Air-cooled server fault detection method, electronic device, medium and program product

Abstract

The application discloses a kind of air-cooled server fault detection method, electronic equipment, medium and program product, belong to server technical field, which comprises: by three-axis disturbance coordination engine controls air-cooled server operation, and sample running data of air-cooled server is acquired regularly;Determine decision diagnosis threshold;Current running data of air-cooled server is acquired regularly;Determine current sensor change rate;According to current sensor change rate and decision diagnosis threshold, fault decision determination processing is carried out, and fault decision determination result is obtained;In response to the operation of rejudgment, current running data is rejudged, and rejudgment result is obtained;Rejudgment result is output as target fault detection result.By using sample running data to determine decision diagnosis threshold, which is finally used for fault detection, the problem of low efficiency of fault detection is solved, and the application achieves the effect of improving the fault detection efficiency of air-cooled server.

Description

TECHNICAL FIELD

[0001] The application relates to the technical field of servers, in particular to a method for detecting faults of an air-cooled server, an electronic device, a medium and a program product. BACKGROUND

[0002] With the rapid development of cloud computing and artificial intelligence, the application scale of air-cooled servers in data centers continues to expand, and the running stability of the air-cooled servers directly affects the continuity of key businesses, so it is necessary to detect the air-cooled servers.

[0003] At present, the detection of the air-cooled servers mainly relies on a passive detection method, but this method has the problem of long detection period, thus resulting in low detection efficiency. Therefore, there is an urgent need for a detection method for the air-cooled servers to solve the above technical problems. SUMMARY

[0004] The application provides a method for detecting faults of an air-cooled server, an electronic device, a medium and a program product, to at least solve the problem of low fault detection efficiency, and the application achieves the effect of improving the fault detection efficiency of the air-cooled server.

[0005] The application provides a method for detecting faults of an air-cooled server, comprising:

[0006] controlling the operation of the air-cooled server through a three-axis disturbance coordination engine, and acquiring sample operation data of the air-cooled server at regular time intervals, wherein the three-axis disturbance coordination engine is a temperature disturbance-power supply disturbance-load disturbance controller for triggering multi-coupling resonance and harmonic resonance;

[0007] determining a decision diagnosis threshold value according to the sample operation data and a preset severity coefficient;

[0008] acquiring current operation data of the air-cooled server at regular time intervals;

[0009] determining a current sensor change rate according to the current operation data;

[0010] performing fault decision judgment processing according to the current sensor change rate and the decision diagnosis threshold value to obtain a fault decision judgment result;

[0011] when it is detected that the fault decision judgment result meets a preset alarm condition, performing rejudgment on the current operation data in response to a rejudgment operation to obtain a rejudgment result;

[0012] when it is detected that the rejudgment result is a real fault, outputting the rejudgment result as a target fault detection result.

[0013] The application also provides an electronic device, comprising at least one processor and a memory.

[0014] The memory stores computer execution instructions;

[0015] The processor executes computer-executed instructions stored in the memory, so that the processor executes the steps of any one of the air-cooled server fault detection methods described above.

[0016] The application also provides a computer-readable storage medium, and the computer-readable storage medium stores a computer program, wherein the computer program is executed by the processor to implement the steps of any one of the air-cooled server fault detection methods described above.

[0017] The application also provides a computer program product, comprising a computer program, wherein the computer program is executed by the processor to implement the steps of any one of the air-cooled server fault detection methods described above.

[0018] According to the application, since the air-cooled server is controlled to run by the three-axis disturbance cooperative engine, the probability of triggering a fault is improved. Meanwhile, the sample running data of the air-cooled server is obtained at a fixed time, and a large amount of sample running data provides the required data support for determining the decision diagnosis threshold value. Then, the decision diagnosis threshold value is determined based on the sample running data and the strictness coefficient, and is used for subsequent judgment of the fault of the air-cooled server. When the current running data obtained is judged for a fault, the current sensor change rate and the decision diagnosis threshold value are calculated for fault decision determination, and then artificial rejudgment is used for secondary determination, so as to ensure the accuracy of the fault detection result, reduce the screening and judgment work of artificial in a large amount of running data, and improve the fault detection efficiency. BRIEF DESCRIPTION OF DRAWINGS

[0019] In order to more clearly illustrate the embodiments of the application, the drawings needed in the embodiments will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the application, and other drawings can be obtained by those skilled in the art without creative labor.

[0020] Figure 1 An application scenario diagram of an air-cooled server fault detection method provided by the embodiments of the application is shown in the figure.

[0021] Figure 2 A flowchart of an air-cooled server fault detection method provided by the embodiments of the application is shown in the figure.

[0022] Figure 3 A control flowchart of a three-axis disturbance cooperative engine provided by the embodiments of the application is shown in the figure.

[0023] Figure 4 A structure diagram of an air-cooled server fault detection device provided by the embodiments of the application is shown in the figure.

[0024] Figure 5A structural schematic diagram of an electronic device provided by an embodiment of the present application is shown. DETAILED DESCRIPTION

[0025] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only some of the embodiments of the present application, but not all the embodiments of the present application. Based on the embodiments in the present application, any other embodiments obtained by a person of ordinary skill in the art without creative work fall within the protection scope of the present application.

[0026] It should be noted that, in the description of the present application, the terms “comprise”, “contain” or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device comprising a series of elements not only includes those elements, but also includes other elements not explicitly listed or inherent to such a process, method, article or device. The terms “first”, “second” and the like in the present application are used to distinguish similar objects, and are not used to describe a specific order or sequence.

[0027] In the air-cooled server fault detection process using the related technology, it is found that passive detection has hysteresis, while the present application can trigger the occurrence of air-cooled server failure through high-frequency disturbance, collect the running data in the air-cooled server failure process as samples, then calculate the failure decision threshold value using these samples, and use the real-time collected current running data and the failure decision threshold value to detect the failure in the subsequent normal operation process of the air-cooled server, quickly determine the running data that may have a failure from a large amount of running data, and perform manual rejudgment, thereby ensuring the accuracy of the fault detection result while greatly improving the fault detection efficiency.

[0028] In order for those skilled in the art to better understand the present application, the present application will be further described in detail below with reference to the drawings and specific embodiments.

[0029] In combination with the specific application environment architecture or specific hardware architecture on which the execution of the air-cooled server fault detection method provided by the embodiments of the present application depends, the specific application environment architecture or specific hardware architecture is described here. For reference Figure 1 , Figure 1 An application scenario diagram of the air-cooled server fault detection method provided by the embodiments of the present application is shown.

[0030] As Figure 1As shown, the application scenario includes a server system composed of multiple air-cooled servers. Since the running data of each air-cooled server is large, and the running data includes capacitor failure, voltage failure, current failure, fan failure, and various component failure conditions that need to be detected in real time, these failures belong to server implicit failures, and passive manual detection will have the problem of low detection efficiency. Therefore, the air-cooled server fault detection method provided by the embodiment of the present application controls the air-cooled server to run through the three-axis disturbance cooperative engine, covers the electrical-thermal-load full-link fault scene, and improves the fault triggering rate. At the same time, the sample running data of the air-cooled server is collected, the decision diagnosis threshold is determined, then the current running data of each air-cooled server is collected, the current sensor change rate is determined, and then the current sensor change rate and the decision diagnosis threshold are used as the determination result of fault detection. Finally, manual rejudgment is performed to ensure the accuracy of the fault detection result. The whole process reduces the workload of manual analysis and judgment from a large amount of running data, and improves the fault detection efficiency.

[0031] Figure 2 The flowchart of the air-cooled server fault detection method provided by the embodiment of the present application is shown in Figure 2 As shown, the embodiment of the present application provides an air-cooled server fault detection method, which is described in detail as follows.

[0032] S201: Control the air-cooled server to run through the three-axis disturbance cooperative engine, and collect the sample running data of the air-cooled server at regular intervals, wherein the three-axis disturbance cooperative engine is a temperature disturbance-power supply disturbance-load disturbance controller that triggers multi-coupling resonance and harmonic resonance.

[0033] In this embodiment, the control process of the three-axis disturbance cooperative engine can be to form a disturbance full-coverage matrix in a fixed time period through the sequence of temperature disturbance-power supply disturbance-load disturbance in a specific time period, trigger multi-coupling resonance and harmonic resonance at high frequency, so as to trigger the occurrence of the fault problem of the air-cooled server with a higher probability.

[0034] In this embodiment, the sample running data of the air-cooled server can be collected through the sensors of the air-cooled server.

[0035] Figure 3 The control flowchart of the three-axis disturbance cooperative engine provided by the embodiment of the present application is shown in

[0036] Specifically, as shown in Figure 3 In an optional embodiment of the present application, the step S201 of controlling the air-cooled server to run through the three-axis disturbance cooperative engine includes:

[0037] S201a: controlling the air-cooled server to change the heat convection power operation by the triaxial disturbance coordination engine according to the temperature disturbance type control in the first period.

[0038] In the embodiment, the first period can be a time length artificially set in advance or a time length automatically generated by the controller, such as 5 minutes, as shown in FIG. 2A. Figure 3 The control strategy of the temperature disturbance type control can be: from automatic speed to automatic speed*95%, from automatic speed*95% to automatic speed, from automatic speed to automatic speed*105%, and from automatic speed*105% to full speed, and the whole process is completed in the first period. So as to achieve the effect of forcibly changing the heat convection efficiency.

[0039] S201b: controlling the air-cooled server to simulate power switching transient impact operation by the triaxial disturbance coordination engine according to the power disturbance type control in the second period.

[0040] In the embodiment, the second period can also be a time length artificially set in advance or a time length automatically generated by the controller, such as 4 minutes, as shown in FIG. 2B. Figure 3 The control strategy of the power disturbance type control can be: from double power to single power 0, from single power 0 to double power, and from double power to single power 1, and the whole process is completed in the second period. So as to achieve the effect of simulating the power switching state transient impact.

[0041] S201c: controlling the air-cooled server to excite power device stress changes by the triaxial disturbance coordination engine according to the load disturbance type control in the third period; wherein the first period, the second period and the third period are cyclically alternated, and the values of each first period, each second period and each third period are in a preset fixed time length.

[0042] In the embodiment, the third period can also be a time length artificially set in advance or a time length automatically generated by the controller, such as 3 minutes, as shown in FIG. 2C. Figure 3 The control strategy of the load disturbance type control can be: from initial load to 30% load, from 30% load to 60% load, and from 60% load to 90% load, and the whole control process is completed in the third period.

[0043] In the embodiment, the first period, the second period and the third period are cyclically alternated, and the values of each first period, each second period and each third period are in a preset fixed time length. The preset fixed time length can be a time length artificially set in advance or a time length automatically generated by the controller, such as 1 hour.

[0044] Based on the above embodiment, in an optional embodiment of the present application, step S201a specifically includes: controlling the fan of the air-cooled server to change the heat convection power operation by the triaxial disturbance coordination engine according to the temperature disturbance type control in the first period, and the fan is sequentially from automatic speed, 95% of automatic speed, automatic speed, 105% of automatic speed to full speed.

[0045] Step S201b comprises: through the three-axis disturbance coordination engine, in the second period, controlling the two power supplies of the air-cooled server in turn from two power supply, first power supply, two power supply to the second power supply according to the power supply disturbance type, so that the air-cooled server simulates power supply switching transient impact operation.

[0046] Step S201c comprises: through the three-axis disturbance coordination engine, in the third period, controlling the load simulator of the air-cooled server in turn from thirty percent load, sixty percent load to ninety percent load according to the load disturbance type, so that the air-cooled server simulates power stress change.

[0047] Based on the above embodiment, in an optional embodiment of the present application, the three-axis disturbance coordination engine in step S201 realizes that the disturbance timing matrix is fully covered within the preset fixed time length through the preset common multiple; the air-cooled server is deployed with a temperature sensor, a voltage sensor, a current sensor and a rotating speed sensor;

[0048] Correspondingly, the current running data of the air-cooled server is acquired in step S201, comprising:

[0049] S201d: acquiring the temperature value of the air-cooled server from the temperature sensor through the intelligent platform management interface.

[0050] S201e: acquiring the voltage value of the air-cooled server from the voltage sensor through the intelligent platform management interface.

[0051] S201f: acquiring the current value of the air-cooled server from the current sensor through the intelligent platform management interface.

[0052] S201g: acquiring the rotating speed value of the air-cooled server from the rotating speed sensor through the intelligent platform management interface.

[0053] In the embodiment, the intelligent platform management interface ipmi interface is used, the ipmit interface collects the sensor values of the server, and the data of the sensor can be collected once every 10 seconds. In the embodiment, the rotating speed sensor can be a sensor for detecting the rotating speed of the fan of the air-cooled server.

[0054] Based on the above embodiment, in an optional embodiment of the present application, the preset common multiple realizes that the disturbance timing matrix is fully covered in 60 minutes, that is, LCM (5, 4, 3) = 60 minutes, until a complete test closed loop is formed, and the disturbance full coverage matrix. The preset common multiple can be a smaller common multiple. Specifically, the following table 1 is used:

[0055]

[0056] S202: determining a decision diagnosis threshold according to the sample running data and a preset strictness coefficient, wherein the preset strictness coefficient is a coefficient for artificially setting a strict or relaxed fault detection standard value according to a fault detection quality department requirement or market performance.

[0057] In this embodiment, the preset strictness coefficient is a coefficient for artificially setting a strict or relaxed fault detection standard value according to a fault detection quality department requirement or market performance. For example, the preset strictness coefficient is 110%, which means that the air-cooled server is in a busy market business period, and the fault detection standard can be appropriately relaxed.

[0058] Specifically, in an optional embodiment of the present application, the sample running data includes sensor data collected at multiple sampling time points, and correspondingly, step S202 specifically includes:

[0059] S202a: performing a change rate feature extraction process according to sensor data collected at any two adjacent sampling time points to obtain sample sensor change rate data.

[0060] In this embodiment, the any two adjacent sampling time points can be two time points that are continuous on a time line. For example, the 10th second and the 20th second are two adjacent sampling time points with a 10-second sampling interval. The sample sensor change rate data can be data representing the change of sensor data in a unit time. The change rate feature extraction process can be to calculate the sensor data corresponding to the two adjacent time points for each sensor by a change rate feature extraction formula.

[0061] S202b: determining the sample quantity according to the sample running data.

[0062] In this embodiment, the sample running data includes running data of multiple sensors, and the exact sample quantity can be obtained by dividing and recording the sensor types or time nodes in the sample running data.

[0063] S202c: determining a sample overall average value according to the sample quantity and the sample sensor change rate data.

[0064] In this embodiment, the sample overall average value can be determined by calculating the average value of the sum of the data and the sample quantity after summing the data.

[0065] Specifically, in an optional embodiment of the present application, the calculation formula of S203c is:

[0066]

[0067] In the formula, is the sample overall average value, is the i-th sample sensor change rate value, n is the sample quantity, and i is a positive integer less than or equal to n.

[0068] S202d: determining a sample population standard deviation according to the sample population mean, the sample quantity and the sample sensor change rate data.

[0069] In this embodiment, the sample population standard deviation can be determined by using the standard deviation formula.

[0070] Specifically, in an optional embodiment of the present application, the calculation formula of step S203 is:

[0071]

[0072] In the formula, SD is the sample population standard deviation, is the sample population mean, is the i-th sample sensor change rate value, n is the sample quantity, and i is a positive integer less than or equal to n.

[0073] S202e: determining a decision diagnosis threshold according to the sample population mean, the sample population standard deviation and a preset strictness coefficient.

[0074] In an optional embodiment of the present application, the decision diagnosis threshold includes an upper decision threshold and a lower decision threshold, and the calculation formula of step S203e is:

[0075]

[0076] In the formula, L is the upper decision threshold, U is the lower decision threshold, T is the preset strictness coefficient, is the sample population mean, and SD is the sample population standard deviation.

[0077] For example, when the voltage change rate value is greater than the upper threshold U, it indicates that the voltage changes too fast, and a fault warning can be triggered. Or when the current change rate value is less than the lower threshold, it indicates that the current changes too slowly.

[0078] Based on the above embodiments, the decision diagnosis threshold can be a value after adjustment. Specifically, in an optional embodiment of the present application, the step S202 of determining the decision diagnosis threshold according to the sample operation data and the preset strictness coefficient further includes:

[0079] S202f: obtaining a fault data set, the fault data set including a plurality of sets of historical fault data obtained after manual rejudgment operations.

[0080] In this embodiment, the fault data set is used to store fault data obtained after fault detection of the air-cooled server. The historical fault data refers to the fault data related to the air-cooled server saved after manual rejudgment.

[0081] S202g: selecting historical fault data from the fault data set, and calculating a historical fault parameter value according to the historical fault data.

[0082] In this embodiment, calculating the corresponding similar fault parameter value according to the historical fault data can be calculating the fault change rate parameter value of each sensor in the historical fault data, recorded as X_f.

[0083] S202f: when detecting that the historical fault parameter value is less than the decision diagnosis threshold value, taking the historical fault parameter value as a new decision diagnosis threshold value.

[0084] In this embodiment, when detecting that the historical fault parameter value X_f is less than the decision diagnosis threshold value U, it indicates that the current threshold value does not capture this fault (i.e. the threshold value is too high), and the threshold value is adjusted to X_f (taking a more stringent value, i.e. lowering the threshold value to detect the fault). Conversely, when detecting that the historical fault parameter value X_f is greater than or equal to the decision diagnosis threshold value U, the current threshold value is maintained (because the fault has been correctly detected).

[0085] S203: periodically acquiring current running data of the air-cooled server.

[0086] In this embodiment, periodically acquiring means a process of acquiring at intervals of a fixed time period or at multiple acquisition time points. The current running data refers to real-time running data of the air-cooled server at the time node of sampling.

[0087] S204: determining a current sensor change rate according to the current running data.

[0088] In this embodiment, the current sensor change rate can be calculated according to the current running data by using a change rate feature extraction formula.

[0089] Specifically, in an optional embodiment of the present application, the current running data includes voltage values, current values, temperature values, power values, rotation speed values and time points of multiple acquisition running data. Correspondingly, step S204 includes:

[0090] S204a: determining a current voltage sensor change rate according to the voltage values and the time points.

[0091] Specifically, in an optional embodiment of the present application, the calculation formula of step S204a is:

[0092] =

[0093] In the formula, is the current voltage sensor change rate, is the first time point of the adjacent two time points, V2 is a voltage value of a second time point of the two adjacent time points, V1 is a voltage value of a first time point of the two adjacent time points, V2 is a voltage value of a second time point of the two adjacent time points.

[0094] S204b: determining a current current sensor change rate according to the current value and the time point.

[0095] Specifically, in an optional embodiment of the present application, the calculation formula of step S204b is:

[0096]

[0097] In the formula, V2 is a voltage value of a second time point of the two adjacent time points, V1 is a voltage value of a first time point of the two adjacent time points, V2 is a voltage value of a second time point of the two adjacent time points.

[0098] S204c: determining a current temperature sensor change rate according to the temperature value and the time point.

[0099] Specifically, in an optional embodiment of the present application, the calculation formula of step S204c is:

[0100]

[0101] In the formula, V2 is a voltage value of a second time point of the two adjacent time points, V1 is a voltage value of a first time point of the two adjacent time points, V2 is a voltage value of a second time point of the two adjacent time points.

[0102] S204d: determining a current temperature sensor change rate according to the temperature value and the time point.

[0103] Specifically, in an optional embodiment of the present application, the calculation formula of step S204d is:

[0104]

[0105] In the formula, V2 is a voltage value of a second time point of the two adjacent time points, V1 is a voltage value of a first time point of the two adjacent time points, V2 is a voltage value of a second time point of the two adjacent time points.

[0106] S204e: determining a current temperature sensor change rate according to the temperature value and the power value.

[0107] Specifically, in an optional embodiment of the present application, the calculation formula of step S204e is:

[0108]

[0109] wherein, is the current unit power consumption temperature rise rate, is the temperature value of the first time point of the adjacent two time points, is the temperature value of the second time point of the adjacent two time points, is the power consumption value of the first time point of the adjacent two time points, is the power consumption value of the second time point of the adjacent two time points.

[0110] S204f: determining the current unit temperature fan speed change rate according to the temperature value and the speed value.

[0111] Specifically, in an optional embodiment of the present application, the calculation formula of step S204f is:

[0112]

[0113] wherein, is the current unit temperature fan speed change rate, is the temperature value of the first time point of the adjacent two time points, is the temperature value of the second time point of the adjacent two time points, is the speed value of the first time point of the adjacent two time points, is the speed value of the second time point of the adjacent two time points.

[0114] S205: performing fault decision determination processing according to the current sensor change rate and the decision diagnosis threshold value to obtain a fault decision determination result.

[0115] In the embodiment, the process of performing fault decision determination processing according to the current sensor change rate and the decision diagnosis threshold value can be to compare each value of the current sensor change rate with each value of the decision diagnosis threshold value, and when one or more values exceed the decision diagnosis threshold value, a preliminary fault conclusion of the fault decision determination result can be obtained.

[0116] Specifically, in an optional embodiment of the present application, the decision diagnosis threshold value includes a voltage diagnosis threshold value, a current diagnosis threshold value, a speed change diagnosis threshold value, a temperature change diagnosis threshold value, a temperature diagnosis threshold value, a unit power consumption diagnosis threshold value, a unit temperature rise slope diagnosis threshold value, and a unit temperature speed diagnosis threshold value; and step S205 includes:

[0117] S205a: Based on the current voltage sensor change rate, the current current sensor change rate, the voltage diagnostic threshold, and the current diagnostic threshold, a capacitor fault determination process is performed to obtain a fault decision result on whether the capacitor value has decayed.

[0118] In this embodiment, when the numerical comparison is performed and the detection is successful... >100mV / s, When the current change is less than 10mA / s, it indicates that the current changes too slowly and the voltage changes too quickly. Therefore, a fault decision result of capacitance decay can be obtained. Conversely, a fault decision result of normal capacitance can be obtained.

[0119] S205b: Based on the current speed sensor change rate and speed diagnostic threshold, the fan fault judgment process is performed to obtain the fault decision result of whether the fan bearing is worn.

[0120] In this embodiment, after numerical comparison, if ΔF / Δt < 10% / s is detected, it indicates that the speed adjustment of individual fans is too slow within the same sampling period, and a fault decision result of fan bearing wear can be obtained. Conversely, a fault decision result of no fan bearing wear can be obtained.

[0121] S205c: Based on the current voltage sensor change rate, the current temperature sensor change rate, the voltage diagnostic threshold, and the temperature diagnostic threshold, perform abnormality judgment processing of the semiconductor device voltage regulator block to obtain the fault decision judgment result of whether the semiconductor device voltage regulator block is abnormal.

[0122] In this embodiment, after numerical comparison, if ΔV / Δt > 0.1V and ΔT / Δt > 1℃ / s are detected, it indicates that the voltage and temperature changes are too rapid, thus obtaining a fault decision result indicating an abnormality in the semiconductor device's voltage regulator. Conversely, if the values ​​are not found, a fault decision result indicating that the semiconductor device's voltage regulator is functioning normally can be obtained. In this embodiment, the semiconductor device can be an integrated circuit component.

[0123] S205d: Based on the temperature value, the current temperature rise rate per unit power consumption, the temperature diagnostic threshold, and the temperature rise slope diagnostic threshold, the radiator fault judgment process is performed to obtain the fault decision result of whether the radiator has poor contact.

[0124] In this embodiment, after numerical comparison, if ΔT > 25℃ and the current unit power consumption temperature rise rate > 3℃ / W, it indicates that the temperature difference between the core and the casing of the air-cooled server is large, and the power consumption temperature rise rate is high. Therefore, a fault decision result indicating poor heat sink contact can be obtained. Conversely, a fault decision result indicating normal heat sink contact can be obtained.

[0125] S206: When it is detected that the fault decision determination result meets the preset alarm condition, the current running data is re-judged in response to the re-judgment operation to obtain a re-judgment result.

[0126] In this embodiment, the preset alarm condition can be that there is a character or symbol related to "fault" in the fault decision determination result. In this embodiment, the re-judgment operation can be a repeated review operation based on artificial or artificial intelligence large model, and at this time, the artificial intelligence large model carefully checks the fault decision determination result to obtain the re-judgment result of whether the air-cooled server has a fault.

[0127] S207: When it is detected that the re-judgment result is a real fault, the re-judgment result is output as a target fault detection result.

[0128] In this embodiment, when it is detected that the re-judgment result is a real fault, it means that the air-cooled server has a fault under the current state. The re-judgment result is output as the target fault detection result for the staff of the air-cooled server fault detection to check in time to find the fault of the air-cooled server.

[0129] Based on the above embodiments, the air-cooled server fault detection method provided in an optional embodiment of the present application further comprises:

[0130] Step A: Constructing a multi-parameter associated fault type decision tree according to sample running data and decision diagnosis threshold.

[0131] In this embodiment, the multi-parameter associated fault type decision tree can be obtained by first calculating the change rate value of each sensor through the sample running data, then comparing the numerical value in combination with the decision diagnosis threshold, and then obtaining the phenomenon description corresponding to different comparison results and the corresponding fault type.

[0132] For example, the multi-parameter associated fault type decision tree provided by the embodiment of the present application is shown in Table 2 as follows:

[0133]

[0134] Step B: When it is detected that the re-judgment result is a real fault, determining the fault type according to the re-judgment result and the multi-parameter associated fault type decision tree.

[0135] In this embodiment, the re-judgment result is a real fault, and the accurate fault type can be determined according to the re-judgment result and the above multi-parameter associated fault type decision tree.

[0136] In an optional embodiment of the present application, the fault type decision is made for each sensor, so as to cover all components on each link of the air-cooled server, such as CPU, memory, hard disk, GPU, network card, and other components, and to more accurately determine the component with unstable working state, i.e., the component with potential hidden fault.

[0137] The air-cooled server fault detection method provided in an optional embodiment of the present application further includes:

[0138] Step C: When it is detected that the fault decision determination result does not satisfy the preset alarm condition, the current running data corresponding to the fault decision determination result is stored in the preset health data set.

[0139] In the embodiment, the preset health data set can be used as historical data for subsequent dynamic calculation of decision threshold, so as to obtain more accurate fault detection result.

[0140] Based on the above embodiments, the air-cooled server fault detection method provided in an optional embodiment of the present application further includes:

[0141] Step D: When it is detected that the re-determination result is not real fault, the current running data corresponding to the re-determination result is stored in the preset health data set.

[0142] In the embodiment, when the re-determination result determined after manual re-determination is that the current air-cooled server has no fault, it indicates that the current running data of the air-cooled server is health data, and thus is stored in the preset health data set.

[0143] The air-cooled server fault detection method provided in an optional embodiment of the present application further includes:

[0144] Step E: The current running data of the air-cooled server is acquired through the integrated circuit bus I2C at regular time.

[0145] In the embodiment, the current running data of the air-cooled server acquired through the integrated circuit bus I2C is more accurate than the data collected from the sensor through the intelligent platform management interface in the above embodiments, considering the delay and error of the sensor, and the running data of each hardware of the air-cooled server is directly acquired through the integrated circuit bus I2C, so that the sampling accuracy is higher, and thus the value of the sensor change rate and the decision diagnosis threshold obtained subsequently are more accurate, and the accuracy of the fault detection result is further improved.

[0146] Based on the above embodiments, the air-cooled server fault detection method provided in an optional embodiment of the present application further includes, before the current running data of the air-cooled server is acquired at regular time:

[0147] Step F: Obtain the historical running data of the air-cooled server, and determine the adjusted decision diagnosis threshold according to the historical running data and the preset strictness coefficient.

[0148] In this embodiment, the adjusted decision diagnosis threshold is dynamically adjusted and generated by using the historical running data to perform one-time overall calculation. Thus, as the historical running data increases, the decision diagnosis threshold gradually tends to be stable and accurate.

[0149] Step G: Determine the target decision diagnosis threshold according to the running data in the preset health data set and the preset strictness coefficient.

[0150] In this embodiment, the running data in the preset health data set, i.e., the running data after the fault data is excluded, can also be used each time the fault detection is performed, so as to determine a more accurate decision diagnosis threshold.

[0151] Specifically, the calculation formula for updating the target decision diagnosis threshold is as follows:

[0152] U = U_new + SD_new * T, L = L_new - SD_new * T U_new = the overall average value of the health running data calculated according to the running data in the preset health data set, and SD_new is the overall standard deviation of the health running data calculated according to the running data in the preset health data set. L is the target lower decision diagnosis threshold.

[0153] In the formula, U is the target upper decision diagnosis threshold,

[0154] In summary, the air-cooled server fault detection provided in the embodiments of the present application first controls the air-cooled server to run through the three-axis disturbance coordination engine, thereby improving the probability of triggering the fault. Meanwhile, the sample running data of the air-cooled server under various fault conditions is obtained at regular time intervals, and a large amount of sample running data provides the required data support for determining the subsequent decision diagnosis threshold. Then, the decision diagnosis threshold is determined based on the sample running data and the preset strictness coefficient, and is used for subsequent judgment of the fault of the air-cooled server. When the current running data obtained is subjected to fault judgment, the current sensor change rate and the decision diagnosis threshold are calculated to make a fault decision, and then artificial rejudgment is used for secondary determination, so as to ensure the accuracy of the fault detection result, reduce the screening and judgment work of the artificial in a large amount of running data, and improve the fault detection efficiency.

[0155] ​​Meanwhile, the air-cooled server fault detection method provided by the embodiment of the application can achieve the effect of traditional 72-hour stability test through 1-hour high-frequency disturbance, greatly improves the fault detection rate, and can find predictive hidden faults, thereby improving the detection efficiency. The method can be applied to assembly line screening, improves the product delivery quality, supports online detection, is more friendly to scenarios such as computer rooms, and can directly perform online detection when the server is idle. The method avoids the use of traditional temperature boxes or external devices, and can be used for batch server rapid health screening. The method performs fault analysis and decision-making from the perspective of devices, makes the fault granularity more fine, avoids frequent cross verification in the traditional maintenance mode, reduces the maintenance difficulty, and improves the maintenance timeliness.

[0156] Based on the above embodiment, in an optional embodiment of the application, the method for determining the decision-making diagnosis threshold in step S202 can also be to obtain a threshold generation model by using the operation data in the historical fault data set and combining artificial intelligence AI model for model training, and the threshold generation model is used in subsequent fault detection to improve the accuracy and reliability of the decision-making diagnosis threshold.

[0157] Through the description of the above embodiments, those skilled in the art can clearly understand that the method according to the above embodiments can be realized by means of software and a necessary general hardware platform, and of course, it can also be realized by hardware, but in many cases, the former is a better embodiment.

[0158] Figure 4 A structural schematic diagram of the air-cooled server fault detection device provided by the embodiment of the application is shown in FIG. 1. As shown in FIG. 1, the embodiment of the application further provides an air-cooled server fault detection device, which comprises: Figure 4

[0159] The data acquisition module 41 is configured to control the air-cooled server to run through three-axis disturbance coordination engines, and acquire sample operation data of the air-cooled server at regular time intervals, wherein the three-axis disturbance coordination engines are temperature disturbance-power supply disturbance-load disturbance to trigger a multi-coupling resonance and harmonic controller.

[0160] The calculation module 42 is configured to determine a decision-making diagnosis threshold according to the sample operation data and a preset strictness coefficient, wherein the preset strictness coefficient is a coefficient for presetting a strict or relaxed fault detection standard value.

[0161] The data acquisition module 41 is further configured to acquire current operation data of the air-cooled server at regular time intervals.

[0162] The calculation module 42 is further configured to determine a current sensor change rate according to the current operation data.

[0163] The fault determination module 43 is configured to perform fault decision-making determination processing according to the current sensor change rate and the decision-making diagnosis threshold, and obtain a fault decision-making determination result.​

[0164] The fault determination module 43 is further configured to, when it is detected that the fault decision determination result meets the preset alarm condition, perform re-determination on the current operation data to obtain a re-determination result in response to a re-determination operation;

[0165] The fault determination module 43 is further configured to, when it is detected that the re-determination result is a real fault, output the re-determination result as a target fault detection result.

[0166] In an optional embodiment of the present application, the data acquisition module 41 is specifically configured to: control the air-cooled server to change the heat convection power operation by the three-axis disturbance coordination engine according to the temperature disturbance type in a first period; control the air-cooled server to simulate power supply switching transient impact operation by the three-axis disturbance coordination engine according to the power supply disturbance type in a second period; and control the air-cooled server to stimulate power device stress changes by the three-axis disturbance coordination engine according to the load disturbance type in a third period; wherein the first period, the second period and the third period are cyclically alternated, and the number of each first period, each second period and each third period is within a preset fixed time length.

[0167] In an optional embodiment of the present application, the three-axis disturbance coordination engine realizes full coverage of the disturbance timing matrix within the preset fixed time length by a preset common multiple; the air-cooled server is deployed with a temperature sensor, a voltage sensor, a current sensor and a rotating speed sensor; and correspondingly, the fault determination module 43 is specifically configured to: acquire the temperature value of the air-cooled server from the temperature sensor through the intelligent platform management interface at a regular time; acquire the voltage value of the air-cooled server from the voltage sensor through the intelligent platform management interface at a regular time; acquire the current value of the air-cooled server from the current sensor through the intelligent platform management interface at a regular time; and acquire the rotating speed value of the air-cooled server from the rotating speed sensor through the intelligent platform management interface at a regular time.

[0168] In an optional embodiment of the present application, the sample operation data includes sensor data collected at a plurality of sampling time points; and correspondingly, the calculation module 42 is specifically configured to: perform change rate feature extraction processing on the sensor data collected at any two adjacent sampling time points to obtain sample sensor change rate data; determine a sample number according to the sample operation data; determine a sample population mean value according to the sample number and the sample sensor change rate data; determine a sample population standard deviation according to the sample population mean value, the sample number and the sample sensor change rate data; and determine a decision diagnosis threshold according to the sample population mean value, the sample population standard deviation and a preset severity coefficient.

[0169] In an optional embodiment of the present application, the calculation module 42 is configured to determine the calculation formula of the sample population mean value according to the sample number and the sample sensor change rate data as follows:

[0170] The calculation formula for determining the sample population average value from the sample sensor change rate data is:

[0171]

[0172] wherein, is the sample population average value, is the i-th sample sensor change rate value, n is the sample quantity, and i is a positive integer less than or equal to n;

[0173] The calculation formula for determining the sample population standard deviation from the sample population average value, the sample quantity, and the sample sensor change rate data is:

[0174]

[0175] wherein, SD is the sample population standard deviation, is the sample population average value, is the i-th sample sensor change rate value, n is the sample quantity, and i is a positive integer less than or equal to n;

[0176] The decision diagnosis threshold value includes a decision upper limit threshold value and a decision lower limit threshold value.

[0177] Correspondingly, the calculation formula for determining the decision diagnosis threshold value from the sample population average value, the sample population standard deviation, and a preset strictness coefficient is:

[0178]

[0179] wherein, L is the decision upper limit threshold value, U is the decision lower limit threshold value, and T is the preset strictness coefficient, is the sample population average value, and SD is the sample population standard deviation.

[0180] In an optional embodiment of the present application, the current operation data includes a voltage value, a current value, a temperature value, a power value, a rotating speed value, and a time point of multiple collected operation data; correspondingly, the calculation module 42 is specifically configured to: determine a current voltage sensor change rate according to the voltage value and the time point; determine a current current sensor change rate according to the current value and the time point; determine a current temperature sensor change rate according to the temperature value and the time point; determine a current rotating speed sensor change rate according to the rotating speed value and the time point; determine a current unit power consumption temperature rise rate according to the temperature value and the power value; and determine a current unit temperature fan rotating speed change rate according to the temperature value and the rotating speed value.

[0181] In an optional embodiment of the present application, the calculation formula for determining the current voltage sensor change rate by the calculation module 42 according to the voltage value and the time point is:

[0182] =

[0183] wherein, is the current voltage sensor rate of change, is the first of two adjacent time points, is the second of two adjacent time points, is the voltage value at the first of two adjacent time points, is the voltage value at the second of two adjacent time points.

[0184] The calculation module 42 is configured to determine, according to the current value and the time point, a calculation formula of the current current sensor rate of change as:

[0185]

[0186] wherein, is the current current sensor rate of change, is the current value at the first of two adjacent time points, is the current value at the second of two adjacent time points.

[0187] The calculation module 42 is configured to determine, according to the temperature value and the time point, a calculation formula of the current temperature sensor rate of change as:

[0188]

[0189] wherein, is the current temperature sensor rate of change, is the temperature value at the first of two adjacent time points, is the temperature value at the second of two adjacent time points.

[0190] The calculation module 42 is configured to determine, according to the rotation speed value and the time point, a calculation formula of the current rotation speed sensor rate of change as:

[0191]

[0192] wherein, is the current rotation speed sensor rate of change, is the rotation speed value at the first of two adjacent time points, is the rotation speed value at the second of two adjacent time points.

[0193] The calculation module 42 is configured to determine, according to the temperature value and the power value, a calculation formula of the current unit power consumption temperature rise rate as:

[0194]

[0195] wherein, is the current unit power consumption temperature rise rate, is a temperature value of a first time point of the two adjacent time points, is a temperature value of a second time point of the two adjacent time points, is a power consumption value of a first time point of the two adjacent time points, is a power consumption value of a second time point of the two adjacent time points.

[0196] The calculation module 42 is configured to determine a calculation formula of the current unit temperature fan speed change rate according to the temperature value and the rotation speed value, as follows:

[0197]

[0198] In the formula, is the current unit temperature fan speed change rate, is a temperature value of a first time point of the two adjacent time points, is a temperature value of a second time point of the two adjacent time points, is a rotation speed value of a first time point of the two adjacent time points, is a rotation speed value of a second time point of the two adjacent time points.

[0199] In an optional embodiment of the present application, the fault determination module 43 is further configured to: construct a multi-parameter associated fault type decision tree according to the sample operation data and a decision diagnosis threshold; and determine a fault type according to the re-determination result and the multi-parameter associated fault type decision tree when it is detected that the re-determination result is a real fault.

[0200] In an optional embodiment of the present application, the fault determination module 43 is further configured to: when it is detected that the fault decision determination result does not satisfy the preset alarm condition, store the current operation data corresponding to the fault decision determination result into a preset health data set.

[0201] In an optional embodiment of the present application, the fault determination module 43 is further configured to: when it is detected that the re-determination result is a non-real fault, store the current operation data corresponding to the re-determination result into the preset health data set.

[0202] In an optional embodiment of the present application, the data acquisition module 41 is further configured to: acquire the current operation data of the air-cooled server through an integrated circuit bus I2C at a regular time.

[0203] In the embodiment, the features of the air-cooled server fault detection device can be understood with reference to the related descriptions of the air-cooled server fault detection method, which will not be repeated here.

[0204] Figure 5 FIG. 1 is a structural schematic diagram of an electronic device according to an embodiment of the present application. Figure 5As shown, the electronic device 50 provided by the embodiment includes at least one processor 501 and a memory 502. Optionally, the electronic device 50 further includes a communication component 503. The processor 501, the memory 502 and the communication component 503 are connected through a bus.

[0205] In the implementation process, the at least one processor 501 executes the computer execution instructions stored in the memory 502, so that the at least one processor 501 executes the air-cooled server fault detection method embodiment described above.

[0206] The specific implementation process of the processor 501 can refer to the method embodiments described above, which has similar implementation principles and technical effects, and will not be described here in detail.

[0207] In the above embodiments, it should be understood that the processor can be a central processing unit (CPU), and can also be other general-purpose processors, digital signal processors (DSP), application specific integrated circuits (ASIC), etc. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor, etc. The steps of the method disclosed in the application can be directly embodied as execution completed by a hardware processor, or executed by a combination of hardware and software modules in the processor.

[0208] The memory can include a random access memory (RAM), and can also include a non-volatile memory (NVM), for example, at least one disk memory.

[0209] The bus can be an industry standard architecture (ISA) bus, a peripheral component (PCI) bus or an extended industry standard architecture (EISA) bus, etc. The bus can be divided into an address bus, a data bus, a control bus, etc. For the convenience of representation, the bus in the drawings of the present application does not limit to only one bus or one type of bus.

[0210] The embodiment of the present application also provides a computer readable storage medium, and the computer readable storage medium stores a computer program, wherein the computer program is set to execute the steps in any of the above XX method embodiments when running.

[0211] In an example embodiment, the computer readable storage medium described above can include, but is not limited to, a U disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a mobile hard disk, a magnetic disk or an optical disk, and various media that can store computer programs.

[0212] Embodiments of the present application also provide a computer program product, which includes a computer program, and the computer program, when executed by a processor, implements the steps in any of the air-cooled server fault detection method embodiments described above.

[0213] Embodiments of the present application also provide another computer program product, which includes a non-volatile computer readable storage medium, and the non-volatile computer readable storage medium stores a computer program, and the computer program, when executed by a processor, implements the steps in any of the air-cooled server fault detection method embodiments described above.

[0214] The skilled in the art can further realize that the units and algorithm steps of the examples described in conjunction with the embodiments disclosed herein can be realized in electronic hardware, computer software or a combination of both. In order to clearly illustrate the interchangeability of hardware and software, the components and steps of the examples have been described in the above description in a general manner. Whether the functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. The skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the present application.

[0215] The above describes in detail a kind of air-cooled server fault detection method provided by the present application. The principle and implementation mode of the present application are described in the specific examples in this paper, the above example is only for helping to understand the method of the present application and its core idea. It should be pointed out that, for the ordinary skilled in the art, under the premise of not departing from the principle of the present application, the present application can be improved and modified in several ways, and these improvements and modifications also fall within the protection scope of the claims of the present application.

Claims

1. A method for fault detection in an air-cooled server, characterized in that, include: Temperature-based disturbance via a three-axis disturbance cooperative engine Power supply disturbance The three-axis perturbation sequence of the load disturbance forms a perturbation full-coverage matrix within a fixed duration according to a preset time period to control the operation of the air-cooled server, and periodically acquires sample operation data of the air-cooled server. The three-axis perturbation collaborative engine is a temperature perturbation engine. Power supply disturbance A controller that triggers multi-coupled resonance and harmonic resonance by load disturbance, wherein the sample operation data includes sensor data collected at multiple sampling time points; The rate of change feature is extracted from sensor data collected at any two adjacent sampling time points to obtain sample sensor rate of change data; the number of samples is determined based on the sample operation data; the overall average value of the samples is determined based on the number of samples and the sample sensor rate of change data; the overall standard deviation of the samples is determined based on the overall average value, the number of samples, and the sample sensor rate of change data; and a decision diagnosis threshold is determined based on the overall average value, the overall standard deviation of the samples, and a preset tightening coefficient, wherein the preset tightening coefficient is a coefficient for pre-set tightening or relaxing fault detection standard values. The current operating data of the air-cooled server is periodically acquired; Based on the current operating data, determine the current sensor change rate; Based on the current sensor change rate and the decision diagnosis threshold, a fault decision determination process is performed to obtain the fault decision determination result. When the fault decision judgment result is detected to meet the preset alarm conditions, a re-judgment operation is performed to re-judge the current running data and obtain the re-judgment result. When the re-judgment result is detected as a real fault, the re-judgment result is output as the target fault detection result.

2. The method according to claim 1, characterized in that, The temperature-based perturbation is achieved through a three-axis perturbation cooperative engine. Power supply disturbance The three-axis perturbation sequence of load disturbances forms a perturbation full-coverage matrix within a fixed duration according to a preset time period to control the operation of the air-cooled server, including: The three-axis perturbation cooperative engine controls the air-cooled server to change its thermal convection power operation according to the temperature perturbation type in the first cycle. The three-axis disturbance cooperative engine controls the air-cooled server to simulate power supply switching transient impact operation according to the power supply disturbance type in the second cycle; The stress change of the power device of the air-cooled server is controlled by the three-axis disturbance cooperative engine according to the load disturbance type in the third cycle; wherein the first cycle, the second cycle and the third cycle alternate cyclically, and the values ​​of each first cycle, each second cycle and each third cycle are within a preset fixed duration.

3. The method according to claim 2, characterized in that, The method of controlling the air-cooled server to change its thermal convection power according to the temperature disturbance type in the first cycle via a three-axis perturbation cooperative engine includes: The three-axis disturbance cooperative engine controls the fan of the air-cooled server sequentially from automatic speed, 95% of automatic speed, automatic speed, 105% of automatic speed to full speed in the first cycle according to the temperature disturbance type, so that the air-cooled server changes the heat convection power to operate. The method of controlling the air-cooled server to simulate power supply switching transient impact operation in the second cycle according to the power supply disturbance type via a three-axis disturbance cooperative engine includes: The three-axis disturbance cooperative engine controls the two power supplies of the air-cooled server to switch from two power supplies, then to the first power supply, and then from two power supplies to the second power supply in the second cycle according to the power supply disturbance type, so that the air-cooled server can simulate the transient impact of power supply switching. The method of controlling the stress change of the power device of the air-cooled server according to the load disturbance type in the third cycle through the three-axis perturbation cooperative engine includes: In the third cycle, the load simulator of the air-cooled server is controlled by a three-axis perturbation cooperative engine according to the load perturbation type, and the load simulator simulates the power stress change of the air-cooled server from 30% load, 60% load to 90% load.

4. The method according to claim 2, characterized in that, The three-axis disturbance cooperative engine achieves full coverage of the disturbance timing matrix within the preset fixed duration by using a preset common multiple; the air-cooled server is equipped with temperature sensors, voltage sensors, current sensors and speed sensors; Accordingly, the periodic acquisition of the current operating data of the air-cooled server includes: The temperature value of the air-cooled server is periodically obtained from the temperature sensor through the intelligent platform management interface; The voltage value of the air-cooled server is periodically obtained from the voltage sensor via the intelligent platform management interface; The current value of the air-cooled server is periodically obtained from the current sensor through the intelligent platform management interface; The rotational speed of the air-cooled server is periodically obtained from the rotational speed sensor via the intelligent platform management interface.

5. The method according to claim 1, characterized in that, The formula for determining the overall average value of the sample based on the sample size and the sample sensor change rate data is as follows: In the formula, The mean of the sample population. Let be the sensor change rate value of the i-th sample, n be the number of samples, and i be a positive integer less than or equal to n; The formula for determining the population standard deviation based on the population mean, the number of samples, and the sample sensor change rate data is as follows: In the formula, SD is the standard deviation of the sample population. The mean of the sample population. Let be the sensor change rate value of the i-th sample, n be the number of samples, and i be a positive integer less than or equal to n; The decision diagnosis thresholds include an upper decision threshold and a lower decision threshold. Accordingly, the calculation formula for determining the decision diagnosis threshold based on the population mean, the population standard deviation, and the preset tightening coefficient is as follows: In the formula, L is the upper threshold of the decision, U is the lower threshold of the decision, and T is the preset tightening coefficient. denoted as the population mean and SD as the population standard deviation.

6. The method according to claim 5, characterized in that, The step of determining the decision diagnosis threshold based on the sample operation data and the preset tightening coefficient further includes: Obtain a fault dataset, which includes historical fault data obtained after multiple re-judgment operations; Select the historical fault data from the fault dataset, and calculate the historical fault parameter values ​​based on the historical fault data; If the historical fault parameter value is detected to be less than the decision diagnosis threshold, then the historical fault parameter value is used as the new decision diagnosis threshold.

7. The method according to claim 1, characterized in that, The current operating data includes voltage, current, temperature, power, speed, and the time points at which the operating data was collected multiple times; Accordingly, determining the current sensor change rate based on the current operating data includes: Determine the current rate of change of the voltage sensor based on the voltage value and the time point; Based on the current value and the time point, determine the current sensor's rate of change. Based on the temperature value and the time point, determine the current rate of change of the temperature sensor; Based on the rotational speed value and the time point, determine the current rate of change of the rotational speed sensor; Determine the current temperature rise rate per unit power consumption based on the temperature value and the power value; Based on the temperature and speed values, determine the rate of change of fan speed per unit temperature.

8. The method according to claim 7, characterized in that, The decision diagnostic thresholds include voltage diagnostic threshold, current diagnostic threshold, speed change diagnostic threshold, temperature change diagnostic threshold, temperature diagnostic threshold, unit power consumption diagnostic threshold, unit temperature rise slope diagnostic threshold, and unit temperature speed diagnostic threshold. The fault decision-making process based on the current sensor change rate and the decision diagnosis threshold, to obtain the fault decision-making result, includes: Based on the current voltage sensor change rate, the current current sensor change rate, the voltage diagnostic threshold, and the current diagnostic threshold, a capacitor fault determination process is performed to obtain a fault decision result on whether the capacitor value has decayed. Based on the current speed sensor change rate and the speed diagnostic threshold, a fan fault determination process is performed to obtain a fault decision result regarding whether the fan bearing is worn. Based on the current voltage sensor change rate, the current temperature sensor change rate, the voltage diagnostic threshold, and the temperature diagnostic threshold, anomaly determination processing of the semiconductor device voltage regulator is performed to obtain a fault decision determination result on whether the semiconductor device voltage regulator is abnormal. The radiator fault determination process is performed based on the temperature value, the current unit power consumption temperature rise rate, the temperature diagnostic threshold, and the unit temperature rise slope diagnostic threshold to obtain the fault decision result of whether the radiator has poor contact.

9. The method according to claim 1, characterized in that, Also includes: Based on the sample operation data and the decision diagnosis threshold, a multi-parameter associated fault type decision tree is constructed. When the re-judgment result is detected as a real fault, the fault type is determined based on the re-judgment result and the multi-parameter associated fault type decision tree.

10. The method according to claim 1, characterized in that, Also includes: When the fault decision judgment result is detected to not meet the preset alarm conditions, the current running data corresponding to the fault decision judgment result is stored in the preset health dataset.

11. The method according to any one of claims 1 to 10, characterized in that, Also includes: When the re-judgment result is detected as a non-real fault, the current running data corresponding to the re-judgment result is stored in a preset health dataset.

12. An electronic device, characterized in that, include: At least one processor and memory; The memory stores computer-executed instructions; The processor executes computer execution instructions stored in the memory, causing the processor to perform the method as described in any one of claims 1 to 11.

13. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-executable instructions, which, when executed by a processor, are used to implement the method as described in any one of claims 1 to 11.

14. A computer program product, characterized in that, Includes a computer program that, when executed by a processor, implements the method of any one of claims 1 to 11.

Images