Air inlet temperature correction method and device, electronic equipment and storage medium

Abstract

The embodiment of the invention provides an air inlet temperature correction method and device, electronic equipment and a storage medium, and relates to the technical field of network equipment heat dissipation. The method comprises the following steps: acquiring first sensor data corresponding to a first temperature sensor and second sensor data corresponding to a second temperature sensor in target network equipment; the first temperature sensor and the second temperature sensor are internal temperature sensors located at different positions in the target network equipment; inputting the first sensor data and the second sensor data into a temperature correction model to obtain a temperature correction value corresponding to the target network equipment; and according to the first sensing data and the temperature correction value, the air inlet temperature of the target network equipment is determined, so that the air inlet temperature of the network equipment is corrected, and the precision of the air inlet temperature of the network equipment is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of heat dissipation of network equipment, and in particular to an air inlet temperature correction method, device, electronic equipment and storage medium. Background Art

[0002] Currently, with the rapid development of information technology, the scale of IDCs (Internet Data Centers) and communication networks continues to expand, and the requirements for the performance and reliability of network equipment are becoming increasingly stringent. Network equipment generally has the prominent characteristic of high operating power consumption, and its heat dissipation has always been a significant factor affecting its performance and lifespan. In particular, the air inlet temperature of network equipment has a key impact on the efficiency of the entire cooling system. If the collected air inlet temperature of the network equipment is lower than the external ambient temperature, the fan speed in the cooling strategy based on this temperature will be too low, which is not conducive to the heat dissipation of the entire network equipment and thus affects the performance and stability of the network equipment. Conversely, the fan speed in the cooling strategy based on this temperature will be too high, resulting in high noise, high power consumption, and shortened fan lifespan.

[0003] Therefore, how to correct the air inlet temperature of network equipment to collect accurate network equipment air inlet temperature is of great significance for researching and improving heat dissipation technology based on network equipment air inlet temperature, improving the operating stability of network equipment and extending the service life of equipment. Summary of the Invention

[0004] Embodiments of the present invention provide an air inlet temperature correction method, device, electronic device, and storage medium to correct the air inlet temperature of a network device and improve the accuracy of the air inlet temperature of the network device.

[0005] A first aspect of an embodiment of the present invention provides a method for correcting air inlet temperature, the method comprising:

[0006] Acquire first sensor data corresponding to a first temperature sensor inside a target network device and second sensor data corresponding to a second temperature sensor; the first temperature sensor and the second temperature sensor are internal temperature sensors located at different positions inside the target network device;

[0007] Inputting the first sensor data and the second sensor data into a temperature correction model to obtain a temperature correction value corresponding to the target network device;

[0008] determining an air inlet temperature of the target network device according to the first sensing data and the temperature correction value;

[0009] In which, the temperature correction model is obtained based on the first sample sensor data corresponding to the first sample temperature sensor inside the sample network device, the second sample sensor data corresponding to the second sample temperature sensor inside the sample network device, and the sample external temperature corresponding to the sample network device; the first sample temperature sensor is the first temperature sensor inside the sample network device used for model training, the second sample temperature sensor is the second temperature sensor inside the sample network device used for model training, and the sample external temperature is the actual temperature of the external environment in which the sample network device is located.

[0010] A second aspect of an embodiment of the present invention provides an air inlet temperature correction device, the device comprising:

[0011] a data acquisition module, configured to acquire first sensor data corresponding to a first temperature sensor inside a target network device and second sensor data corresponding to a second temperature sensor inside the target network device; the first temperature sensor and the second temperature sensor are internal temperature sensors located at different positions inside the target network device;

[0012] a model prediction module, configured to input the first sensor data and the second sensor data into a temperature correction model to obtain a temperature correction value corresponding to the target network device;

[0013] a temperature correction module, configured to determine an air inlet temperature of the target network device based on the first sensing data and the temperature correction value;

[0014] In which, the temperature correction model is obtained based on the first sample sensor data corresponding to the first sample temperature sensor inside the sample network device, the second sample sensor data corresponding to the second sample temperature sensor inside the sample network device, and the sample external temperature corresponding to the sample network device; the first sample temperature sensor is the first temperature sensor inside the sample network device used for model training, the second sample temperature sensor is the second temperature sensor inside the sample network device used for model training, and the sample external temperature is the actual temperature of the external environment in which the sample network device is located.

[0015] A third aspect of an embodiment of the present invention provides an electronic device, comprising a processor, a memory, and a computer program stored in the memory and executable on the processor. When the computer program is executed by the processor, the steps of the air inlet temperature correction method as described in the first aspect of the embodiment of the present invention are implemented.

[0016] A fourth aspect of an embodiment of the present invention provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the steps of the air inlet temperature correction method as described in the first aspect of the embodiment of the present invention are implemented.

[0017] A fifth aspect of an embodiment of the present invention provides a computer program product, comprising a computer program / instruction, which, when executed by a processor, implements the steps of the air inlet temperature correction method as described in the first aspect of the embodiment of the present invention.

[0018] An embodiment of the present invention provides an air inlet temperature correction method. First, first sensor data corresponding to a first temperature sensor and second sensor data corresponding to a second temperature sensor within a target network device are obtained. Second, the first sensor data and the second sensor data are input into a temperature correction model to obtain a temperature correction value corresponding to the target network device. Finally, the air inlet temperature of the target network device is determined based on the first sensor data and the temperature correction value. In this embodiment, the influence of internal temperature sensors at different locations within the target network device is fully considered. Sensor data corresponding to the first temperature sensor and the second temperature sensor within the target network device are obtained separately. Dynamic and accurate prediction of the air inlet temperature deviation value is performed based on these sensor data and the temperature correction model. The air inlet temperature calculation no longer relies solely on a fixed temperature deviation value and the data of the internal temperature sensor closest to the air inlet. This achieves accurate correction of the air inlet temperature of the network device and improves the accuracy of the air inlet temperature of the network device. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments of the present invention. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0020] Figure 1 This is a flowchart of a method for correcting air inlet temperature according to an embodiment of the present invention;

[0021] Figure 2 This is a method for correcting the air inlet temperature of a switch provided in the related art;

[0022] Figure 3 This is a principle block diagram of a switch air inlet temperature collection system according to an embodiment of the present invention;

[0023] Figure 4The present invention provides a method for correcting the temperature of a switch air inlet based on multi-sensor data fusion and dynamic correction algorithm.

[0024] Figure 5 This is a structural block diagram of an air inlet temperature correction device provided by one embodiment of the present invention;

[0025] Figure 6 FIG. 1 is a schematic diagram of an electronic device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0026] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of them. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0027] Please refer to Figure 1 , Figure 1 FIG. 1 is a flowchart of a method for correcting air inlet temperature according to an embodiment of the present invention. The method for correcting air inlet temperature according to this embodiment may include the following steps:

[0028] Step S11: Acquire first sensor data corresponding to a first temperature sensor and second sensor data corresponding to a second temperature sensor inside a target network device; the first temperature sensor and the second temperature sensor are internal temperature sensors located at different positions inside the target network device.

[0029] In this embodiment, the target network device is the network device for which air inlet temperature prediction is to be performed, and the first temperature sensor and the second temperature sensor are internal temperature sensors located at different locations within the target network device. That is, the first temperature sensor and the second temperature sensor are internal temperature sensors located at different locations within the target network device. In this embodiment, first sensor data corresponding to the first temperature sensor within the target network device can be obtained, and second sensor data corresponding to the second temperature sensor within the target network device can be obtained.

[0030] Step S12: inputting the first sensor data and the second sensor data into a temperature correction model to obtain a temperature correction value corresponding to the target network device.

[0031] In this embodiment, a temperature correction model is pre-trained. This temperature correction model is used to predict a temperature correction value corresponding to a target network device based on first sensor data and second sensor data of the target network device. The temperature correction model is trained based on first sample sensor data corresponding to a first sample temperature sensor within the sample network device, second sample sensor data corresponding to a second sample temperature sensor within the sample network device, and a sample external temperature corresponding to the sample network device. The sample network device in this embodiment is a network device used to train the temperature correction model; the first sample temperature sensor is the first temperature sensor within the sample network device used for model training; the second sample temperature sensor is the second temperature sensor within the sample network device used for model training; and the sample external temperature is the actual temperature of the external environment in which the sample network device is located.

[0032] Step S13: determining the air inlet temperature of the target network device according to the first sensing data and the temperature correction value.

[0033] In this embodiment, after obtaining the temperature correction value output by the temperature correction model, the air inlet temperature of the target network device can be determined based on the first sensor data corresponding to the target network device and the temperature correction value, thereby achieving accurate prediction of the air inlet temperature of the target network device.

[0034] For example, in an optional implementation, the sum of the first sensing data and the temperature correction value may be determined as the air inlet temperature of the target network device.

[0035] In response to the problem of low accuracy of the air inlet temperature caused by the subjective experience of the correction of the air inlet temperature of the network equipment in the method of determining the air inlet temperature based on a manually set fixed temperature correction value and the temperature sensor data closest to the heat dissipation air duct inside the switch in the related art, this embodiment no longer relies solely on the temperature sensor data closest to the air inlet to calculate the air inlet temperature, but fully considers the impact of different positions inside the target network equipment on the internal temperature sensor, obtains the sensor data corresponding to the first temperature sensor and the second temperature sensor inside the target network equipment respectively, and objectively and dynamically predicts the air inlet temperature deviation value based on these sensor data and the temperature correction model, thereby realizing the accurate correction of the air inlet temperature of the network equipment and improving the accuracy of the air inlet temperature of the network equipment.

[0036] In combination with the above embodiments, in one embodiment, the present invention further provides an air inlet temperature correction method. In this method, in addition to the above steps, step S21 may also be included:

[0037] Step S21: dividing the internal temperature sensor in the target network device into the first temperature sensor and the second temperature sensor according to the relevant information of the heat dissipation duct in the target network device.

[0038] In this embodiment, the internal temperature sensors in the target network device can be divided into first temperature sensors and second temperature sensors based on the relevant information about the heat dissipation duct in the target network device. The relevant information about the heat dissipation duct in this embodiment includes at least one of the following: the location of the heat dissipation duct, the location of the heat-generating component, and the heat dissipation direction.

[0039] In this embodiment, when determining the air inlet temperature, we no longer only consider the temperature value of the temperature sensor closest to the air inlet inside the target network device, but instead consider the impact of the heat dissipation duct inside the target network device on the internal temperature sensor. Based on the relevant information of the heat dissipation duct, the internal temperature sensor inside the target network device is divided into a first temperature sensor and a second temperature sensor, and the sensor data of the two types of temperature sensors are included in the determination of the air inlet temperature according to different weights, so as to further obtain the accurate air inlet temperature dynamically in real time to ensure that the heat dissipation strategy based on the air inlet temperature is more reasonable.

[0040] In conjunction with the above embodiments, in one embodiment, the present invention further provides an air inlet temperature correction method, wherein a first temperature sensor divided based on relevant information of the heat dissipation duct is closer to the heat dissipation duct inside the target network device than a second temperature sensor.

[0041] When the relevant information about the heat dissipation duct only includes the location of the heat dissipation duct, the first temperature sensor can be an internal temperature sensor located within a first threshold range from the location of the heat dissipation duct, and the second temperature sensor can be an internal temperature sensor located outside the first threshold range from the location of the heat dissipation duct. It will be understood that in this embodiment, the internal temperature sensor close to the heat dissipation duct is the first temperature sensor, and the internal temperature sensor farther from the heat dissipation duct is the second temperature sensor. The first threshold range represents the maximum threshold range indicating the proximity of the temperature sensor to the heat dissipation duct, and the specific value can be flexibly set based on experience or needs.

[0042] In the case where the relevant information of the heat dissipation duct only includes the location of the heat-generating component, the first temperature sensor is an internal temperature sensor arranged within the second threshold range from the location of the heat-generating component, and the second temperature sensor is an internal temperature sensor arranged outside the second threshold range from the location of the heat-generating component. It is understood that in this embodiment, the internal temperature sensor close to the heat-generating component is the first temperature sensor, and the internal temperature sensor far from the heat-generating component is the second temperature sensor. The second threshold range is the maximum threshold range that represents the proximity of the temperature sensor to the heat-generating component, and the specific value can be flexibly set based on experience or needs.

[0043] It can be understood that the internal temperature sensor in the direction of the heat dissipation air can be exposed to the outside air and is closer to the outside temperature; therefore, when the relevant information of the heat dissipation air duct only includes the heat dissipation air direction, the first temperature sensor is the internal temperature sensor in the direction of the heat dissipation air, and the second temperature sensor is the internal temperature sensor not in the direction of the heat dissipation air.

[0044] In addition, when classifying internal temperature sensors, the relevant information of the heat dissipation duct that needs to be considered is the following: the position of the heat dissipation duct, the position of the heat-generating component, and the heat dissipation wind direction. As long as an internal temperature sensor is determined to be the first temperature sensor based on any one of the multiple items, the internal temperature sensor is determined to be the first temperature sensor.

[0045] In conjunction with the above embodiments, in one embodiment, the present invention further provides an air inlet temperature correction method. In this method, the "obtaining first sensor data corresponding to the first temperature sensor inside the target network device and second sensor data corresponding to the second temperature sensor" in the above step S11 may specifically include steps S41 to S43:

[0046] Step S41: obtaining first temperatures corresponding to a plurality of the first temperature sensors, and obtaining second temperatures corresponding to a plurality of the second temperature sensors.

[0047] In this embodiment, the target network device includes a plurality of first temperature sensors and a plurality of second temperature sensors, and the first temperatures corresponding to the plurality of first temperature sensors and the second temperatures corresponding to the plurality of second temperature sensors can be obtained.

[0048] Step S42: Determine the first weights corresponding to the plurality of first temperature sensors respectively based on the sensor accuracy, location and relationship between the first temperature sensor and the heat dissipation duct; and determine the second weights corresponding to the plurality of second temperature sensors respectively based on the sensor accuracy, location and relationship between the second temperature sensor and the heat dissipation duct.

[0049] In this embodiment, taking into account the impact of external factors such as the cooling air duct, cooling air direction, and temperature sensor accuracy on the ambient temperature, first weights corresponding to the plurality of first temperature sensors are determined based on the sensor accuracy corresponding to the first temperature sensor, the location of the first temperature sensor, and the relationship between the first temperature sensor and the cooling air duct. Furthermore, second weights corresponding to the plurality of second temperature sensors are determined based on the sensor accuracy corresponding to the second temperature sensor, the location of the second temperature sensor, and the relationship between the second temperature sensor and the cooling air duct.

[0050] Step S43: determining the first sensor data based on the first temperatures and the first weights respectively corresponding to the plurality of first temperature sensors; and determining the second sensor data based on the second temperatures and the second weights respectively corresponding to the plurality of second temperature sensors.

[0051] In this embodiment, after obtaining the first temperature and the first weight corresponding to each first temperature sensor, the first sensor data corresponding to the first temperature sensor may be determined based on the first temperatures and the first weights respectively corresponding to the plurality of first temperature sensors.

[0052] In an optional example, the first sensor data may be obtained using a weighted average algorithm, for example: the first sensor data T is determined based on the first formula in :

[0053] First formula:

[0054] Where m is the number of first temperature sensors, t j is the first temperature of the jth first temperature sensor, w j is the first weight of the j-th first temperature sensor.

[0055] Furthermore, after obtaining the second temperature and the second weight corresponding to each second temperature sensor, this embodiment may determine the second sensor data corresponding to the second temperature sensor based on the second temperatures and the second weights respectively corresponding to the plurality of second temperature sensors.

[0056] In an optional example, the second sensor data may be obtained using a weighted average algorithm, for example: the second sensor data T is determined based on the second formula exn :

[0057] Second formula:

[0058] Where u is the number of the second temperature sensor, t k is the second temperature of the kth second temperature sensor, w kis the second weight of the kth second temperature sensor. In one example, m may be 3, and u may be 15.

[0059] In this embodiment, for multiple first temperature sensors and multiple second temperature sensors, each first temperature sensor and second temperature sensor can be assigned a weight based on the sensor accuracy, location, and relationship with the heat dissipation duct, so as to determine accurate first sensor data and second sensor data based on the weights and temperatures corresponding to each of the first temperature sensor and the second temperature sensor.

[0060] In combination with the above embodiments, in one embodiment, the present invention further provides an air inlet temperature correction method. In this method, the training step of the temperature correction model includes at least steps S51 to S57:

[0061] Step S51: obtaining a sample external temperature corresponding to the sample network device.

[0062] In this embodiment, when performing model training, it is necessary to first obtain multiple sample network devices and obtain sample external temperatures corresponding to the sample network devices, wherein the sample external temperature is the actual temperature of the external environment in which the sample network devices are located.

[0063] Step S52: acquiring first sample sensor data corresponding to a first sample temperature sensor inside the sample network device at the sample external temperature of the sample network device.

[0064] Step S53: obtaining a sample temperature deviation value corresponding to the sample network device based on the first sample sensor data and the sample external temperature.

[0065] Step S54: obtaining second sample sensor data corresponding to a second sample temperature sensor inside the sample network device at the sample external temperature of the sample network device.

[0066] In this embodiment, first sample sensor data corresponding to a first sample temperature sensor inside the sample network device at a sample external temperature may also be obtained; and second sample sensor data corresponding to a second sample temperature sensor inside the sample network device at a sample external temperature may also be obtained. The first sample temperature sensor is the first temperature sensor used for model training, and the second sample temperature sensor is the second temperature sensor used for model training. The first sample sensor data is the first sensor data corresponding to the first sample temperature sensor, and the second sample sensor data is the second sensor data corresponding to the second sample temperature sensor.

[0067] In this embodiment, the division of the first sample temperature sensor and the second sample temperature sensor is the same as or similar to the aforementioned division method of the first temperature sensor and the second temperature sensor; the method for obtaining the first sample sensor data and the second sample sensor data is the same as or similar to the aforementioned method for obtaining the first sensor data and the second sensor data, and is not further described.

[0068] That is, in this embodiment, the acquired sample external temperature corresponding to the sample network device, the first sample sensor data corresponding to the sample network device, and the second sample sensor data corresponding to the sample network device are three temperature data corresponding to the sample network device at the same time.

[0069] In this embodiment, after obtaining the first sample sensor data and the sample external temperature corresponding to the sample network device, the sample temperature deviation value corresponding to the sample network device can be obtained based on the first sample sensor data and the sample external temperature. For example, the first sample sensor data T in Subtract the sample external temperature T out , get the sample temperature deviation value ΔT: ΔT = T in -T out .

[0070] Step S55: Determine the first sample sensor data, the second sample sensor data, and the sample temperature deviation value as a set of training data.

[0071] In this embodiment, the determined first sample sensor data, second sample sensor data, and sample temperature deviation value are determined as a set of training data.

[0072] Step S56: cyclically acquiring multiple sets of training data for the multiple sample network devices to obtain a training sample library.

[0073] In this embodiment, multiple sets of training data are acquired for multiple sample network devices based on the aforementioned steps S51 to S54, and a training sample library is constructed based on these multiple sets of training data. In other words, in this embodiment, the first sample sensor data, the second sample sensor data, and the sample temperature deviation value acquired for each sample network device at each moment are determined as one set of training data, thereby obtaining multiple sets of training data for training the initial model.

[0074] Step S57: Using the first sample sensor data and the second sample sensor data in the training sample library as training samples, and the sample temperature deviation values ​​in the training sample library as labels, the initial model is trained based on a support vector regression algorithm until the initial model converges to obtain the temperature correction model.

[0075] In this embodiment, after obtaining the training sample library, the first sample sensor data T in a set of training data can be in and the second sample sensor data T exn As a set of training samples, that is, the input feature vector: X i =[T exni ,T ini ]; The sample temperature deviation value in the training data set is used as the label of the training sample, that is, the target vector: y i =ΔT i Where i is the number of training sample groups, that is, the number of times the temperature was collected. Then, based on the training samples and the corresponding labels, the initial model is trained using the support vector regression algorithm until the initial model converges to obtain the temperature correction model.

[0076] In an optional example, a Support Vector Regression (SVR) model may be determined as the initial model, and a decision function of a Gaussian kernel function may be used as the loss function of the initial model:

[0077]

[0078] Among them, the Gaussian kernel function is defined as: K(X i ,X)=expπ(-γ||X i -X|| 2 ).

[0079] In this embodiment, n is the total number of training data groups in the training sample library, the kernel function parameter γ = 0.5, the bias term b = 0.5, exp(x) is the exponential operator, that is, e to the power of x; X i is the input feature vector, X is the number of training sample libraries except X i Other input feature vectors. Lagrange multiplier α i and The values ​​of can be determined by solving an optimization problem, which can be obtained by transforming the original optimization problem into a dual problem using the Lagrange dual method.

[0080] For SVR, the original optimization problem is:

[0081]

[0082]

[0083] By using the Lagrange multiplier method, this problem can be transformed into its dual form:

[0084]

[0085] in, In this embodiment, n is set to 300, and the penalty parameter C=1.0. After the above transformation, the Lagrange multiplier can be obtained through a standard convex quadratic programming problem solver.

[0086] In combination with the above embodiments, in one embodiment, during the training of the initial model, the first sample sensor data and the second sample sensor data as a group of training samples respectively correspond to different levels of importance, so that in the process of predicting the temperature deviation value, the influence of the heat dissipation duct inside the sample network device on the internal temperature is fully considered to predict a more accurate temperature deviation value. Furthermore, during the initial model training process, the weight of the first sample sensor data as a training sample can be set to be greater than the weight of the second sample sensor data as a training sample, so as to train the initial model based on the first sample sensor data and its corresponding weight, and the second sample sensor data and its corresponding weight. Specifically, during the model training process, the specific weights corresponding to the first sample sensor data and the second sample sensor data can be flexibly and freely set based on manual experience or project requirements, and there is no restriction on the numerical value of the specific weight.

[0087] In combination with the above embodiments, in one embodiment, the present invention further provides an air inlet temperature correction method. In this method, the above step S51 may specifically include steps S61 to S64:

[0088] Step S61: Arrange a plurality of external temperature sensors around the sample network device according to the volume and placement position of the sample network device.

[0089] In this embodiment, multiple external temperature sensors can be arranged at different locations around the sample network device based on the size and placement of the sample network device. The number of arranged external temperature sensors is not specifically limited, and can be, for example, 10 external temperature sensors.

[0090] Step S62: Acquire third temperatures corresponding to the plurality of external temperature sensors.

[0091] In this embodiment, sensor data corresponding to a plurality of external temperature sensors may be collected respectively to obtain third temperatures corresponding to the plurality of external temperature sensors.

[0092] Step S63: determining third weights corresponding to the plurality of external temperature sensors respectively according to historical data accuracy, location importance, and reliability of the external temperature sensors.

[0093] In this embodiment, the third weight corresponding to each of the plurality of external temperature sensors may be determined based on historical data accuracy of the external temperature sensor, location importance of the external temperature sensor, and reliability of the external temperature sensor.

[0094] Step S64: Determine a sample external temperature corresponding to the sample network device based on the third temperatures and the third weights respectively corresponding to the plurality of external temperature sensors.

[0095] In this embodiment, after obtaining the third temperature and the third weight corresponding to each external temperature sensor, the sample external temperature corresponding to the sample network device can be determined based on the third temperatures and third weights respectively corresponding to the multiple external temperature sensors.

[0096] In an optional example, a weighted average algorithm may be used to obtain the sample external temperature. For example, the sample external temperature T is determined based on the third formula: out :

[0097] The third formula:

[0098] Where n is the number of external temperature sensors, t i is the third temperature of the i-th external temperature sensor, w i is the third weight of the i-th external temperature sensor. In one example, n may be 10.

[0099] In this embodiment, it is ensured that the sum of all weights is not zero (i.e. ); when determining the weight, it should be set according to the actual situation. For example, if the location of some sensors can better reflect the ambient temperature, they should be given a higher weight; if the historical data of some sensors is more accurate, they should also be given a higher weight.

[0100] In this embodiment, in order to address the defects in the related art of inaccurate detection of the external ambient temperature of network equipment and the inability of temperature data collected by a single temperature sensor to accurately reflect the actual temperature outside the network equipment, an algorithm based on multi-sensor data fusion is proposed. By collecting ambient temperature data at different locations around the network equipment, the sensor weight is determined according to factors such as the historical data accuracy, location importance, and reliability of the temperature sensor, and the data fusion technology is used to correct the peripheral temperature of the network equipment to make it more accurately reflect the ambient temperature and obtain a more accurate external temperature value.

[0101] In addition, in one embodiment, after collecting the sensing data of each sensor (such as the first temperature, the second temperature, and the third temperature), before determining the first sensor data, the second sensor data, or the external temperature of the sample, these data can be preprocessed to remove abnormal data or perform data smoothing to improve the accuracy of the results.

[0102] In combination with the above embodiments, in one embodiment, the present invention further provides an air inlet temperature correction method. In this method, the above step S56 may specifically include steps S71 to S73:

[0103] Step S71: setting a plurality of temperature intervals based on the working environment and working status of the sample network device.

[0104] In this embodiment, multiple temperature intervals can be set based on the operating environment and operating state of the sample network device. It will be understood that the multiple temperature intervals set in this embodiment are temperature intervals when the network device is in an operating state (including normal operating state and abnormal operating state). In an optional embodiment, the multiple temperature intervals set can be multiple consecutive temperature intervals. This embodiment does not impose any restrictions on the specific number of temperature intervals.

[0105] For example, three temperature intervals can be set, in units of ° C: (T0, T1), (T1, T2), (T2, T3); for example, T0=10, T1=23, T2=35, T3=50.

[0106] Step S72: for each temperature interval in the multiple temperature intervals, when the sample external temperature corresponding to the sample network device is within the temperature interval, obtain multiple sets of training data for each of the multiple sample network devices.

[0107] In this embodiment, for each of the multiple temperature intervals, when the sample external temperature corresponding to the sample network device is within each temperature interval, multiple sets of training data are obtained for each of the multiple sample network devices.

[0108] That is, this embodiment collects multiple sets of training data for the sample external temperature corresponding to the sample network device in each temperature interval. For example, for the first temperature interval, a sample external temperature corresponding to a sample network device in the first temperature interval is obtained, and then the first sample sensor data and the second sample sensor data corresponding to the sample network device at the sample external temperature are obtained. Based on the sample external temperature and the first sample sensor data, a sample temperature deviation value is obtained, thereby obtaining a set of training data for the first temperature interval (first sample sensor data, second sample sensor data, and sample temperature deviation value). In an optional embodiment, a certain amount of training data can be collected for each temperature interval, and the number of sets of training data corresponding to each temperature interval can be the same or different. For example, 100 sets of training data are collected for each temperature interval, and the training database includes a total of 300 sets of training data.

[0109] Step S73: obtaining the training sample library based on multiple sets of training data corresponding to multiple sample network devices.

[0110] In this embodiment, a training sample library may be constructed based on multiple sets of training data corresponding to the multiple sample network devices.

[0111] In this embodiment, the temperature correction model is trained based on the training data collected in different temperature ranges, so that the accurate air inlet temperature can be dynamically obtained in real time in different temperature ranges, thereby ensuring that the heat dissipation strategy based on the air inlet temperature is more reasonable, while reducing power consumption and meeting the heat dissipation requirements while ensuring stable and safe operation of network equipment.

[0112] In conjunction with the above embodiments, this embodiment provides an air inlet temperature correction method. In this method, the target network device or the sample network device includes any of the following: a switch, a server, and a computer. In other words, this embodiment can be applied to the heat dissipation technology of switches, servers, mainframe PCs, or personal PCs. During operation, these network devices can achieve an air inlet temperature that is closer to the external temperature by dynamically correcting the air duct temperature in real time.

[0113] like Figure 2 As shown, Figure 2 This is a method for correcting the air inlet temperature of a switch provided in the related art. Figure 2In this method, the internal temperature sensor data of the switch is first collected, and then the temperature value is segmented. Then, the corresponding correction value is set according to the temperature segmentation interval. For example, the temperature sensor data range is divided into three intervals, that is, a fixed correction value ΔTx (x = 1, 2, 3) is set in each temperature interval. Finally, based on the collected internal temperature sensor data T of the switch, according to the correction value corresponding to the temperature interval to which the T value belongs, the arithmetic operation of the temperature value and the correction value is performed (T + ΔTx (x = 1, 2, 3)) to obtain the switch inlet temperature.

[0114] It can be seen that there are at least two flaws in the relevant calculations: First, the actual theoretical value of the switch's external ambient temperature is obtained only through a single temperature sensor. Considering that the data collected by the sensor varies depending on its installation location, and the accuracy of temperature sensors of different models varies, the collected temperature data cannot accurately reflect the actual external temperature of the switch. Second, although the relevant technology has corrected the internal temperature sensor data of the switch to a certain extent, so that the corrected data is as close as possible to the external ambient temperature of the switch, the selection and setting of the correction value ΔTx (x = 1, 2, 3) and the division of the temperature range are subjective and empirical. For example, the influence and interference of various factors such as the cabinet location, the heat dissipation direction of the switch, the location of the heat dissipation duct, and the accuracy of the temperature sensor used are not fully considered. In some scenarios, there is a disadvantage of inaccurate correction value, which may deviate too much from the external ambient temperature of the switch, thereby causing the switch speed to be too high or too low.

[0115] Based on this, in order to solve some or all of the above technical problems, in one embodiment, an air inlet temperature collection system is provided. Figure 3 As shown, Figure 3 The present invention is a block diagram showing the principle of a switch air inlet temperature collection system according to an embodiment of the present invention. Figure 3 In the present invention, the switch air inlet temperature collection system includes at least: a switch ambient temperature collection module, a switch internal temperature collection module, a data processing and prediction module, and an air inlet temperature calculation module. The switch ambient temperature collection module is connected to at least 10 external temperature sensors, which can be placed at different locations around the switch based on the size and location of the switch. The switch internal temperature collection module is provided by the switch itself and can collect data from various internal temperature sensors within the switch. The data processing and prediction module and the air inlet temperature calculation module can be implemented using software algorithms. The data processing and prediction module is used to process the data collected by the switch ambient temperature collection module and the switch internal temperature collection module and predict the correction value using a temperature correction model. The air inlet temperature calculation module is used to calculate the switch air inlet temperature based on the predicted correction value and the internal temperature sensor data.

[0116] Furthermore, in one embodiment, Figure 4 As shown, Figure 4 This invention proposes a method for correcting the temperature of a switch's air inlet based on multi-sensor data fusion and a dynamic correction algorithm. First, the method obtains the fusion data of the switch's external temperature sensor, Tout; the fusion data of the switch's internal duct temperature sensor (a first sample temperature sensor, Tin); and the fusion data of the switch's internal non-duct temperature sensor (a second sample temperature sensor, Texn). Then, based on the fusion data of the switch's external temperature sensor, Tout, and the fusion data of the switch's internal duct temperature sensor, Tin, the deviation between the duct temperature and the theoretical external ambient temperature is calculated: ΔT = Tin - Tout. Then, based on Tin, Texn, and ΔT, a sample library is established to determine whether the number of samples reaches a preset threshold, N. If not, sample data is continuously acquired. If so, an SVR model is trained based on the sample library to obtain a trained temperature correction model. Finally, when predicting the switch's air inlet temperature, we first obtain the fusion data Texn' of the switch's internal non-duct temperature sensor and the fusion data Tin' of the switch's internal duct temperature sensor. Then, we predict the compensation temperature △T' based on Texn', Tin', and the SVR model. Then, based on the compensation temperature △T' and the fusion data Tin' of the switch's internal duct temperature sensor, we calculate the switch's air inlet temperature: T = Tin' + △T'.

[0117] The air inlet temperature correction method provided in this embodiment collects ambient temperature data at different locations around the switch device, and uses data fusion technology to correct the peripheral temperature of the switch so that it more accurately reflects the ambient temperature. This method can shield the influence of external factors such as cabinet position, switch air duct, heat dissipation direction, and temperature sensor accuracy on the ambient temperature from the data source to the greatest extent possible; at the same time, inside the switch, according to the sensor offset position in the heat dissipation air duct, it is divided into near air duct sensor and far air duct sensor, and the data of the near air duct temperature sensor and the far air duct temperature sensor are cyclically collected to establish a sample library and use the vector regression algorithm to train the SVR model. The temperature correction deviation is dynamically predicted in real time by the SVR model, and finally the switch air inlet temperature close to the actual ambient temperature of the switch is obtained. This technical solution avoids the problem of inaccurate detection of the switch's external ambient temperature. The temperature data collected by a single temperature sensor cannot accurately reflect the actual temperature outside the switch. When the inlet temperature is obtained by correcting the data of the internal (near) air duct temperature sensor of the switch, the correction deviation cannot change dynamically according to the internal and external temperatures. The accurate inlet temperature can be obtained in real time in different temperature ranges, thereby ensuring that the heat dissipation strategy based on the air inlet temperature is more reasonable, while reducing power consumption, it can not only meet the heat dissipation requirements but also ensure the stable and safe operation of the switch.

[0118] It should be noted that for the sake of simplicity, the method embodiments are described as a series of actions. However, those skilled in the art should be aware that the embodiments of the present invention are not limited by the order of the actions described, because according to the embodiments of the present invention, certain steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in this specification are all preferred embodiments, and the actions involved are not necessarily required by the embodiments of the present invention.

[0119] Based on the same inventive concept, an embodiment of the present invention provides an air inlet temperature correction device 500. Figure 5 , Figure 5 : is a structural block diagram of an air inlet temperature correction device provided by an embodiment of the present invention. The air inlet temperature correction device 500 includes:

[0120] A data acquisition module 501 is configured to acquire first sensor data corresponding to a first temperature sensor and second sensor data corresponding to a second temperature sensor inside a target network device; the first temperature sensor and the second temperature sensor are internal temperature sensors located at different positions inside the target network device;

[0121] A model prediction module 502 is configured to input the first sensor data and the second sensor data into a temperature correction model to obtain a temperature correction value corresponding to the target network device;

[0122] a temperature correction module 503, configured to determine the air inlet temperature of the target network device based on the first sensing data and the temperature correction value;

[0123] In which, the temperature correction model is obtained based on the first sample sensor data corresponding to the first sample temperature sensor inside the sample network device, the second sample sensor data corresponding to the second sample temperature sensor inside the sample network device, and the sample external temperature corresponding to the sample network device; the first sample temperature sensor is the first temperature sensor inside the sample network device used for model training, the second sample temperature sensor is the second temperature sensor inside the sample network device used for model training, and the sample external temperature is the actual temperature of the external environment in which the sample network device is located.

[0124] Optionally, the apparatus 500 further includes:

[0125] a sensor determination module, configured to classify an internal temperature sensor in the target network device into a first temperature sensor and a second temperature sensor based on relevant information of a heat dissipation duct in the target network device;

[0126] The relevant information of the heat dissipation duct includes at least one of the following: the position of the heat dissipation duct, the position of the heat-generating component, and the heat dissipation direction.

[0127] Optionally, when the relevant information of the heat dissipation duct only includes the position of the heat dissipation duct, the first temperature sensor is an internal temperature sensor arranged within a first threshold range from the position of the heat dissipation duct, and the second temperature sensor is an internal temperature sensor arranged outside the first threshold range from the position of the heat dissipation duct;

[0128] When the relevant information of the heat dissipation duct includes only the position of the heat-generating component, the first temperature sensor is an internal temperature sensor arranged within a second threshold range from the position of the heat-generating component, and the second temperature sensor is an internal temperature sensor arranged outside the second threshold range from the position of the heat-generating component;

[0129] When the relevant information of the heat dissipation air duct only includes the heat dissipation air direction, the first temperature sensor is an internal temperature sensor in the heat dissipation air direction, and the second temperature sensor is an internal temperature sensor not in the heat dissipation air direction.

[0130] Optionally, the target network device includes a plurality of the first temperature sensors and a plurality of the second temperature sensors; the data acquisition module 501 includes:

[0131] a first temperature determining module, configured to obtain first temperatures respectively corresponding to the plurality of first temperature sensors, and to obtain second temperatures respectively corresponding to the plurality of second temperature sensors;

[0132] a first weight determination module for determining first weights corresponding to each of the plurality of first temperature sensors based on the sensor accuracy, location, and relationship between the first temperature sensor and the heat dissipation duct; and determining second weights corresponding to each of the plurality of second temperature sensors based on the sensor accuracy, location, and relationship between the second temperature sensor and the heat dissipation duct;

[0133] The sensor data determination module is configured to determine the first sensor data based on the first temperatures and the first weights respectively corresponding to the plurality of first temperature sensors; and to determine the second sensor data based on the second temperatures and the second weights respectively corresponding to the plurality of second temperature sensors.

[0134] Optionally, the apparatus 500 further includes a model training module, which is used to train the temperature correction model. The model training module at least includes:

[0135] a sample temperature determination module, configured to obtain a sample external temperature corresponding to the sample network device;

[0136] a first data determining module, configured to obtain first sample sensor data corresponding to a first sample temperature sensor inside the sample network device at the sample external temperature of the sample network device;

[0137] a deviation value determining module, configured to obtain a sample temperature deviation value corresponding to the sample network device based on the first sample sensor data and the sample external temperature;

[0138] a second data determination module, configured to obtain second sample sensor data corresponding to a second sample temperature sensor inside the sample network device when the sample network device is at the sample external temperature;

[0139] a training data determining module, configured to determine the first sample sensor data, the second sample sensor data, and the sample temperature deviation value as a set of training data;

[0140] A training sample generation module is used to cyclically acquire multiple sets of training data for multiple sample network devices to obtain a training sample library;

[0141] A model parameter adjustment module is used to use the first sample sensor data and the second sample sensor data in the training sample library as training samples, and the sample temperature deviation values ​​in the training sample library as labels, to train the initial model based on a support vector regression algorithm until the initial model converges, thereby obtaining the temperature correction model.

[0142] Optionally, the sample temperature determination module includes:

[0143] a sensor setting module, configured to arrange a plurality of external temperature sensors around the sample network device according to the volume and placement of the sample network device;

[0144] A second temperature determination module is used to obtain third temperatures corresponding to the plurality of external temperature sensors respectively;

[0145] a second weight determination module, configured to determine third weights corresponding to the plurality of external temperature sensors respectively according to historical data accuracy, location importance, and reliability of the external temperature sensors;

[0146] The external sensor data determination module is configured to determine a sample external temperature corresponding to the sample network device based on the third temperatures and the third weights respectively corresponding to the plurality of external temperature sensors.

[0147] Optionally, the training sample generation module includes:

[0148] A temperature division module, configured to set a plurality of temperature intervals based on the working environment and working status of the sample network device;

[0149] a training data acquisition module, configured to acquire, for each of the plurality of temperature intervals, a plurality of sets of training data for each of the plurality of sample network devices when the sample external temperature corresponding to the sample network device is within the temperature interval;

[0150] The training sample library generation module is used to obtain the training sample library based on multiple groups of training data corresponding to multiple sample network devices.

[0151] Optionally, the target network device or the sample network device includes any one of the following: a switch, a server, and a computer.

[0152] Based on the same inventive concept, another embodiment of the present invention provides an electronic device 600, such as Figure 6 shown. Figure 6 This is a schematic diagram of an electronic device according to an embodiment of the present invention. The electronic device includes a processor 601, a memory 602, and a computer program stored in the memory 602 and executable on the processor 601. When executed by the processor, the computer program implements the steps of the air inlet temperature correction method described in any of the above embodiments of the present invention.

[0153] Based on the same inventive concept, another embodiment of the present invention provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, it implements the steps in the air inlet temperature correction method as described in any of the above embodiments of the present invention.

[0154] Based on the same inventive concept, another embodiment of the present invention provides a computer program product, including a computer program / instruction, which, when executed by a processor, implements the steps of the air inlet temperature correction method as described in any of the above embodiments of the present invention.

[0155] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referenced to each other.

[0156] It should be noted that, in this document, the terms "comprises," "includes," or any other variations thereof are intended to encompass non-exclusive inclusion, such that a process, method, article, or apparatus comprising a series of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of other identical elements in the process, method, article, or apparatus comprising the element.

[0157] Through the description of the above embodiments, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus the necessary general hardware platform, and of course can also be implemented by hardware, but in many cases the former is a better embodiment. Based on this understanding, the technical solution of the present invention, or the part that contributes to the relevant technology, can be embodied in the form of a software product. The computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk), and includes a number of instructions for enabling a terminal (which can be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in each embodiment of the present invention.

[0158] 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 guidance 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 of which are protected by the present invention.

[0159] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed in the embodiments of the present invention can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0160] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0161] In the embodiments provided by the present invention, it should be understood that the disclosed systems and methods can be implemented in other ways. For example, the system embodiments described above are merely illustrative. For example, the division of the units is merely a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interface, device or unit, which can be electrical, mechanical or other forms.

[0162] The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected to achieve the purpose of this embodiment according to actual needs.

[0163] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0164] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the relevant technology, or part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes various media that can store program code, such as a USB flash drive, a mobile hard drive, ROM, RAM, a magnetic disk, or an optical disk.

[0165] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any modifications or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be based on the scope of protection of the claims.

Claims

1. A method for correcting air inlet temperature, characterized in that: The method comprises: Acquire first sensor data corresponding to a first temperature sensor inside a target network device and second sensor data corresponding to a second temperature sensor; the first temperature sensor and the second temperature sensor are internal temperature sensors located at different positions inside the target network device; Inputting the first sensor data and the second sensor data into a temperature correction model to obtain a temperature correction value corresponding to the target network device; Determining the air inlet temperature of the target network device according to the first sensing data and the temperature correction value; Among them, the temperature correction model is obtained based on the first sample sensor data corresponding to the first sample temperature sensor inside the sample network device, the second sample sensor data corresponding to the second sample temperature sensor inside the sample network device, and the sample external temperature corresponding to the sample network device; the first sample temperature sensor is the first temperature sensor inside the sample network device used for model training, the second sample temperature sensor is the second temperature sensor inside the sample network device used for model training, and the sample external temperature is the actual temperature of the external environment in which the sample network device is located.

2. The method according to claim 1, characterized in that The method further comprises: According to the relevant information of the heat dissipation duct in the target network device, the internal temperature sensor in the target network device is divided into the first temperature sensor and the second temperature sensor; The relevant information of the heat dissipation duct includes at least one of the following: the position of the heat dissipation duct, the position of the heat generating component, and the heat dissipation direction.

3. The method according to claim 2, characterized in that In a case where the relevant information of the heat dissipation duct only includes the position of the heat dissipation duct, the first temperature sensor is an internal temperature sensor arranged within a first threshold range from the position of the heat dissipation duct, and the second temperature sensor is an internal temperature sensor arranged outside the first threshold range from the position of the heat dissipation duct; In a case where the relevant information of the heat dissipation duct only includes the position of the heat-generating component, the first temperature sensor is an internal temperature sensor arranged within a second threshold range from the position of the heat-generating component, and the second temperature sensor is an internal temperature sensor arranged outside the second threshold range from the position of the heat-generating component; When the relevant information of the heat dissipation air duct only includes the heat dissipation air direction, the first temperature sensor is an internal temperature sensor in the heat dissipation air direction, and the second temperature sensor is an internal temperature sensor not in the heat dissipation air direction.

4. The method according to claim 1, characterized in that: The target network device includes a plurality of the first temperature sensors and a plurality of the second temperature sensors inside the target network device; and obtaining first sensor data corresponding to the first temperature sensor inside the target network device and second sensor data corresponding to the second temperature sensor inside the target network device includes: Acquire first temperatures respectively corresponding to a plurality of the first temperature sensors, and acquire second temperatures respectively corresponding to a plurality of the second temperature sensors; Determine first weights corresponding to the plurality of first temperature sensors respectively according to the sensor accuracy, the position of the first temperature sensor, and the relationship between the first temperature sensor and the heat dissipation duct; and determine second weights corresponding to the plurality of second temperature sensors respectively according to the sensor accuracy, the position of the second temperature sensor, and the relationship between the second temperature sensor and the heat dissipation duct; The first sensor data is determined based on the first temperatures respectively corresponding to the plurality of first temperature sensors and the first weight; and the second sensor data is determined based on the second temperatures respectively corresponding to the plurality of second temperature sensors and the second weight.

5. The method according to claim 1, characterized in that The training step of the temperature correction model at least includes: Obtaining a sample external temperature corresponding to the sample network device; Acquire first sample sensor data corresponding to a first sample temperature sensor inside the sample network device at the sample external temperature of the sample network device; Based on the first sample sensor data and the sample external temperature, obtaining a sample temperature deviation value corresponding to the sample network device; Acquire second sample sensor data corresponding to a second sample temperature sensor inside the sample network device at the sample external temperature of the sample network device; Determine the first sample sensor data, the second sample sensor data, and the sample temperature deviation value as a set of training data; Circularly acquiring multiple sets of training data for multiple sample network devices to obtain a training sample library; The first sample sensor data and the second sample sensor data in the training sample library are used as training samples, and the sample temperature deviation values ​​in the training sample library are used as labels. The initial model is trained based on a support vector regression algorithm until the initial model converges to obtain the temperature correction model.

6. The method according to claim 5, characterized in that The obtaining of the sample external temperature corresponding to the sample network device includes: Arranging a plurality of external temperature sensors around the sample network device according to the volume and placement position of the sample network device; Acquire third temperatures respectively corresponding to the plurality of external temperature sensors; Determine third weights corresponding to the plurality of external temperature sensors respectively according to historical data accuracy, location importance, and reliability of the external temperature sensors; Based on the third temperatures and the third weights respectively corresponding to the plurality of external temperature sensors, a sample external temperature corresponding to the sample network device is determined.

7. The method according to claim 5, characterized in that The step of cyclically acquiring multiple sets of training data for multiple sample network devices to obtain a training sample library includes: Based on the working environment and working status of the sample network device, setting multiple temperature intervals; For each temperature interval in the multiple temperature intervals, when the sample external temperature corresponding to the sample network device is within the each temperature interval, respectively acquiring multiple sets of training data for each of the multiple sample network devices; The training sample library is obtained based on multiple groups of training data respectively corresponding to multiple sample network devices.

8. The method according to any one of claims 1 to 7, characterized in that: The target network device or the sample network device includes any one of the following: a switch, a server, and a computer.

9. An air inlet temperature correction device, characterized in that: The device comprises: A data acquisition module, used to acquire first sensor data corresponding to a first temperature sensor inside a target network device and second sensor data corresponding to a second temperature sensor; the first temperature sensor and the second temperature sensor are internal temperature sensors located at different positions inside the target network device; A model prediction module, used for inputting the first sensor data and the second sensor data into a temperature correction model to obtain a temperature correction value corresponding to the target network device; a temperature correction module, used to determine the air inlet temperature of the target network device according to the first sensing data and the temperature correction value; Among them, the temperature correction model is obtained based on the first sample sensor data corresponding to the first sample temperature sensor inside the sample network device, the second sample sensor data corresponding to the second sample temperature sensor inside the sample network device, and the sample external temperature corresponding to the sample network device; the first sample temperature sensor is the first temperature sensor inside the sample network device used for model training, the second sample temperature sensor is the second temperature sensor inside the sample network device used for model training, and the sample external temperature is the actual temperature of the external environment in which the sample network device is located.

10. An electronic device, characterized in that: include: A processor, a memory, and a computer program stored in the memory and executable on the processor, wherein the computer program, when executed by the processor, implements the steps in the air inlet temperature correction method as described in any one of claims 1 to 8.

11. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the air inlet temperature correction method according to any one of claims 1 to 8 are implemented.

Images