Computer performance detection method and system based on artificial intelligence and medium

Abstract

The invention discloses a computer performance detection method and system based on artificial intelligence and a medium, and the method comprises the steps: obtaining a utilization rate of a CPU of a to-be-detected computer in a preset detection period, and determining a temperature coefficient of the CPU based on the utilization rate; if the temperature coefficient belongs to a risk coefficient, determining a first temperature of the CPU based on a preset network model, and obtaining a reference temperature interval corresponding to the CPU; determining a to-be-cooled temperature according to the reference temperature interval and the first temperature, and obtaining a cooling duration corresponding to the to-be-cooled temperature; and after the heat dissipation element operates for the heat dissipation duration, detecting a second temperature of the CPU, and detecting the heat dissipation performance of the to-be-detected computer according to the second temperature. The first temperature is determined through the preset network model in combination with various factors of the environment where the CPU is located, so that the first temperature accurately reflects the actual temperature of the CPU, the determined to-be-cooled temperature and the cooling duration are more accurate, and the cooling performance detection accuracy of the to-be-detected computer is improved.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence technology, and in particular to a computer performance detection method, system and medium based on artificial intelligence. Background Art

[0002] Computer performance refers to the computer's ability to process tasks and execute operations, and involves multiple performance indicators in terms of the central processing unit (CPU), memory (RAM), graphics processing unit (GPU), storage, motherboard, display, and heat dissipation. Among them, heat dissipation performance affects the temperature of the hardware in the computer. If the hardware temperature is too high, on the one hand, the system will automatically reduce the frequency of the processor to prevent overheating and thus reduce computer performance. It may also cause problems such as blue screen, freeze, and restart due to system instability. On the other hand, long-term operation of the hardware in a high temperature environment will accelerate aging, causing the performance of key components such as the processor, graphics card, and memory to decline, or even burn out. Therefore, a good heat dissipation performance of the computer can effectively prevent the hardware temperature from being too high, and ensure that the system and hardware run stably under high load.

[0003] When a computer is running, its temperature can usually be detected, such as the CPU temperature, GPU temperature, etc. Once the temperature is detected to be too high, the cooling is started. However, there is no relevant detection mechanism for the heat dissipation performance itself. How to accurately and effectively detect the heat dissipation performance of a computer has become one of the technical problems that need to be solved. Summary of the invention

[0004] The main purpose of the present invention is to provide a computer performance detection method, system and medium based on artificial intelligence, aiming to solve the technical problem of how to accurately and effectively detect the heat dissipation performance of a computer.

[0005] To achieve the above object, the present invention provides a computer performance detection method based on artificial intelligence, the computer performance detection method comprising:

[0006] Obtaining a usage rate of a CPU of a computer to be detected within a preset detection period, and determining a temperature coefficient of the CPU based on the usage rate;

[0007] Determine whether the temperature coefficient belongs to the risk coefficient, and if it belongs to the risk coefficient, determine the first temperature of the CPU based on a preset network model, and obtain a reference temperature range corresponding to the CPU;

[0008] Determine a temperature to be dissipated according to the reference temperature interval and the first temperature, and obtain a heat dissipation time corresponding to the temperature to be dissipated;

[0009] After the heat dissipation element corresponding to the CPU runs for the heat dissipation time, the second temperature of the CPU is detected, and the heat dissipation performance of the computer to be detected is detected based on the second temperature.

[0010] Preferably, the step of determining the first temperature of the CPU based on a preset network model includes:

[0011] Obtaining a voltage value of a thermistor corresponding to the CPU;

[0012] According to the preset functional relationship between the voltage and temperature corresponding to the thermistor, the voltage value is analyzed and calculated to obtain the temperature value of the CPU, and the preset functional relationship is:

[0013]

[0014] Wherein, V represents the voltage value, T represents the temperature value, w, A, and R represent the test constants corresponding to the preset function relationship, k represents the Boltzmann constant, q represents the electron charge constant, and B represents the current density;

[0015] Determine the components to be tested on the circuit board where the CPU is located, and determine the heating temperature corresponding to the circuit board according to each of the components to be tested;

[0016] Acquire the area parameters of the circuit board, the arrangement parameters of the electronic components on the circuit board, and the usage time and heat dissipation maintenance parameters of the computer to be tested;

[0017] Analyze and process the heating temperature, area parameter, arrangement parameter, usage time and heat dissipation maintenance parameter based on the preset network model to generate a correction coefficient;

[0018] The temperature value is corrected according to the correction coefficient to obtain the first temperature.

[0019] Preferably, the step of determining the heating temperature corresponding to the circuit board according to each of the components under test comprises:

[0020] The temperature equation corresponding to the circuit board is read, and the component temperature of each component to be measured is calculated according to the temperature equation, where the temperature equation is:

[0021]

[0022] Wherein, Tk represents the component temperature of the kth component to be tested, δ represents the temperature coefficient of the ambient temperature corresponding to the preset detection period, Ik represents the current value of the kth component to be tested, Rk represents the resistance value of the kth component to be tested, hk represents the contact heat transfer coefficient of the kth component to be tested, Sk represents the projection area of ​​the kth component to be tested on the circuit board, and Wk represents the test thermal resistance corresponding to the kth component to be tested on the circuit board;

[0023] The heat generation temperature is generated based on each of the element temperatures.

[0024] Preferably, the step of analyzing and processing the heating temperature, area parameter, arrangement parameter, usage time and heat dissipation maintenance parameter based on the preset network model to generate a correction coefficient includes:

[0025] The parameter names corresponding to the heating temperature, area parameter, arrangement parameter, use time and heat dissipation maintenance parameter are simultaneously formed into matrix rows and matrix columns;

[0026] Determine the similarity values ​​between the parameter values ​​corresponding to the heating temperature, area parameter, arrangement parameter, usage time and heat dissipation maintenance parameter, and form each of the similarity values ​​into a matrix element based on the arrangement order of each parameter name in the matrix row and matrix column;

[0027] The matrix formed by the matrix rows, matrix columns and matrix elements is analyzed based on a preset network model to generate the correction coefficient.

[0028] Preferably, the step of determining the first temperature of the CPU based on a preset network model includes:

[0029] Acquire a large amount of sample data associated with computer heat dissipation performance, and divide the sample data into training samples and verification samples;

[0030] The preset initial model is trained based on the training samples, and when the single training time reaches the preset time, the preset initial model is tested and verified based on the verification samples to generate a verification result;

[0031] A loss function value of the preset initial model is generated based on the verification result, and it is determined whether the loss function value is less than a preset threshold value. If it is less than the preset threshold value, the preset initial model is generated as a preset network model.

[0032] Preferably, the step of determining whether the loss function value is less than a preset threshold value includes:

[0033] If the loss function value is greater than or equal to a preset threshold, the preset duration and the model parameters of the preset initial model are updated and calculated based on a preset update formula, and the preset update formula is:

[0034]

[0035] Among them, t represents the preset duration after the update, t0 represents the preset duration before the update, p represents the current number of training times, and w ij represents the updated weight value from the i-th neuron in the input layer to the j-th neuron in the hidden layer in the preset initial model, τ represents the training rate of the preset initial model, n represents the number of neurons in the input layer, m represents the number of neurons in the hidden layer, yj represents the verification output value of the j-th neuron in the hidden layer, yj0 represents the reference output value of the j-th neuron in the hidden layer, and w ij0 represents the weight value before updating from the i-th neuron in the input layer to the j-th neuron in the hidden layer in the preset initial model, w jp represents the updated weight value of the jth neuron from the output layer to the hidden layer in the preset initial model, yp represents the verification output value of the output layer, yp0 represents the reference output value of the output layer, and w jp0 Represents the weight value before updating the jth neuron from the output layer to the hidden layer in the preset initial model;

[0036] For the updated preset initial model, a step of training the preset initial model based on the training samples is performed.

[0037] Preferably, the step of determining the temperature coefficient of the CPU based on the usage rate comprises:

[0038] Reading the mapping relationship between the preset usage rate interval and the preset temperature coefficient corresponding to the computer to be detected, and comparing the usage rate with each of the preset usage rate intervals to determine the target preset usage rate interval where the usage rate is located;

[0039] A target mapping relationship corresponding to the target preset usage rate interval in each of the mapping relationships is obtained, and a preset temperature coefficient mapped in the target mapping relationship is determined as the temperature coefficient.

[0040] Preferably, the step of detecting the heat dissipation performance of the computer to be detected according to the second temperature includes:

[0041] Determining whether the second temperature matches the reference temperature interval, and if so, determining that the heat dissipation performance of the computer to be tested is qualified;

[0042] If the second temperature does not match the reference temperature range, heat dissipation warning information corresponding to the computer to be detected is output.

[0043] Furthermore, to achieve the above-mentioned purpose, the present invention also provides a computer performance detection system based on artificial intelligence, wherein the computer performance detection system based on artificial intelligence includes a memory, a processor, a communication bus, and a control program stored in the memory:

[0044] The communication bus is used to realize the connection and communication between the processor and the memory;

[0045] The processor is used to execute the control program to implement the steps of the computer performance detection method based on artificial intelligence as described above.

[0046] Furthermore, to achieve the above-mentioned purpose, the present invention also provides a medium, which is a readable storage medium, on which a control program is stored, and when the control program is executed by a processor, the steps of the computer performance detection method based on artificial intelligence as described above are implemented.

[0047] The computer performance detection method, system and medium based on artificial intelligence of the present invention are pre-set with a preset detection cycle, the utilization rate of the CPU of the detected computer in the preset detection cycle is obtained, and the temperature coefficient of the CPU is determined according to the obtained utilization rate; it is judged whether the temperature coefficient belongs to the risk coefficient, if it belongs to the risk coefficient, the first temperature of the CPU is determined according to the preset network model, and the reference temperature range of the CPU is obtained; then the temperature to be dissipated is determined according to the reference temperature range and the first temperature, and the heat dissipation time corresponding to the temperature to be dissipated is obtained; after the operation time of the heat dissipation element for dissipating the CPU reaches the heat dissipation time, the second temperature of the CPU is detected, and the heat dissipation performance of the computer to be detected is detected by the second temperature. Among them, the reference temperature range is a temperature range that is conducive to the operation of the CPU, and the heat dissipation time corresponding to the temperature to be dissipated indicates the time required for the computer to be detected to reduce the excessively high first temperature to the reference temperature range when the heat dissipation performance is good. After the heat dissipation element runs for the heat dissipation time, if the detected second temperature is within the reference temperature range, it means that the heat dissipation performance of the computer to be detected is good, otherwise the heat dissipation performance of the computer to be detected needs to be optimized, so as to realize the detection of the heat dissipation performance of the computer to be detected. Furthermore, the first temperature of the CPU in the computer to be tested is determined by combining various factors of the environment where the CPU is located with a preset network model, so that the determined first temperature more accurately reflects the actual temperature of the CPU, and thus the determined temperature to be cooled and the cooling time are more accurate, thereby improving the accuracy of the cooling performance test of the computer to be tested. In this way, accurate and effective testing of the cooling performance of the computer to be tested is achieved. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 It is a flowchart of a first embodiment of a computer performance detection method based on artificial intelligence of the present invention;

[0049] Figure 2 It is a flow chart of a second embodiment of a computer performance detection method based on artificial intelligence of the present invention;

[0050] Figure 3 It is a flowchart of a third embodiment of a computer performance detection method based on artificial intelligence of the present invention;

[0051] Figure 4 It is a structural schematic diagram of the hardware operating environment involved in an embodiment of a computer performance detection system based on artificial intelligence of the present invention.

[0052] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0053] It should be understood that the specific embodiments described herein are only used to explain the present invention, and are not used to limit the present invention.

[0054] The present invention provides a computer performance detection method based on artificial intelligence, please refer to Figure 1 , Figure 1 It is a flowchart of the first embodiment of the computer performance detection method based on artificial intelligence of the present invention.

[0055] The embodiment of the present invention provides an embodiment of a computer performance detection method based on artificial intelligence. It should be noted that although a logical order is shown in the flowchart, in some cases, the steps shown or described may be performed in a different order than that here. Specifically, the computer performance detection method based on artificial intelligence in this embodiment includes:

[0056] Step S10, obtaining the usage rate of the CPU of the computer to be detected within a preset detection period, and determining the temperature coefficient of the CPU based on the usage rate.

[0057] The computer performance detection method based on artificial intelligence in this embodiment is applied to a computer performance detection system. The detection system can be a server connected to a computer communication, and the server is connected to multiple computers to detect the performance of multiple computers at the same time. In addition, computer performance involves multiple aspects of the computer, such as computing performance, graphics processing performance, display performance, heat dissipation performance, etc. The computer performance of this embodiment is preferably heat dissipation performance. Specifically, a computer connected to a server for heat dissipation performance detection is used as a computer to be detected, and different preset detection cycles are pre-set for different models of computers to be detected based on comprehensive factors such as their respective CPU performance and heat dissipation performance. When it is monitored that a certain computer to be detected has reached its preset detection cycle, the usage rate of the CPU of the computer to be detected within the preset detection cycle is obtained, and the temperature coefficient of the CPU is determined based on the obtained usage rate. Among them, the step of determining the temperature coefficient of the CPU based on the usage rate includes:

[0058] Step S11, reading the mapping relationship between the preset usage rate interval and the preset temperature coefficient corresponding to the computer to be detected, and comparing the usage rate with each of the preset usage rate intervals to determine the target preset usage rate interval in which the usage rate is located;

[0059] Step S12, obtaining a target mapping relationship corresponding to the target preset usage rate interval in each of the mapping relationships, and determining the preset temperature coefficient mapped in the target mapping relationship as the temperature coefficient.

[0060] Furthermore, a correspondence between a usage interval and a temperature coefficient is pre-set for the computer to be tested through testing. The higher the usage rate represented by the usage interval, the higher the corresponding temperature coefficient, indicating that the temperature of the CPU is higher in a high usage scenario. The set correspondence is stored as a mapping relationship between a preset usage interval and a preset temperature coefficient corresponding to the computer to be tested. After obtaining the CPU usage rate, the stored mapping relationship is read, and the usage rate is compared with each preset usage interval in the mapping relationship to determine the preset usage interval where the usage rate is located. The preset usage interval is the target preset usage interval. Then, the target mapping relationship where the target preset usage interval is located in each mapping relationship is searched, and the corresponding preset temperature coefficient in the target mapping relationship is the temperature coefficient used to predict the high and low temperature of the CPU.

[0061] It should be noted that the preset detection period is a time period, not a time point, such as 8:00 a.m. to 6:00 p.m. on the 20th of each month. The CPU usage rate is obtained regularly within the time period, such as every hour or irregularly, and then the temperature coefficient corresponding to the CPU is determined according to the usage rate obtained each time. In this way, the CPU temperature is represented by multiple temperature coefficients to avoid missing the detection of the computer's heat dissipation performance within the preset detection period due to the low CPU usage rate obtained at a certain time, which makes the determined temperature coefficient low.

[0062] Step S20, determining whether the temperature coefficient is a risk coefficient, if it is a risk coefficient, determining the first temperature of the CPU based on a preset network model, and obtaining a reference temperature range corresponding to the CPU.

[0063] Furthermore, in order to indicate the high or low temperature coefficient, a preset temperature coefficient threshold is pre-set, and the temperature coefficient is compared with the preset temperature coefficient threshold to determine whether the temperature coefficient is greater than the preset temperature coefficient threshold. If it is greater than the preset temperature coefficient threshold, it means that the value of the temperature coefficient is large, and then the temperature of the CPU is high. At this time, it is determined that the temperature coefficient belongs to the risk coefficient. On the contrary, if the temperature coefficient is not greater than the preset temperature coefficient threshold, it means that the temperature of the CPU is not high and the temperature coefficient does not belong to the risk coefficient.

[0064] Furthermore, the detection system is pre-trained with a preset network model through a large number of data samples that affect the CPU temperature. After determining that the temperature coefficient belongs to the risk coefficient by comparing the temperature coefficient with the preset temperature coefficient threshold, the first temperature of the CPU is determined by the preset network model. The first temperature reflects the actual temperature of the CPU, which at least includes the temperature of the CPU's own heat and the temperature caused by the heat of other nearby electronic components. The preset network model predicts the actual temperature of the CPU based on the temperature of the CPU's own heat and the temperature caused by the heat of nearby electronic components, and obtains the final first temperature. At the same time, the CPU usually has good performance only when it operates normally within an appropriate temperature range, and the appropriate temperature range is obtained as the reference temperature range corresponding to the CPU.

[0065] Step S30, determining a temperature to be dissipated according to the reference temperature interval and the first temperature, and obtaining a heat dissipation time corresponding to the temperature to be dissipated.

[0066] Furthermore, the temperature to be dissipated is determined based on the CPU's appropriate reference temperature range and its first temperature. The temperature to be dissipated is the temperature that the CPU needs to lower. The upper and lower boundary values ​​of the reference temperature range can be averaged to obtain relevant calculation results, and then determined by the difference between the first temperature and the calculation result.

[0067] Furthermore, the heat dissipation device of the computer to be tested is determined, such as fan heat dissipation, heat pipe heat dissipation, etc. Different heat dissipation devices have different heat dissipation performance. The heat dissipation performance of the computer to be tested is tested in advance. When the heat dissipation performance of the computer to be tested is good, the time required for the CPU temperature of the computer to be tested to be reduced by one degree at different ambient temperatures is tested, and the time obtained by the test is stored in correspondence with the ambient temperature. After determining the temperature to be dissipated, first search for the target time corresponding to the ambient temperature in each stored time according to the ambient temperature of the preset detection cycle. Then determine the heat dissipation time corresponding to the temperature to be dissipated based on the target time. The result of the operation obtained by multiplying the target time by the temperature to be dissipated is used as the heat dissipation time.

[0068] In addition, the temperature reduction during heat dissipation may be nonlinear, and different cooling intervals can be set for this purpose. The time required for cooling the CPU of the computer to be tested in each cooling interval under different ambient temperatures can be tested, and the corresponding relationship between the ambient temperature, cooling interval and time is obtained and stored. At this time, for the temperature to be dissipated, the cooling interval in which it is located is determined, and the ambient temperature corresponding to the preset detection cycle is determined, and the target corresponding relationship with the cooling interval and the ambient temperature in the stored corresponding relationships is searched. The time in the target corresponding relationship is the heat dissipation time corresponding to the temperature to be dissipated.

[0069] Step S40, after the heat dissipation element corresponding to the CPU runs for the heat dissipation time, detecting a second temperature of the CPU, and detecting the heat dissipation performance of the computer to be detected based on the second temperature.

[0070] Furthermore, the device for cooling the CPU in the computer to be tested is used as a cooling element corresponding to the CPU. The operation time of the cooling element is counted, and when the operation time reaches the cooling time, the second temperature of the CPU is detected. The second temperature can be the temperature of the CPU itself, or it can be determined by a preset network model combined with the temperature of the CPU itself and the temperature generated by the heat of adjacent electronic components.

[0071] It is understandable that if the heat dissipation performance of the computer to be tested is good, the temperature of the CPU should be reduced to a temperature suitable for the normal operation of the CPU after the heat dissipation element has been running for a long time. Therefore, the heat dissipation performance of the computer to be tested can be tested by whether the second temperature of the CPU is reduced to a temperature suitable for the normal operation of the CPU. Specifically, the step of testing the heat dissipation performance of the computer to be tested according to the second temperature includes:

[0072] Step S41, determining whether the second temperature matches the reference temperature interval, and if so, determining that the heat dissipation performance of the computer to be tested is qualified;

[0073] Step S42: if the second temperature does not match the reference temperature range, outputting heat dissipation warning information corresponding to the computer to be detected.

[0074] Further, the detected second temperature is compared with the reference temperature interval to determine whether the second temperature is within the temperature range represented by the reference temperature interval. Whether the second temperature matches the reference temperature interval is determined by determining whether the second temperature is within the temperature range. If the second temperature is within the temperature range represented by the reference temperature interval, it is determined that the second temperature matches the reference temperature interval. At this time, the computer to be detected reduces the temperature of the CPU to a temperature range suitable for the operation of the CPU within the heat dissipation time, and the computer to be detected has good heat dissipation performance, so it is determined that the heat dissipation performance of the computer to be detected is qualified. On the contrary, if it is determined by comparison that the second temperature is not within the temperature range represented by the reference temperature interval, it is determined that the second temperature does not match the reference temperature interval. At this time, the computer to be detected does not reduce the temperature of the CPU to a temperature range suitable for the operation of the CPU within the heat dissipation time, and the computer to be detected does not have good heat dissipation performance. The model information of the computer to be detected is obtained, and the model information, the heat dissipation time, the second temperature, etc. are formed together as heat dissipation warning information corresponding to the computer to be detected, and the heat dissipation warning information is output to the computer to be detected. The heat dissipation warning information is displayed by a display device connected to the computer to be tested, so as to optimize the heat dissipation performance of the computer to be tested in time and ensure the normal operation and long service life of the computer to be tested.

[0075] The computer performance detection method based on artificial intelligence implemented in this embodiment is pre-set with a preset detection cycle, and the CPU usage rate of the computer to be detected in the preset detection cycle is obtained, and the temperature coefficient of the CPU is determined based on the obtained usage rate; it is judged whether the temperature coefficient belongs to the risk coefficient. If it belongs to the risk coefficient, the first temperature of the CPU is determined based on the preset network model, and the reference temperature range of the CPU is obtained; then the temperature to be dissipated is determined based on the reference temperature range and the first temperature, and the heat dissipation time corresponding to the temperature to be dissipated is obtained; after the running time of the heat dissipation element for dissipating the CPU reaches the heat dissipation time, the second temperature of the CPU is detected, and the heat dissipation performance of the computer to be detected is detected through the second temperature. Among them, the reference temperature range is a temperature range that is conducive to the operation of the CPU, and the heat dissipation time corresponding to the temperature to be dissipated indicates the time required for the computer to be detected to reduce the excessively high first temperature to the reference temperature range when the heat dissipation performance is good. After the heat dissipation element runs for the heat dissipation time, if the detected second temperature is within the reference temperature range, it means that the heat dissipation performance of the computer to be detected is good, otherwise the heat dissipation performance of the computer to be detected needs to be optimized, so as to realize the detection of the heat dissipation performance of the computer to be detected. Furthermore, the first temperature of the CPU in the computer to be tested is determined by combining various factors of the environment where the CPU is located with a preset network model, so that the determined first temperature more accurately reflects the actual temperature of the CPU, and thus the determined temperature to be cooled and the cooling time are more accurate, thereby improving the accuracy of the cooling performance test of the computer to be tested. In this way, accurate and effective testing of the cooling performance of the computer to be tested is achieved.

[0076] For further information, please refer to Figure 2 Based on the first embodiment of the computer performance detection method based on artificial intelligence of the present invention, a second embodiment of the computer performance detection method based on artificial intelligence of the present invention is proposed.

[0077] The difference between the second embodiment of the computer performance detection method based on artificial intelligence and the first embodiment of the computer performance detection method based on artificial intelligence is that the step of determining the first temperature of the CPU based on the preset network model includes:

[0078] Step S21, obtaining a voltage value of a thermistor corresponding to the CPU;

[0079] Step S22, analyzing and calculating the voltage value according to a preset functional relationship between the voltage and temperature corresponding to the thermistor to obtain a temperature value of the CPU;

[0080] Step S23, determining the components to be tested on the circuit board where the CPU is located, and determining the heating temperature corresponding to the circuit board according to each of the components to be tested;

[0081] Furthermore, the first temperature of the CPU at least includes the temperature of the CPU itself and the temperature of the electronic components around the CPU. For the temperature of the CPU itself, a thermal element, such as a thermal PN junction, is provided in the CPU. A preset functional relationship between the voltage and temperature of such thermal element is set in advance through testing, so that the temperature of the CPU itself can be obtained by detecting the voltage value of the thermal element.

[0082] Specifically, the voltage value of the thermistor is obtained, and according to the preset functional relationship between the voltage and temperature corresponding to the thermistor, the voltage value is analyzed and calculated to obtain the temperature value of the CPU. The preset functional relationship can be specifically referred to in the following formula (1).

[0083]

[0084] Among them, V represents the voltage value, T represents the temperature value, w, A, and R represent the test constants corresponding to the preset functional relationship, that is, the constants determined in the process of forming the preset functional relationship through testing, k represents the Boltzmann constant, q represents the electron charge constant, and B represents the current density.

[0085] Furthermore, there are many electronic components involved around the CPU, and different electronic components generate different amounts of heat when running, thus affecting the CPU temperature to different degrees. This embodiment pre-tests the heat generation of each electronic component under the same environment, and pre-sets a heating threshold indicating a greater impact on the CPU temperature, compares each tested heat generation with the heating threshold, and determines a target heat generation greater than the heating threshold among each heat generation. Then, the target electronic component that generates each target heat generation is searched, and the type of each target electronic component is stored as a target type.

[0086] Furthermore, the types of the electronic components on the circuit board where the CPU is located are obtained, and the types of the electronic components are compared with the stored target types respectively, and the electronic components belonging to the target types are screened out. The screened out electronic components are electronic components that generate more heat and have a greater impact on the CPU temperature. This type of electronic component is determined as the component to be tested on the circuit board where the CPU is located, and the heating temperature corresponding to the circuit board is determined by measuring the heat generated by this type of component to be tested. The heating temperature reflects the temperature generated by the heat dissipation of other electronic components on the circuit board that may have a greater impact on the CPU temperature. Specifically, the step of determining the heating temperature corresponding to the circuit board according to each of the components to be tested includes:

[0087] Step S231, reading a temperature equation corresponding to the circuit board, and calculating the component temperature of each of the components to be tested according to the temperature equation;

[0088] Step S232, generating the heating temperature according to the temperature of each component.

[0089] Furthermore, a temperature equation corresponding to the circuit board is pre-tested and set, and the temperature value generated by each electronic component in the circuit board is reflected by the temperature equation. For each component to be tested, the temperature equation is read, and the component temperature of each component to be tested is calculated by the temperature equation. The temperature equation can be specifically referred to as the following formula (2).

[0090]

[0091] Among them, Tk represents the component temperature of the kth component to be tested, δ represents the temperature coefficient of the ambient temperature corresponding to the preset detection cycle, Ik represents the current value of the kth component to be tested, Rk represents the resistance value of the kth component to be tested, hk represents the contact heat transfer coefficient of the kth component to be tested, and the contact heat transfer coefficient is related to the connection method between the circuit board and the kth component to be tested, as well as the number of layers and materials of the circuit board, Sk represents the projected area of ​​the kth component to be tested on the circuit board, and Wk represents the test thermal resistance corresponding to the kth component to be tested on the circuit board, which is determined by testing the thermal resistance between the actual working part of the kth component to be tested and the outermost layer of the circuit board. In addition, different preset detection cycles correspond to different ambient temperatures. At different ambient temperatures, the heat generated by the heat generated by each electronic component has different influences on the CPU temperature, so the temperature coefficient representing the influence of different ambient temperatures on the CPU temperature is set. For the preset detection cycle of the current detection, its ambient temperature is obtained, and the temperature coefficient corresponding to the ambient temperature is found. The temperature generated by the heat generation of each component to be tested is corrected by the temperature coefficient to obtain the component temperature of the component to be tested, so that the component temperature of the component to be tested can more accurately reflect the impact on the CPU temperature.

[0092] Furthermore, after obtaining the component temperature of each component to be tested through the temperature equation, the heating temperature can be generated according to each component temperature. The heating temperature reflects the impact of the overall heating of each component to be tested on the CPU temperature. The generation method can be to add up the temperatures of each component, or to add up the values ​​after removing the decimal values ​​with large offsets. For example, if the difference between the minimum value and the second minimum value of each component temperature is large, the minimum value is removed and then the sum is calculated.

[0093] Step S24, obtaining area parameters of the circuit board, arrangement parameters of the electronic components on the circuit board, and usage time and heat dissipation maintenance parameters of the computer to be tested;

[0094] Step S25, analyzing and processing the heating temperature, area parameter, arrangement parameter, usage time and heat dissipation maintenance parameter based on the preset network model to generate a correction coefficient;

[0095] Step S26, correcting the temperature value according to the correction coefficient to obtain the first temperature.

[0096] Furthermore, the actual first temperature of the CPU is related not only to its own heat generation and the heat generation of the surrounding electronic components, but also to the area of ​​the circuit board and the arrangement of the electronic components on the circuit board, as well as the usage time and heat dissipation maintenance of the computer to be tested. For the area of ​​the circuit board and the arrangement of its electronic components, the smaller the area and the denser the arrangement, the smaller the effective area of ​​the circuit board for heat dissipation, making it easier for heat that affects the actual temperature of the CPU to accumulate, thereby causing the first temperature to rise. For the usage time of the computer to be tested, the longer the usage time, the higher the degree of aging of the CPU and other electronic components, and the heat generation increases with aging, causing the first temperature to rise. The heat dissipation maintenance parameters are parameters such as the time and number of times for cleaning the heat dissipation device of the computer to be cooled. The fewer the number of cleanings and the longer the time, the worse the heat dissipation, and the more difficult it is to lower the first temperature.

[0097] Furthermore, the area parameters of the circuit board, the arrangement parameters of the electronic components in the circuit board, the usage time and the heat dissipation maintenance coefficient of the computer to be tested are obtained, and the heating temperature of the circuit board, the obtained area parameters, arrangement parameters, usage time and heat dissipation maintenance parameters are analyzed and processed by a preset network model to generate a correction coefficient for correcting the temperature value of the CPU. Then, the temperature value of the CPU is corrected by the correction coefficient to obtain a first temperature reflecting the actual temperature of the CPU.

[0098] The step of analyzing and processing the heating temperature, area parameter, arrangement parameter, usage time and heat dissipation maintenance parameter based on the preset network model to generate a correction coefficient includes:

[0099] Step S251, forming the parameter names corresponding to the heating temperature, area parameter, arrangement parameter, use time and heat dissipation maintenance parameter into matrix rows and matrix columns at the same time;

[0100] Step S252, determining similarity values ​​between the parameter values ​​corresponding to the heating temperature, area parameter, arrangement parameter, usage time and heat dissipation maintenance parameter, and forming each of the similarity values ​​into a matrix element based on the arrangement order of each parameter name in the matrix row and matrix column;

[0101] Step S253: Analyze the matrix formed by the matrix rows, matrix columns and matrix elements based on a preset network model to generate the correction coefficient.

[0102] It can be understood that the heating parameters, area parameters, arrangement parameters, usage parameters and heat dissipation maintenance parameters all exist in the form of parameter names and parameter values. The parameter names are the names of the heating parameters, area parameters, arrangement parameters, usage parameters and heat dissipation maintenance parameters, and the parameter values ​​are the numerical values ​​corresponding to the heating parameters, area parameters, arrangement parameters, usage parameters and heat dissipation maintenance parameters. Each parameter name is used as a matrix row and a matrix column at the same time to form a matrix including five matrix rows and five matrix columns. The parameter names corresponding to each matrix row and each matrix column are the same. For example, from the first matrix row to the second matrix row, the parameters are arranged in the order of the parameter names of the heating parameters, area parameters, arrangement parameters, usage parameters and heat dissipation maintenance parameters, and from the first matrix column to the second matrix column, the parameters are also arranged in the order of the parameter names of the heating parameters, area parameters, arrangement parameters, usage parameters and heat dissipation maintenance parameters.

[0103] Further, similarity values ​​are calculated for the corresponding parameter values ​​of heating temperature, area parameter, arrangement parameter, usage time and heat dissipation maintenance parameter. For example, if the corresponding parameter values ​​of heating parameter, area parameter, arrangement parameter, usage parameter and heat dissipation maintenance parameter are a1, a2, a3, a4 and a5 respectively, then the similarity between a1 and a2, a3, a4 and a5 is calculated to obtain similarity values ​​a1a2, a1a3, a1a4 and a1a5, the similarity between a2 and a3, a4 and a5 is calculated to obtain similarity values ​​a2a3, a2a4 and a2a5, the similarity between a3 and a4 and a5 is calculated to obtain similarity values ​​a3a4 and a3a5, and the similarity between a4 and a5 is calculated to obtain similarity value a4a5. The similarity value can be calculated by cosine distance.

[0104] Furthermore, the calculated similarity values ​​are formed into matrix operations according to the arrangement order of the parameter names in the matrix rows and matrix columns. For example, for the matrix rows and matrix columns formed according to the arrangement of the parameter names of the heating parameters, area parameters, arrangement parameters, usage parameters and heat dissipation maintenance parameters, a1a1, a1a2, a1a3, a1a4, a1a5 can be formed as the matrix elements of the first row, and a1a2, a2a2, a1a3, a1a4, a1a5 can be formed as the matrix elements of the second row... Among them, a1a1 and a2a2 represent the similarity of the heating parameters themselves and the similarity of the area parameters themselves, respectively, and the value is 1.

[0105] Furthermore, the matrix rows, matrix columns and matrix elements are formed into a matrix, and the formed matrix is ​​analyzed and processed through a preset network model to obtain a correction coefficient indicating that the CPU temperature needs to be corrected due to the influence of parameters such as heating temperature, area parameters, arrangement parameters, usage time and heat dissipation maintenance parameters.

[0106] In this embodiment, for the first temperature of the CPU, the temperature of the circuit board where the CPU is located, the area parameters of the circuit board and the arrangement parameters of the electronic components thereon, the usage time of the computer to be tested and the heat dissipation maintenance parameters and other factors that have an impact on the actual temperature of the CPU are corrected for the temperature value of the circuit board detected by the thermistor, so that the determined first temperature of the CPU is more consistent with the actual temperature of the CPU, thereby improving the accuracy of the first temperature of the CPU.

[0107] For further information, please refer to Figure 3 Based on the first and second embodiments of the computer performance detection method based on artificial intelligence of the present invention, a third embodiment of the computer performance detection method based on artificial intelligence of the present invention is proposed.

[0108] The third embodiment of the computer performance detection method based on artificial intelligence is different from the first and second embodiments of the computer performance detection method based on artificial intelligence in that the step of determining the first temperature of the CPU based on a preset network model includes:

[0109] Step S50, obtaining a large amount of sample data associated with the heat dissipation performance of the computer, and dividing the sample data into training samples and verification samples;

[0110] Step S60, training a preset initial model based on the training sample, and when a single training time reaches a preset time, testing and verifying the preset initial model based on the verification sample to generate a verification result;

[0111] Step S70, generating a loss function value of the preset initial model based on the verification result, and determining whether the loss function value is less than a preset threshold value, and if so, generating the preset initial model as a preset network model.

[0112] Furthermore, in order to make the correction coefficient generated by the preset network model for analyzing and processing the heating temperature, area parameter, arrangement parameter, use time and heat dissipation maintenance parameter more accurate, it is necessary to train the preset network model with a large amount of sample data associated with the computer heat dissipation performance. The sample data associated with the computer heat dissipation performance may be sample data consisting of historical heating temperature, area parameter, arrangement parameter, use time, heat dissipation maintenance parameter and correction coefficient.

[0113] Specifically, a large amount of sample data associated with the heat dissipation performance of the computer is obtained, and the sample data is divided into training samples and verification samples according to a preset allocation ratio. The preset allocation ratio should make the number of training samples greater than the number of verification samples, for example, a ratio of 8 to 2, or a ratio of 7 to 3, etc., to ensure the accuracy of the training. Then, the preset initial model set in advance is trained according to the training samples, and the preset initial model is preferably a neural network model including an input layer, a hidden layer, and an output layer. A preset duration is preset for a single training. In the process of training the preset initial model through the training samples, the training duration is counted as a single training duration. When the statistical single training duration reaches the preset duration, the trained preset initial model is tested and verified through the verification sample, and the verification sample is analyzed and processed by the trained preset initial model to obtain a processing result as a verification result.

[0114] Further, in order to indicate the accuracy of the preset initial model in processing the verification sample after training, a loss function, such as a cross entropy loss function, is set for the preset initial model. The verification result is calculated by the preset loss function to obtain the loss function value of the preset initial model. The loss function value indicates the difference between the verification result obtained by the preset initial model in analyzing and processing the verification sample and the reference result corresponding to the verification sample. In order to indicate the size of the loss function value, a preset threshold is preset. The generated loss function value is compared with the preset threshold to determine whether the loss function value is less than the preset threshold. If it is less than the preset threshold, it means that the difference between the verification result and the reference result is small, the degree of proximity between the two is high, and the accuracy of the preset initial model in processing the verification sample is high. At this time, the training of the preset initial model is stopped and it is generated as a preset network model. On the contrary, if it is determined that the loss function value is not less than the preset threshold, it means that the difference between the verification result and the reference result is large, the degree of proximity between the two is low, the accuracy of the preset initial model in processing the verification sample is low, and it is necessary to continue to iterate the training of the preset initial model.

[0115] Specifically, the step of determining whether the loss function value is less than a preset threshold value includes:

[0116] Step S80, if the loss function value is greater than or equal to a preset threshold, updating and calculating the model parameters of the preset duration and the preset initial model based on a preset update formula;

[0117] Step S90, for the updated preset initial model, executing the step of training the preset initial model based on the training sample.

[0118] Furthermore, a preset update formula for updating the preset duration and model parameters of the preset initial model is preset. After determining that the loss function value is not less than the preset threshold, the preset duration and model parameters are updated and calculated using the preset update formula. The preset update formula can be found in the following formula (3).

[0119]

[0120] Among them, t represents the preset duration after the update, t0 represents the preset duration before the update, p represents the current number of training times, and w ij represents the updated weight value from the i-th neuron in the input layer to the j-th neuron in the hidden layer in the preset initial model, τ represents the training rate of the preset initial model, n represents the number of neurons in the input layer, m represents the number of neurons in the hidden layer, yj represents the verification output value of the j-th neuron in the hidden layer, that is, the output value of the j-th neuron in the hidden layer during the test and verification of the verification sample by the preset initial model, yj0 represents the reference output value of the j-th neuron in the hidden layer, that is, the reference value preset at the output of the j-th neuron in the hidden layer for the test and verification of the verification sample by the preset initial model, w ij0 represents the weight value before updating from the i-th neuron in the input layer to the j-th neuron in the hidden layer in the preset initial model, w jp represents the updated weight value of the jth neuron from the output layer to the hidden layer in the preset initial model, yp represents the verification output value of the output layer, that is, the output value of the output layer during the test and verification process of the verification sample by the preset initial model, yp0 represents the reference output value of the output layer, that is, the reference value preset in the output of the output layer for the test and verification of the verification sample by the preset initial model, w jp0 Represents the weight value before updating the jth neuron from the output layer to the hidden layer in the preset initial model.

[0121] It should be noted that in order to avoid a single training session being too long, a maximum value of the preset duration can be set. After each update, determine whether the preset duration is greater than or equal to its maximum value. If it is greater than or equal to its maximum value, the preset duration is set to the maximum value, and the preset duration is no longer updated subsequently; if it is determined that the preset duration is less than its maximum value, the preset duration is updated by the preset update formula. Alternatively, a dynamic adjustment mechanism for the preset duration can be set, such as setting a first-level preset duration and a second-level preset duration, and determining whether the preset duration reaches the first-level preset duration after each update. If it reaches it, the formula for updating the preset duration in the above preset update formula is subjected to a ratio reduction process, such as dividing the left part of the formula by a certain value, such as 2, or 3, or 4, etc., to reduce the rate of increase of the preset duration and reduce the duration of a single training session until the preset duration increases to the second-level preset duration. At the same time, the input layer and hidden layer of the preset initial network are both provided with multiple neurons, and the model parameters include the weight values ​​between each neuron in the input layer and each neuron in the hidden layer, and the weight values ​​between each neuron in the hidden layer and the output layer. By updating each weight value, the preset initial model can extract features of the training sample more accurately. After the preset duration and model parameter update of the preset initial model are completed, the preset initial model is trained again with the training sample. After the single training duration reaches the updated preset duration, it is re-verified with the verification sample, and the loss function value is calculated. The training of the preset initial model is stopped until the calculated loss function value is less than the preset threshold, and it is generated as the preset network model.

[0122] This embodiment iteratively trains the preset initial model through a large amount of sample data related to the computer's heat dissipation performance, and updates the preset duration and model parameters of the preset initial model during the training process, so that the preset network model generated by the training can accurately process data related to the computer's heat dissipation performance, thereby improving the accuracy of determining the actual CPU temperature during the computer's heat dissipation performance detection process.

[0123] In addition, the embodiment of the present invention also provides a computer performance detection system based on artificial intelligence. Figure 4 , Figure 4 It is a structural schematic diagram of the equipment hardware operating environment involved in the embodiment of the computer performance detection system based on artificial intelligence of the present invention.

[0124] like Figure 4As shown, the computer performance detection system based on artificial intelligence may include: a processor 1001, such as a CPU, a communication bus 1002, a user interface 1003, a network interface 1004, and a memory 1005. Among them, the communication bus 1002 is used to realize the connection and communication between these components. The user interface 1003 may include a display screen (Display), an input unit such as a keyboard (Keyboard), and the optional user interface 1003 may also include a standard wired interface and a wireless interface. The network interface 1004 may optionally include a standard wired interface and a wireless interface (such as a WI-FI interface). The memory 1005 may be a high-speed RAM memory, or it may be a stable memory (non-volatile memory), such as a disk memory. The memory 1005 may also be a storage device independent of the aforementioned processor 1001.

[0125] Those skilled in the art will understand that Figure 4 The hardware structure of the artificial intelligence-based computer performance detection system shown in the figure does not constitute a limitation on the artificial intelligence-based computer performance detection system, and may include more or fewer components than shown in the figure, or a combination of certain components, or a different arrangement of components.

[0126] like Figure 4 As shown, the memory 1005 as a medium may include an operating system, a network communication module, a user interface module and a control program. The operating system is a program for managing and controlling the computer performance detection system and software resources based on artificial intelligence, and supports the operation of the network communication module, the user interface module, the control program and other programs or software; the network communication module is used to manage and control the network interface 1004; the user interface module is used to manage and control the user interface 1003.

[0127] exist Figure 4 In the hardware structure of the computer performance detection system based on artificial intelligence shown in the figure, the network interface 1004 is mainly used to connect to the system server and communicate data with the system server; the user interface 1003 is mainly used to connect to the client (user end) and communicate data with the client; the processor 1001 can call the control program stored in the memory 1005 and perform the following operations:

[0128] Obtaining a usage rate of a CPU of a computer to be detected within a preset detection period, and determining a temperature coefficient of the CPU based on the usage rate;

[0129] Determine whether the temperature coefficient belongs to the risk coefficient, and if it belongs to the risk coefficient, determine the first temperature of the CPU based on a preset network model, and obtain a reference temperature range corresponding to the CPU;

[0130] Determine a temperature to be dissipated according to the reference temperature interval and the first temperature, and obtain a heat dissipation time corresponding to the temperature to be dissipated;

[0131] After the heat dissipation element corresponding to the CPU runs for the heat dissipation time, the second temperature of the CPU is detected, and the heat dissipation performance of the computer to be detected is detected based on the second temperature.

[0132] Furthermore, the step of determining the first temperature of the CPU based on a preset network model includes:

[0133] Obtaining a voltage value of a thermistor corresponding to the CPU;

[0134] According to the preset functional relationship between the voltage and temperature corresponding to the thermistor, the voltage value is analyzed and calculated to obtain the temperature value of the CPU, and the preset functional relationship is:

[0135]

[0136] Wherein, V represents the voltage value, T represents the temperature value, w, A, and R represent the test constants corresponding to the preset function relationship, k represents the Boltzmann constant, q represents the electron charge constant, and B represents the current density;

[0137] Determine the components to be tested on the circuit board where the CPU is located, and determine the heating temperature corresponding to the circuit board according to each of the components to be tested;

[0138] Acquire the area parameters of the circuit board, the arrangement parameters of the electronic components on the circuit board, and the usage time and heat dissipation maintenance parameters of the computer to be tested;

[0139] Analyze and process the heating temperature, area parameter, arrangement parameter, usage time and heat dissipation maintenance parameter based on the preset network model to generate a correction coefficient;

[0140] The temperature value is corrected according to the correction coefficient to obtain the first temperature.

[0141] Furthermore, the step of determining the heating temperature corresponding to the circuit board according to each of the components to be tested includes:

[0142] The temperature equation corresponding to the circuit board is read, and the component temperature of each component to be measured is calculated according to the temperature equation, where the temperature equation is:

[0143]

[0144] Wherein, Tk represents the component temperature of the kth component to be tested, δ represents the temperature coefficient of the ambient temperature corresponding to the preset detection period, Ik represents the current value of the kth component to be tested, Rk represents the resistance value of the kth component to be tested, hk represents the contact heat transfer coefficient of the kth component to be tested, Sk represents the projection area of ​​the kth component to be tested on the circuit board, and Wk represents the test thermal resistance corresponding to the kth component to be tested on the circuit board;

[0145] The heat generation temperature is generated based on each of the element temperatures.

[0146] Furthermore, the step of analyzing and processing the heating temperature, area parameter, arrangement parameter, usage time and heat dissipation maintenance parameter based on the preset network model to generate a correction coefficient includes:

[0147] The parameter names corresponding to the heating temperature, area parameter, arrangement parameter, use time and heat dissipation maintenance parameter are simultaneously formed into matrix rows and matrix columns;

[0148] Determine the similarity values ​​between the parameter values ​​corresponding to the heating temperature, area parameter, arrangement parameter, usage time and heat dissipation maintenance parameter, and form each of the similarity values ​​into a matrix element based on the arrangement order of each parameter name in the matrix row and matrix column;

[0149] The matrix formed by the matrix rows, matrix columns and matrix elements is analyzed based on a preset network model to generate the correction coefficient.

[0150] Further, before the step of determining the first temperature of the CPU based on the preset network model, the processor 1001 may call a control program stored in the memory 1005 and perform the following operations:

[0151] Acquire a large amount of sample data associated with computer heat dissipation performance, and divide the sample data into training samples and verification samples;

[0152] The preset initial model is trained based on the training samples, and when the single training time reaches the preset time, the preset initial model is tested and verified based on the verification samples to generate a verification result;

[0153] A loss function value of the preset initial model is generated based on the verification result, and it is determined whether the loss function value is less than a preset threshold value. If it is less than the preset threshold value, the preset initial model is generated as a preset network model.

[0154] Further, after the step of determining whether the loss function value is less than a preset threshold, the processor 1001 may call the control program stored in the memory 1005 and perform the following operations:

[0155] If the loss function value is greater than or equal to a preset threshold, the preset duration and the model parameters of the preset initial model are updated and calculated based on a preset update formula, and the preset update formula is:

[0156]

[0157] Among them, t represents the preset duration after the update, t0 represents the preset duration before the update, p represents the current number of training times, and w ij represents the updated weight value from the i-th neuron in the input layer to the j-th neuron in the hidden layer in the preset initial model, τ represents the training rate of the preset initial model, n represents the number of neurons in the input layer, m represents the number of neurons in the hidden layer, yj represents the verification output value of the j-th neuron in the hidden layer, yj0 represents the reference output value of the j-th neuron in the hidden layer, and w ij0 represents the weight value before updating from the i-th neuron in the input layer to the j-th neuron in the hidden layer in the preset initial model, w jp represents the updated weight value of the jth neuron from the output layer to the hidden layer in the preset initial model, yp represents the verification output value of the output layer, yp0 represents the reference output value of the output layer, and w jp0 Represents the weight value before updating the jth neuron from the output layer to the hidden layer in the preset initial model;

[0158] For the updated preset initial model, a step of training the preset initial model based on the training samples is performed.

[0159] Further, the step of determining the temperature coefficient of the CPU based on the usage rate includes:

[0160] Reading the mapping relationship between the preset usage rate interval and the preset temperature coefficient corresponding to the computer to be detected, and comparing the usage rate with each of the preset usage rate intervals to determine the target preset usage rate interval where the usage rate is located;

[0161] A target mapping relationship corresponding to the target preset usage rate interval in each of the mapping relationships is obtained, and a preset temperature coefficient mapped in the target mapping relationship is determined as the temperature coefficient.

[0162] Furthermore, the step of detecting the heat dissipation performance of the computer to be detected according to the second temperature includes:

[0163] Determining whether the second temperature matches the reference temperature interval, and if so, determining that the heat dissipation performance of the computer to be tested is qualified;

[0164] If the second temperature does not match the reference temperature range, heat dissipation warning information corresponding to the computer to be detected is output.

[0165] The specific implementation of the artificial intelligence-based computer performance detection system of the present invention is basically the same as the above-mentioned embodiments of the artificial intelligence-based computer performance detection method, and will not be repeated here.

[0166] The embodiment of the present invention further provides a medium. The medium is a readable storage medium, on which a control program is stored, and when the control program is executed by a processor, the steps of the computer performance detection method based on artificial intelligence are implemented.

[0167] The storage medium of the present invention may be a computer-readable storage medium, and its implementation method is basically the same as the above-mentioned embodiments of the computer performance detection method based on artificial intelligence, and will not be repeated here.

[0168] The embodiments of the present invention are described above in conjunction with the accompanying drawings, but the present invention is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the enlightenment of the present invention, ordinary technicians in this field can also make many forms without departing from the scope of protection of the present invention and the claims. All equivalent structures or equivalent process changes made using the contents of the specification and drawings of the present invention, or directly or indirectly used in other related technical fields, are protected by the present invention.

Claims

1. A computer performance detection method based on artificial intelligence, characterized in that: The computer performance detection method comprises: Obtaining a usage rate of a CPU of a computer to be detected within a preset detection period, and determining a temperature coefficient of the CPU based on the usage rate; Determine whether the temperature coefficient belongs to the risk coefficient, and if it belongs to the risk coefficient, determine the first temperature of the CPU based on a preset network model, and obtain a reference temperature range corresponding to the CPU; Determine a temperature to be dissipated according to the reference temperature interval and the first temperature, and obtain a heat dissipation time corresponding to the temperature to be dissipated; After the heat dissipation element corresponding to the CPU runs for the heat dissipation time, the second temperature of the CPU is detected, and the heat dissipation performance of the computer to be detected is detected based on the second temperature.

2. The computer performance monitoring method according to claim 1, characterized in that: The step of determining the first temperature of the CPU based on a preset network model comprises: Obtaining a voltage value of a thermistor corresponding to the CPU; According to the preset functional relationship between the voltage and temperature corresponding to the thermistor, the voltage value is analyzed and calculated to obtain the temperature value of the CPU, and the preset functional relationship is: Wherein, V represents the voltage value, T represents the temperature value, w, A, and R represent the test constants corresponding to the preset function relationship, k represents the Boltzmann constant, q represents the electron charge constant, and B represents the current density; Determine the components to be tested on the circuit board where the CPU is located, and determine the heating temperature corresponding to the circuit board according to each of the components to be tested; Acquire the area parameters of the circuit board, the arrangement parameters of the electronic components on the circuit board, and the usage time and heat dissipation maintenance parameters of the computer to be tested; Analyze and process the heating temperature, area parameter, arrangement parameter, usage time and heat dissipation maintenance parameter based on the preset network model to generate a correction coefficient; The temperature value is corrected according to the correction coefficient to obtain the first temperature.

3. The computer performance monitoring method according to claim 2, characterized in that: The step of determining the heating temperature corresponding to the circuit board according to each of the components to be tested comprises: The temperature equation corresponding to the circuit board is read, and the component temperature of each component to be measured is calculated according to the temperature equation, where the temperature equation is: Wherein, Tk represents the component temperature of the kth component to be tested, δ represents the temperature coefficient of the ambient temperature corresponding to the preset detection period, Ik represents the current value of the kth component to be tested, Rk represents the resistance value of the kth component to be tested, hk represents the contact heat transfer coefficient of the kth component to be tested, Sk represents the projection area of ​​the kth component to be tested on the circuit board, and Wk represents the test thermal resistance corresponding to the kth component to be tested on the circuit board; The heat generation temperature is generated based on each of the element temperatures.

4. The computer performance monitoring method according to claim 2, characterized in that: The step of analyzing and processing the heating temperature, area parameter, arrangement parameter, usage time and heat dissipation maintenance parameter based on the preset network model to generate a correction coefficient includes: The parameter names corresponding to the heating temperature, area parameter, arrangement parameter, use time and heat dissipation maintenance parameter are simultaneously formed into matrix rows and matrix columns; Determine the similarity values ​​between the parameter values ​​corresponding to the heating temperature, area parameter, arrangement parameter, usage time and heat dissipation maintenance parameter, and form each of the similarity values ​​into a matrix element based on the arrangement order of each parameter name in the matrix row and matrix column; The matrix formed by the matrix rows, matrix columns and matrix elements is analyzed based on a preset network model to generate the correction coefficient.

5. The computer performance monitoring method according to claim 1, characterized in that: The step of determining the first temperature of the CPU based on a preset network model includes: Acquire a large amount of sample data associated with computer heat dissipation performance, and divide the sample data into training samples and verification samples; The preset initial model is trained based on the training samples, and when the single training time reaches the preset time, the preset initial model is tested and verified based on the verification samples to generate a verification result; A loss function value of the preset initial model is generated based on the verification result, and it is determined whether the loss function value is less than a preset threshold value. If it is less than the preset threshold value, the preset initial model is generated as a preset network model.

6. The computer performance monitoring method according to claim 5, characterized in that: The step of determining whether the loss function value is less than a preset threshold value includes: If the loss function value is greater than or equal to a preset threshold, the preset duration and the model parameters of the preset initial model are updated and calculated based on a preset update formula, and the preset update formula is: Among them, t represents the preset duration after the update, t0 represents the preset duration before the update, p represents the current number of training times, and w ij represents the updated weight value from the i-th neuron in the input layer to the j-th neuron in the hidden layer in the preset initial model, τ represents the training rate of the preset initial model, n represents the number of neurons in the input layer, m represents the number of neurons in the hidden layer, yj represents the verification output value of the j-th neuron in the hidden layer, yj0 represents the reference output value of the j-th neuron in the hidden layer, and w ij0 represents the weight value before updating from the i-th neuron in the input layer to the j-th neuron in the hidden layer in the preset initial model, w jp represents the updated weight value of the jth neuron from the output layer to the hidden layer in the preset initial model, yp represents the verification output value of the output layer, yp0 represents the reference output value of the output layer, and w jp0 Represents the weight value before updating the jth neuron from the output layer to the hidden layer in the preset initial model; For the updated preset initial model, a step of training the preset initial model based on the training samples is performed.

7. The computer performance monitoring method according to any one of claims 1 to 6, characterized in that: The step of determining the temperature coefficient of the CPU based on the usage rate comprises: Reading the mapping relationship between the preset usage rate interval and the preset temperature coefficient corresponding to the computer to be detected, and comparing the usage rate with each of the preset usage rate intervals to determine the target preset usage rate interval where the usage rate is located; A target mapping relationship corresponding to the target preset usage rate interval in each of the mapping relationships is obtained, and a preset temperature coefficient mapped in the target mapping relationship is determined as the temperature coefficient.

8. The computer performance monitoring method according to any one of claims 1 to 6, characterized in that: The step of detecting the heat dissipation performance of the computer to be detected according to the second temperature includes: Determining whether the second temperature matches the reference temperature interval, and if so, determining that the heat dissipation performance of the computer to be tested is qualified; If the second temperature does not match the reference temperature range, heat dissipation warning information corresponding to the computer to be detected is output.

9. A computer performance detection system based on artificial intelligence, characterized in that: The computer performance detection system based on artificial intelligence includes a memory, a processor, a communication bus, and a control program stored in the memory: The communication bus is used to realize the connection and communication between the processor and the memory; The processor is used to execute the control program to implement the steps of the computer performance detection method based on artificial intelligence as described in any one of claims 1-8.

10. A medium, characterized in that The medium is a readable storage medium, on which a control program is stored. When the control program is executed by a processor, the steps of the computer performance detection method based on artificial intelligence as described in any one of claims 1 to 8 are implemented.

Images