance of computers has improved rapidly, but at the same time, the improvement in performance is also accompanied by an increase in the heat generated by components when they are working; the accumulation of heat will not only affect the performance and stability of the computer, but may also cause damage to components and shorten their service life; therefore, how to effectively dissipate heat has become a key issue in the field of computer hardware design.

[0003] At present, a Chinese invention with publication number CN118131876B provides a computer system temperature control method and device, which controls the computer temperature by identifying the computer process, and controls the speed of the cooling fan with the temperature when the computer temperature is in a stable state. Although the computer software will not rise rapidly when it is opened, the noise of the computer cooling device is reduced, and the user experience is improved, but there is no prediction of the temperature of the computer system in the future, and the temperature adjustment strategy cannot be formulated in advance. In addition, the temperature adjustment method is relatively simple, and the effect of rapid adjustment for high temperature of the computer is limited. Failure to respond in time may cause hardware damage, and the performance of the computer processor is not evaluated and predicted based on the existing temperature data, which is not conducive to a specific understanding of the impact of computer temperature on processor performance. Summary of the invention

[0004] The technical problem solved by the present invention is that the prior art does not predict the temperature of a computer system in the future, and cannot formulate a temperature control strategy in advance. The temperature control method is relatively simple, and the effect of rapid adjustment for high temperature of the computer is limited. Failure to respond in time may cause hardware damage. Moreover, the performance of the computer processor is not evaluated and predicted based on the existing temperature data, which is not conducive to a specific understanding of the impact of computer temperature on processor performance.

[0005] In order to solve the above technical problems, in a first aspect, the present invention provides a computer heat dissipation control method based on temperature prediction, comprising the following steps:

[0006] Step S1, obtaining computer historical temperature data, and establishing a temperature prediction model according to the computer historical temperature data;

[0007] Step S2, obtaining indoor environment data, computer component data and computer process data, inputting the indoor environment data, the computer component data and the computer process data into a temperature prediction model, and obtaining a computer temperature curve;

[0008] Step S3, obtaining a temperature peak value through the computer temperature curve, and determining a temperature adjustment method according to the temperature peak value, and formulating a temperature adjustment strategy, wherein the temperature adjustment strategy includes a fan temperature adjustment strategy, a water cooling temperature adjustment strategy, and a process temperature adjustment strategy;

[0009] Step S4, acquiring an actual temperature change curve, obtaining processor performance change data through the actual temperature change curve, and estimating the remaining life of the processor based on the processor performance change data.

[0010] As a preferred solution of the computer heat dissipation control method based on temperature prediction described in the present invention, wherein:

[0011] The step S1 specifically includes the following steps:

[0012] Step S101, obtaining computer historical temperature data, wherein the computer historical temperature data includes historical indoor environment data, historical computer component data, historical computer process data and historical computer temperature data;

[0013] Step S102, dividing the computer historical temperature data into a training set and a test set according to a certain ratio, inputting the historical indoor environment data, historical computer component data, historical computer process data and historical computer temperature data in the training set into a machine learning model for training, wherein the historical indoor environment data, the historical computer component data and the historical computer process data are used as model inputs, and the historical computer temperature data are used as model outputs, to obtain a primary temperature prediction model;

[0014] Step S103, inputting the historical indoor environment data, historical computer component data, and historical computer process data in the test set into the primary temperature prediction model, obtaining the test computer temperature data, improving the model loss function to obtain the improved loss function, using the improved loss function to obtain the model loss value according to the historical computer temperature data and the test computer temperature data, if the model loss value is less than the loss expectation threshold, the model training is completed, and the temperature prediction model is obtained, if the model loss value is greater than or equal to the loss expectation threshold, the model parameters of the primary temperature prediction model are adjusted according to the model loss value until the model loss value is less than the loss expectation threshold, and the temperature prediction model is obtained;

[0015] The mathematical expression of the improved loss function is as follows:

[0016] Loss = ∑w h *f(x)*(y h-p -y h-t ) 2 +Σw l *g(x)*(y l-p -yl-t ) 2 +Σ(y n-p -y n-t ) 2

[0017]

[0018] Among them, f(x) is the extreme temperature importance factor when the computer temperature is high, g(x) is the extreme temperature importance factor when the computer temperature is low, a is the extreme temperature importance factor, w h is the weight coefficient when the computer temperature is high, w l is the weight coefficient when the computer temperature is low, y h-p is the predicted temperature when the computer temperature is high, y h-t is the actual temperature when the computer temperature is high, y l-p is the predicted temperature when the computer temperature is low, y l-t is the actual temperature when the computer temperature is low, y n-p is the predicted temperature when the computer temperature is normal, y n-t is the actual temperature when the computer temperature is normal;

[0019] As a preferred solution of the computer heat dissipation control method based on temperature prediction described in the present invention, wherein:

[0020] The step S2 specifically includes the following steps:

[0021] Step S201, divide the indoor space into several subspaces, and obtain the numbers 001, 002...011, 012... of each subspace, use sensors to obtain environmental data of each subspace respectively, obtain environmental data of the subspace where the computer is located, as indoor environmental data, use computer query to obtain computer component data and computer process data, the environmental data includes temperature, humidity, illumination and cleanliness, the computer component data includes component name, component model, component brand and component usage time, the computer process data includes process name, process type, process running time, process idle time, process CPU ratio and process memory occupied, the process type includes idle process and non-idle process;

[0022] Step S202, inputting the indoor environment data, computer component data and computer process data into a temperature prediction model, obtaining temperature increment data within a time period, obtaining the current temperature of the computer, and obtaining a computer temperature curve within a period according to the current temperature of the computer and the temperature increment data, which is defined as a first computer temperature curve;

[0023] As a preferred solution of the computer heat dissipation control method based on temperature prediction described in the present invention, wherein:

[0024] The step S3 specifically comprises the following steps:

[0025] Step S301, obtaining a first peak value of the temperature curve of the first computer, if the first peak value is greater than or equal to a first expected threshold value and less than a second expected threshold value, then turning on a fan temperature adjustment mode, and dividing the temperature curve of the first computer into a plurality of temperature segments evenly according to time, and obtaining a fan temperature adjustment strategy for the time period according to the temperature segments, wherein the fan temperature adjustment strategy includes a change in fan speed over time;

[0026] Step S302, if the first peak value is greater than or equal to the second expected threshold value and less than the third expected threshold value, then start the water cooling temperature adjustment mode, and obtain the water cooling temperature adjustment strategy for each time period according to the temperature curve, wherein the water cooling temperature adjustment strategy includes the change of the solution temperature and the solution flow rate according to time;

[0027] Step S303, if the first peak value is greater than or equal to the third expected threshold, then obtain the process idle time and process CPU ratio of the idle process according to the computer process data, obtain the high-quality priority rate of each idle process according to the process idle time and the process CPU ratio, and arrange the idle processes in ascending order according to the high-quality priority rate, obtain a process list, and determine a process temperature control strategy according to the process list, the process temperature control strategy includes terminating the first process according to the process arrangement order, obtaining the updated computer process data, and obtaining the second computer temperature curve and the second peak value according to the updated computer process data by using the temperature prediction model, if the second peak value is still greater than or equal to the third expected threshold, then terminate the second process until the peak value is less than or equal to the third expected threshold, and then implement steps S301 to S302;

[0028] The first process is the idle process ranked first in the process list, and the second process is the idle process ranked first in the process list after being updated once;

[0029] As a preferred solution of the computer heat dissipation control method based on temperature prediction described in the present invention, wherein:

[0030] The step S4 specifically comprises the following steps:

[0031] Step S401, obtaining CPU throughput, CPU utilization and CPU response time according to historical computer temperature data to evaluate processor performance per unit time and obtain a performance evaluation value;

[0032] Step S402, obtaining a computer temperature curve within a unit time according to historical computer temperature data, and obtaining a segment temperature value within a unit time according to the computer temperature curve, performing a correlation analysis on the segment temperature value and the performance evaluation value using a Pearson correlation coefficient method, and obtaining a correlation formula between temperature and performance evaluation value;

[0033] Step S403, obtaining an actual temperature curve of the computer within a time period, obtaining a performance evaluation value change curve of the processor within a time period according to the actual temperature curve of the computer and a correlation formula, and obtaining a performance loss difference within a period according to the performance evaluation change curve, which is defined as a period performance loss difference, setting a performance threshold of the processor, and obtaining the remaining life of the processor through the period performance loss difference and the performance threshold;

[0034] As a preferred solution of the computer heat dissipation control method based on temperature prediction described in the present invention, wherein:

[0035] The fan temperature adjustment strategies for the time period are obtained based on the temperature segment:

[0036] Obtaining a segment temperature value of the temperature segment, obtaining an initial fan speed according to the segment temperature value, and sequentially reducing the initial fan speed according to time intervals, wherein the segment temperature value and the initial fan speed are in a one-to-one correspondence;

[0037] The mathematical expression for obtaining the temperature value of the curved segment is as follows:

[0038]

[0039] Where T is the temperature of the curved section, is the area of ​​the temperature curve and the x-axis, t is the x-axis length of the curve, β1 is the average temperature coefficient, T h is the highest temperature of the temperature curve, β2 is the highest temperature coefficient;

[0040] As a preferred solution of the computer heat dissipation control method based on temperature prediction described in the present invention, wherein:

[0041] The water cooling temperature control strategies for each time period obtained based on the temperature curve include:

[0042] Obtaining a segment temperature value of the temperature segment, obtaining an initial solution temperature and an initial solution flow rate according to the segment temperature value, and sequentially reducing the initial solution flow rate according to a time interval, wherein the initial solution temperature, the initial solution flow rate and the segment temperature value are in a one-to-one correspondence relationship;

[0043] As a preferred solution of the computer heat dissipation control method based on temperature prediction described in the present invention, wherein:

[0044] The historical indoor environment data includes temperature, humidity, lighting and cleanliness;

[0045] The historical computer component data includes component name, component model, component brand and component usage time;

[0046] The historical computer process data includes process name, process type, process running time, process idle time, process CPU percentage and process memory usage;

[0047] The historical computer temperature data includes time, computer temperature, CPU throughput, CPU utilization and CPU response time;

[0048] As a preferred solution of the computer heat dissipation control method based on temperature prediction described in the present invention, wherein:

[0049] The mathematical expression for obtaining the high-quality priority rate of a process is as follows:

[0050]

[0051] Among them, C is the high quality priority rate, T s is the process idle time, λ1 is the process idle time weight coefficient, W is the process CPU proportion, and λ2 is the process CPU proportion weight coefficient.

[0052] In a second aspect, the present invention provides a computer heat dissipation control system based on temperature prediction, including a temperature prediction module, a temperature control strategy module and a life assessment module:

[0053] The temperature prediction module is used to obtain computer historical temperature data, and establish a temperature prediction model using machine learning model technology based on the computer historical temperature data and the improved loss function;

[0054] The temperature adjustment strategy module is used to obtain a computer temperature curve using a temperature prediction model, and determine a temperature adjustment method and formulate a temperature adjustment strategy according to the computer temperature curve;

[0055] The life evaluation module is used to obtain the correlation formula of temperature and performance evaluation value according to the historical computer temperature data, and use the correlation formula to estimate the remaining life of the processor according to the actual temperature curve of the computer.

[0056] The beneficial effects of the present invention are as follows: the present invention establishes a temperature prediction model based on the historical temperature data of the computer using machine learning model technology, and uses the improved loss function to adjust the model parameters, and uses the temperature prediction model to predict the temperature curve of the computer in a future time period, thereby providing accurate temperature data for timely formulation of temperature control strategies, and the improved loss function effectively improves the prediction accuracy, which is conducive to formulating accurate and effective temperature control strategies, ensuring that the computer can dissipate heat in a timely manner, and is conducive to stabilizing the performance of computer hardware and extending the service life of computer hardware.

[0057] In addition to the original traditional cooling methods of fan cooling and water cooling, a process cooling method is added. When the temperature reaches a certain threshold, a process temperature control strategy is used to enable the computer to dissipate heat and lower the temperature in a timely and effective manner when the computer is at high temperature, thereby ensuring the stability and efficiency of the computer system operation and facilitating the safe use of computer hardware.

[0058] The correlation between temperature and processor performance is obtained through computer temperature data and processor performance data, and the performance loss difference of the processor within a cycle is obtained based on the correlation using the actual temperature curve of the computer. The remaining life of the processor is estimated by the performance loss difference, which is conducive to a detailed understanding of the impact of temperature on computer hardware and computer usage consumption. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] Figure 1 A basic flow chart of a computer heat dissipation control method based on temperature prediction provided by an embodiment of the present invention.

[0060] Figure 2 A basic flow chart of a computer heat dissipation control system based on temperature prediction is provided for one embodiment of the present invention. DETAILED DESCRIPTION

[0061] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the drawings. It is obvious that the described embodiments are only part of the embodiments of the present invention, but not all of the embodiments.

[0062] Example 1, reference Figure 1 , as an embodiment of the present invention, provides a computer heat dissipation control method based on temperature prediction, comprising the following steps:

[0063] Step S1, obtaining computer historical temperature data, and establishing a temperature prediction model according to the computer historical temperature data;

[0064] Step S2, obtaining indoor environment data, computer component data and computer process data, inputting the indoor environment data, the computer component data and the computer process data into a temperature prediction model, and obtaining a computer temperature curve;

[0065] Step S3, obtaining a temperature peak value through the computer temperature curve, and determining a temperature adjustment method according to the temperature peak value, and formulating a temperature adjustment strategy, wherein the temperature adjustment strategy includes a fan temperature adjustment strategy, a water cooling temperature adjustment strategy, and a process temperature adjustment strategy;

[0066] Step S4, acquiring an actual temperature change curve, obtaining processor performance change data through the actual temperature change curve, and estimating the remaining life of the processor based on the processor performance change data.

[0067] In this embodiment, a temperature prediction model is established based on the historical temperature data of the computer using machine learning model technology, and the model parameters are adjusted using the improved loss function. The temperature prediction model is used to predict the temperature curve of the computer in a future time period, which provides accurate temperature data for timely formulation of temperature control strategies. The improved loss function effectively improves the prediction accuracy, which is conducive to formulating accurate and effective temperature control strategies, ensuring that the computer can dissipate heat in a timely manner, and is conducive to stabilizing the performance of computer hardware and extending the service life of computer hardware.

[0068] In addition to the original traditional cooling methods of fan cooling and water cooling, a process cooling method is added. When the temperature reaches a certain threshold, the process temperature adjustment strategy is used, so that the computer can dissipate heat and lower the temperature in a timely and effective manner when it is at high temperature, ensuring the stability and efficiency of the computer system operation, and is conducive to maintaining the safe use of computer hardware;

[0069] The correlation between temperature and processor performance is obtained through computer temperature data and processor performance data, and the performance loss difference of the processor within a cycle is obtained based on the correlation using the actual temperature curve of the computer. The remaining life of the processor is estimated by the performance loss difference, which is conducive to a detailed understanding of the impact of temperature on computer hardware and computer usage consumption.

[0070] The step S1 specifically includes the following steps:

[0071] Step S101, obtaining computer historical temperature data, wherein the computer historical temperature data includes historical indoor environment data, historical computer component data, historical computer process data and historical computer temperature data;

[0072] Step S102, dividing the computer historical temperature data into a training set and a test set according to a certain ratio, inputting the historical indoor environment data, historical computer component data, historical computer process data and historical computer temperature data in the training set into a machine learning model for training, wherein the historical indoor environment data, the historical computer component data and the historical computer process data are used as model inputs, and the historical computer temperature data are used as model outputs, to obtain a primary temperature prediction model;

[0073] Step S103, inputting the historical indoor environment data, historical computer component data, and historical computer process data in the test set into the primary temperature prediction model, obtaining the test computer temperature data, improving the model loss function to obtain the improved loss function, using the improved loss function to obtain the model loss value according to the historical computer temperature data and the test computer temperature data, if the model loss value is less than the loss expectation threshold, the model training is completed, and the temperature prediction model is obtained, if the model loss value is greater than or equal to the loss expectation threshold, the model parameters of the primary temperature prediction model are adjusted according to the model loss value until the model loss value is less than the loss expectation threshold, and the temperature prediction model is obtained;

[0074] The mathematical expression of the improved loss function is as follows:

[0075] Loss=Σw h *f(x)*(y h-p -y h-t ) 2 +∑w l *g(x)*(y l-p -y l-t ) 2 +∑(y n-p -y n-t ) 2

[0076]

[0077] Among them, f(x) is the extreme temperature importance factor when the computer temperature is high, g(x) is the extreme temperature importance factor when the computer temperature is low, a is the extreme temperature importance factor, w h is the weight coefficient when the computer temperature is high, w l is the weight coefficient when the computer temperature is low, y h-p is the predicted temperature when the computer temperature is high, y h-t is the actual temperature when the computer temperature is high, y l-p is the predicted temperature when the computer temperature is low, y l-t is the actual temperature when the computer temperature is low, y n-p is the predicted temperature when the computer temperature is normal, yn-t It is the actual temperature when the computer is at normal temperature.

[0078] The machine learning model is trained by inputting massive training data to grasp the potential laws contained in the data, and then accurately classify or predict the newly input data;

[0079] The loss expectation threshold is 0.2;

[0080] In this embodiment, a temperature prediction model is established based on the computer's historical temperature data using machine learning model technology, and the model parameters are adjusted using an improved loss function, thereby providing an accurate and effective machine learning model for predicting the computer's temperature curve within a time period. The improved loss function improves the prediction accuracy, which is conducive to accurately and effectively predicting the computer's temperature curve.

[0081] The step S2 specifically includes the following steps:

[0082] Step S201, divide the indoor space into several subspaces, and obtain the numbers 001, 002...011, 012... of each subspace, use sensors to obtain environmental data of each subspace respectively, obtain environmental data of the subspace where the computer is located, as indoor environmental data, use computer query to obtain computer component data and computer process data, the environmental data includes temperature, humidity, illumination and cleanliness, the computer component data includes component name, component model, component brand and component usage time, the computer process data includes process name, process type, process running time, process idle time, process CPU ratio and process memory occupied, the process type includes idle process and non-idle process;

[0083] Step S202, input the indoor environment data, computer component data and computer process data into the temperature prediction model, obtain the temperature increment data within a time period, obtain the current temperature of the computer, and obtain the computer temperature curve within a period based on the current temperature of the computer and the temperature increment data, which is defined as the first computer temperature curve.

[0084] The cleanliness is used to indicate the dust content per unit space in an indoor environment;

[0085] The process running time represents the time for the process to execute an operation;

[0086] The process idle time refers to the time when the process does not perform any operation;

[0087] In this embodiment, by acquiring indoor environment data, computer component data and computer process data and using a temperature prediction model to obtain a first computer temperature curve, it is beneficial to timely understand the computer operating temperature and provide accurate data support for selecting a temperature control method and formulating a temperature control strategy, which is beneficial to timely heat dissipation of the computer and ensures smooth and efficient operation of the computer.

[0088] The step S3 specifically comprises the following steps:

[0089] Step S301, obtaining a first peak value of the temperature curve of the first computer, if the first peak value is greater than or equal to a first expected threshold value and less than a second expected threshold value, then turning on a fan temperature adjustment mode, and dividing the temperature curve of the first computer into a plurality of temperature segments evenly according to time, and obtaining a fan temperature adjustment strategy for the time period according to the temperature segments, wherein the fan temperature adjustment strategy includes a change in fan speed over time;

[0090] Step S302, if the first peak value is greater than or equal to the second expected threshold value and less than the third expected threshold value, then start the water cooling temperature adjustment mode, and obtain the water cooling temperature adjustment strategy for each time period according to the temperature curve, wherein the water cooling temperature adjustment strategy includes the change of the solution temperature and the solution flow rate according to time;

[0091] Step S303, if the first peak value is greater than or equal to the third expected threshold, then obtain the process idle time and process CPU ratio of the idle process according to the computer process data, obtain the high-quality priority rate of each idle process according to the process idle time and the process CPU ratio, and arrange the idle processes in ascending order according to the high-quality priority rate, obtain a process list, and determine a process temperature control strategy according to the process list, the process temperature control strategy includes terminating the first process according to the process arrangement order, obtaining the updated computer process data, and obtaining the second computer temperature curve and the second peak value according to the updated computer process data by using the temperature prediction model, if the second peak value is still greater than or equal to the third expected threshold, then terminate the second process until the peak value is less than or equal to the third expected threshold, and then implement steps S301 to S302;

[0092] The first process is the idle process ranked first in the process list, and the second process is the idle process ranked first in the process list after being updated once.

[0093] The first expected threshold is 45°C;

[0094] The second expected threshold is 65°C;

[0095] The third expected threshold is 80°C;

[0096] In this embodiment, a temperature control method is selected according to the first computer temperature curve, and a detailed and specific temperature control strategy is formulated, which can flexibly and effectively dissipate computer heat, so that the computer can dissipate heat and lower the temperature in a timely and effective manner when it is at high temperature, thereby ensuring the stability and efficiency of the computer system operation, and facilitating the safe use of computer hardware.

[0097] The step S4 specifically comprises the following steps:

[0098] Step S401, evaluating the processor performance per unit time according to the CPU throughput, CPU utilization and CPU response time in the historical computer temperature data, and obtaining a performance evaluation value;

[0099] Step S402, obtaining a computer temperature curve within a unit time according to historical computer temperature data, and obtaining a segment temperature value within a unit time according to the computer temperature curve, performing a correlation analysis on the segment temperature value and the performance evaluation value using a Pearson correlation coefficient method, and obtaining a correlation formula between temperature and performance evaluation value;

[0100] Step S403, obtain the actual temperature curve of the computer within a time period, obtain the performance evaluation value change curve of the processor within a time period according to the actual temperature curve of the computer and the correlation formula, and obtain the performance loss difference within a period according to the performance evaluation change curve, which is defined as the periodic performance loss difference, set the performance threshold of the processor, and obtain the remaining life of the processor through the periodic performance loss difference and the performance threshold.

[0101] The CPU throughput represents the amount of data processed by the processor per unit time;

[0102] The CPU utilization rate indicates the usage of the CPU per unit time;

[0103] The CPU response time indicates the time required to complete an operation task.

[0104] In this embodiment, the correlation between temperature and processor performance is obtained through computer temperature data and processor performance data, and the actual temperature curve of the computer is used to obtain the performance loss difference of the processor within a cycle based on the correlation, and the remaining life of the processor is estimated by the performance loss difference, which is conducive to a detailed understanding of the impact of temperature on computer hardware and the usage consumption of the computer.

[0105] The fan temperature adjustment strategies for the time period are obtained based on the temperature segment:

[0106] Obtaining a segment temperature value of the temperature segment, obtaining an initial fan speed according to the segment temperature value, and sequentially reducing the initial fan speed according to time intervals, wherein the segment temperature value and the initial fan speed are in a one-to-one correspondence;

[0107] The mathematical expression for obtaining the temperature value of the curved segment is as follows:

[0108]

[0109] Where T is the temperature of the curved section, is the area of ​​the temperature curve and the x-axis, t is the x-axis length of the curve, β1 is the average temperature coefficient, T h is the highest temperature of the temperature curve, β2 is the highest temperature coefficient;

[0110] In this embodiment, the initial fan speed is obtained according to the segment temperature value, and the fan speed is gradually reduced in a computer temperature segment. Timely adjustment of the fan speed according to the computer temperature can ensure the heat dissipation of the computer in a timely and effective manner. Gradually reducing the fan speed can both ensure the heat dissipation of the computer and reduce the energy consumption for heat dissipation.

[0111] The water cooling temperature control strategies for each time period obtained based on the temperature curve include:

[0112] The segment temperature value of the temperature segment is obtained, the initial solution temperature and the initial solution flow rate are obtained through the segment temperature value, and the initial solution flow rate is reduced sequentially according to the time interval, and the initial solution temperature, the initial solution flow rate and the segment temperature value are all in a one-to-one correspondence.

[0113] In this embodiment, in this embodiment, the initial solution temperature and the initial solution flow rate are obtained according to the segment temperature value, and the solution flow rate is gradually reduced in a computer temperature segment, which can ensure the heat dissipation of the computer in a timely and effective manner. The gradual solution flow rate not only ensures the heat dissipation of the computer, but also reduces the energy consumption for heat dissipation.

[0114] The historical indoor environment data includes temperature, humidity, lighting and cleanliness;

[0115] The historical computer component data includes component name, component model, component brand and component usage time;

[0116] The historical computer process data includes process name, process type, process running time, process idle time, process CPU percentage and process memory usage;

[0117] The historical computer temperature data includes time, computer temperature, CPU throughput, CPU utilization, and CPU response time.

[0118] In this embodiment, accurate and specific training data and test data are provided for establishing a temperature prediction model by acquiring historical indoor environment data, historical computer component data, historical computer process data and historical computer temperature data.

[0119] The mathematical expression for obtaining the high-quality priority rate of a process is as follows:

[0120]

[0121] Among them, C is the high quality priority rate, T s is the process idle time, λ1 is the process idle time weight coefficient, W is the process CPU proportion, and λ2 is the process CPU proportion weight coefficient.

[0122] In this embodiment, the high-quality priority rate of the idle process is obtained based on the process idle time and the process CPU ratio. In order to terminate the idle process through the high-quality priority rate, the computer can dissipate heat and lower the temperature in a timely and effective manner when it is at high temperature, thereby ensuring the stability and efficiency of the computer system operation and facilitating the safe use of computer hardware.

[0123] Example 2, reference Figure 2 , which is another embodiment of the present invention. This embodiment is different from the first embodiment in that it provides a computer heat dissipation control system based on temperature prediction, characterized in that it includes a temperature prediction module, a temperature control strategy module and a life assessment module:

[0124] The temperature prediction module is used to obtain computer historical temperature data, and establish a temperature prediction model using machine learning model technology based on the computer historical temperature data and the improved loss function;

[0125] The temperature adjustment strategy module is used to obtain a computer temperature curve using a temperature prediction model, and determine a temperature adjustment method and formulate a temperature adjustment strategy according to the computer temperature curve;

[0126] The life evaluation module is used to obtain a correlation formula between temperature and performance evaluation values ​​according to historical computer temperature data, and to estimate the remaining life of the processor according to the actual temperature curve of the computer using the correlation formula.

[0127] In this embodiment, a temperature prediction model is established based on the historical temperature data of the computer using machine learning model technology, and the model parameters are adjusted using the improved loss function. The temperature prediction model is used to predict the temperature curve of the computer in a future time period, which provides accurate temperature data for timely formulation of temperature control strategies. The improved loss function effectively improves the prediction accuracy, which is conducive to formulating accurate and effective temperature control strategies, ensuring that the computer can dissipate heat in a timely manner, and is conducive to stabilizing the performance of computer hardware and extending the service life of computer hardware.

[0128] In addition to the original traditional cooling methods of fan cooling and water cooling, a process cooling method is added. When the temperature reaches a certain threshold, the process temperature adjustment strategy is used to enable the computer to dissipate heat and lower the temperature in a timely and effective manner when the computer is at high temperature, ensuring the stability and efficiency of the computer system operation, and is conducive to maintaining the safe use of computer hardware;

[0129] The correlation between temperature and processor performance is obtained through computer temperature data and processor performance data, and the performance loss difference of the processor within a cycle is obtained based on the correlation using the actual temperature curve of the computer. The remaining life of the processor is estimated by the performance loss difference, which is conducive to a detailed understanding of the impact of temperature on computer hardware and computer usage consumption.

[0130] It should be understood by those skilled in the art that the embodiments of the present invention can be provided as methods, systems or computer program products. Therefore, the present invention can take the form of a complete hardware embodiment, a complete software embodiment or an embodiment combining software and hardware. Moreover, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media containing computer-usable program codes. Among them, the storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (Static Random Access Memory, referred to as SRAM), electrically erasable programmable read-only memory (Electrically Erasable Programmable Read-Only Memory, referred to as EEPROM), erasable programmable read-only memory (Erasable Programmable Read Only Memory, referred to as EPROM), programmable read-only memory (Programmable Red-Only Memory, referred to as PROM), read-only memory (Read-Only Memory, referred to as ROM), magnetic memory, flash memory, disk or optical disk. These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0131] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A computer heat dissipation control method based on temperature prediction, characterized in that: The following steps are involved: Step S1, obtaining computer historical temperature data, and establishing a temperature prediction model according to the computer historical temperature data; Step S2, obtaining indoor environment data, computer component data and computer process data, inputting the indoor environment data, the computer component data and the computer process data into a temperature prediction model, and obtaining a computer temperature curve; Step S3, obtaining a temperature peak value through the computer temperature curve, and determining a temperature adjustment method according to the temperature peak value, and formulating a temperature adjustment strategy, wherein the temperature adjustment strategy includes a fan temperature adjustment strategy, a water cooling temperature adjustment strategy, and a process temperature adjustment strategy; Step S4, acquiring an actual temperature change curve, obtaining processor performance change data through the actual temperature change curve, and estimating the remaining life of the processor based on the processor performance change data.

2. A computer heat dissipation control method based on temperature prediction as claimed in claim 1, characterized in that: The step S1 specifically includes the following steps: Step S101, obtaining computer historical temperature data, wherein the computer historical temperature data includes historical indoor environment data, historical computer component data, historical computer process data and historical computer temperature data; Step S102, dividing the computer historical temperature data into a training set and a test set according to a certain ratio, inputting the historical indoor environment data, historical computer component data, historical computer process data and historical computer temperature data in the training set into a machine learning model for training, wherein the historical indoor environment data, the historical computer component data and the historical computer process data are used as model inputs, and the historical computer temperature data are used as model outputs, to obtain a primary temperature prediction model; Step S103, inputting the historical indoor environment data, historical computer component data, and historical computer process data in the test set into the primary temperature prediction model, obtaining the test computer temperature data, improving the model loss function to obtain the improved loss function, using the improved loss function to obtain the model loss value according to the historical computer temperature data and the test computer temperature data, if the model loss value is less than the loss expectation threshold, the model training is completed, and the temperature prediction model is obtained, if the model loss value is greater than or equal to the loss expectation threshold, the model parameters of the primary temperature prediction model are adjusted according to the model loss value until the model loss value is less than the loss expectation threshold, and the temperature prediction model is obtained; The mathematical expression of the improved loss function is as follows: Loss=Σw h *f(x)*(y h-p -y h-t ) 2 +Σw l *g(x)*(y l-p -y l-t ) 2 +∑(y n-p -y n-t ) 2 Among them, f(x) is the extreme temperature importance factor when the computer temperature is high, g(x) is the extreme temperature importance factor when the computer temperature is low, a is the extreme temperature importance factor, w h is the weight coefficient when the computer temperature is high, w l is the weight coefficient when the computer temperature is low, y h-p is the predicted temperature when the computer temperature is high, y h-t is the actual temperature when the computer temperature is high, y l-p is the predicted temperature when the computer temperature is low, y l-t is the actual temperature when the computer temperature is low, y n-p is the predicted temperature when the computer temperature is normal, y n-t It is the actual temperature when the computer is at normal temperature.

3. A computer heat dissipation control method based on temperature prediction as claimed in claim 1, characterized in that: The step S2 specifically includes the following steps: Step S201, divide the indoor space into several subspaces, and obtain the numbers 001, 002...011, 012... of each subspace, use sensors to obtain environmental data of each subspace respectively, obtain environmental data of the subspace where the computer is located, as indoor environmental data, use computer query to obtain computer component data and computer process data, the environmental data includes temperature, humidity, illumination and cleanliness, the computer component data includes component name, component model, component brand and component usage time, the computer process data includes process name, process type, process running time, process idle time, process CPU ratio and process memory occupied, the process type includes idle process and non-idle process; Step S202, input the indoor environment data, computer component data and computer process data into the temperature prediction model, obtain the temperature increment data within a time period, obtain the current temperature of the computer, and obtain the computer temperature curve within a period based on the current temperature of the computer and the temperature increment data, which is defined as the first computer temperature curve.

4. A computer heat dissipation control method based on temperature prediction as claimed in claim 3, characterized in that: The step S3 specifically comprises the following steps: Step S301, obtaining a first peak value of the temperature curve of the first computer, if the first peak value is greater than or equal to a first expected threshold value and less than a second expected threshold value, then turning on a fan temperature adjustment mode, and dividing the temperature curve of the first computer into a plurality of temperature segments evenly according to time, and obtaining a fan temperature adjustment strategy for the time period according to the temperature segments, wherein the fan temperature adjustment strategy includes a change in fan speed over time; Step S302, if the first peak value is greater than or equal to the second expected threshold value and less than the third expected threshold value, then start the water cooling temperature adjustment mode, and obtain the water cooling temperature adjustment strategy for each time period according to the temperature curve, wherein the water cooling temperature adjustment strategy includes the change of the solution temperature and the solution flow rate according to time; Step S303, if the first peak value is greater than or equal to the third expected threshold, then obtain the process idle time and process CPU ratio of the idle process according to the computer process data, obtain the high-quality priority rate of each idle process according to the process idle time and the process CPU ratio, and arrange the idle processes in ascending order according to the high-quality priority rate, obtain a process list, and determine a process temperature control strategy according to the process list, the process temperature control strategy includes terminating the first process according to the process arrangement order, obtaining the updated computer process data, and obtaining the second computer temperature curve and the second peak value according to the updated computer process data by using the temperature prediction model, if the second peak value is still greater than or equal to the third expected threshold, then terminate the second process until the peak value is less than or equal to the third expected threshold, and then implement steps S301 to S302; The first process is the idle process ranked first in the process list, and the second process is the idle process ranked first in the process list after being updated once.

5. The computer heat dissipation control method based on temperature prediction according to claim 1, characterized in that: The step S4 specifically comprises the following steps: Step S401, evaluating the processor performance per unit time according to the CPU throughput, CPU utilization and CPU response time in the historical computer temperature data, and obtaining a performance evaluation value; Step S402, obtaining a computer temperature curve within a unit time according to historical computer temperature data, and obtaining a segment temperature value within a unit time according to the computer temperature curve, performing a correlation analysis on the segment temperature value and the performance evaluation value using a Pearson correlation coefficient method, and obtaining a correlation formula between temperature and performance evaluation value; Step S403, obtain the actual temperature curve of the computer within a time period, obtain the performance evaluation value change curve of the processor within a time period according to the actual temperature curve of the computer and the correlation formula, and obtain the performance loss difference within a period according to the performance evaluation change curve, which is defined as the periodic performance loss difference, set the performance threshold of the processor, and obtain the remaining life of the processor through the periodic performance loss difference and the performance threshold.

6. A computer heat dissipation control method based on temperature prediction as claimed in claim 4, characterized in that: The fan temperature adjustment strategies for the time period are obtained based on the temperature segment: Obtaining a segment temperature value of the temperature segment, obtaining an initial fan speed according to the segment temperature value, and sequentially reducing the initial fan speed according to time intervals, wherein the segment temperature value and the initial fan speed are in a one-to-one correspondence; The mathematical expression for obtaining the temperature value of the curved segment is as follows: Where T is the temperature of the curved section, is the area of ​​the temperature curve and the x-axis, t is the x-axis length of the curve, β1 is the average temperature coefficient, T h is the highest temperature of the temperature curve, and β2 is the maximum temperature coefficient.

7. A computer heat dissipation control method based on temperature prediction as claimed in claim 1, characterized in that: The water cooling temperature control strategies for each time period obtained based on the temperature curve include: The segment temperature value of the temperature segment is obtained, the initial solution temperature and the initial solution flow rate are obtained through the segment temperature value, and the initial solution flow rate is reduced sequentially according to the time interval, and the initial solution temperature, the initial solution flow rate and the segment temperature value are all in a one-to-one correspondence.

8. A computer heat dissipation control method based on temperature prediction as claimed in claim 2, characterized in that: The historical indoor environment data includes temperature, humidity, lighting and cleanliness; The historical computer component data includes component name, component model, component brand and component usage time; The historical computer process data includes process name, process type, process running time, process idle time, process CPU percentage and process memory usage; The historical computer temperature data includes time, computer temperature, CPU throughput, CPU utilization, and CPU response time.

9. The computer heat dissipation control method based on temperature prediction according to claim 1, characterized in that: The mathematical expression for obtaining the high-quality priority rate of a process is as follows: Among them, C is the high quality priority rate, T s is the process idle time, λ1 is the process idle time weight coefficient, W is the process CPU proportion, and λ2 is the process CPU proportion weight coefficient.

10. A computer heat dissipation control system based on temperature prediction, characterized in that: Including temperature prediction module, temperature control strategy module and life assessment module: The temperature prediction module is used to obtain computer historical temperature data, and establish a temperature prediction model using machine learning model technology based on the computer historical temperature data and the improved loss function; The temperature adjustment strategy module is used to obtain a computer temperature curve using a temperature prediction model, and determine a temperature adjustment method and formulate a temperature adjustment strategy according to the computer temperature curve; The life evaluation module is used to obtain a correlation formula between temperature and performance evaluation values ​​according to historical computer temperature data, and to estimate the remaining life of the processor according to the actual temperature curve of the computer using the correlation formula.

Images