Method and apparatus for determining stability of target device

Abstract

The application relates to the technical field of server data centers, and discloses a method and device for determining the runtime stability degree of a target device. When a trigger instruction is received, a plurality of preset indexes are obtained respectively corresponding to a plurality of data sets. According to a first data set corresponding to a first preset index, a weight factor corresponding to the first preset index is determined. According to weight factors corresponding to all preset indexes, a weight value corresponding to each preset index is determined. The index values included in each performance data in the first data set are subjected to a scaling operation to obtain a second data set. The second data set is input into a time-frequency conversion model to obtain a third data set. The index value corresponding to a target frequency in the third data set is determined as a characteristic value of the first preset index. According to the characteristic value and the weight value, the runtime stability degree of the target device is determined. The application can improve the efficiency of determining the runtime stability degree of the device, and the determined runtime stability degree is more accurate.

Description

Technical Field

[0001] This invention relates to the field of server data center technology, and more specifically to a method and apparatus for determining the operational stability of a target device. Background Technology

[0002] In the field of server and data center technology, data centers contain a large number of server devices. The stability of server operation affects data security and business processing in the data center. Therefore, technicians need to monitor the stability of server operation through server performance data in order to promptly inspect and repair servers with low stability.

[0003] Generally, servers can statistically analyze the acquired performance data to obtain the variance, mean deviation, and other parameters for each performance metric. Furthermore, technical personnel can analyze the server's operational stability based on the variance and mean deviation of each metric. Alternatively, servers can display the acquired performance data in charts, allowing technical personnel to analyze the server's operational stability using these charts.

[0004] In related technologies, the analysis of server stability by technicians is time-consuming, inefficient, and the results are highly susceptible to the subjectivity of the technicians, which may lead to lower accuracy. Summary of the Invention

[0005] In view of this, the present invention provides a method, apparatus, computer equipment and storage medium for determining the operational stability of a target device, in order to solve the problems of low accuracy and low efficiency of analysis results.

[0006] In a first aspect, the present invention provides a method for determining the operational stability of a target device, the method comprising:

[0007] When a trigger command is received, a data set corresponding to each of the multiple preset indicators is obtained, wherein the data set includes performance data in the target device corresponding to the preset indicator;

[0008] Based on the first data set corresponding to the first preset indicator, a weight factor corresponding to the first preset indicator is determined, wherein the first preset indicator is any one of the plurality of preset indicators;

[0009] Based on the weight factors corresponding to all the preset indicators, determine the weight value corresponding to each preset indicator;

[0010] The performance data in the first dataset is scaled to obtain the second dataset.

[0011] The second data set is input into the time-frequency conversion model to obtain the third data set;

[0012] The index value corresponding to the target frequency in the third data set is determined as the feature value corresponding to the first preset index;

[0013] The stability of the target device during operation is determined based on the feature value and the weight value.

[0014] The method for determining the operational stability of a target device provided by this invention has the following advantages:

[0015] Since the operational stability of a target device is closely related to its performance data at each point in time, its stability can be determined using this data. Because different preset indicators have varying impacts on the device's stability, the influence of each preset indicator on its stability can be determined by statistically analyzing the performance data of all preset indicators. Time-frequency conversion of the performance data ensures that each converted data point reflects the drastic changes in the preset indicators. In summary, determining the operational stability of a target device by assigning weights and corresponding characteristic values ​​to each preset indicator considers both the impact of each indicator on the device's operation and the drastic changes in each indicator, resulting in a more accurate determination of operational stability. Furthermore, technicians only need to take appropriate measures based on the final given operational stability level, without needing to meticulously analyze the statistical results of the performance data to determine whether the target device is operationally stable. This significantly improves the efficiency of determining the operational stability of target devices and the efficiency of repairing malfunctioning equipment.

[0016] In one optional implementation, determining the weight value corresponding to each preset indicator based on the weight factors corresponding to all the preset indicators includes:

[0017] The total weight factor is constructed based on the weight factors corresponding to all the preset indicators;

[0018] Based on the weight factor corresponding to each of the preset indicators and the total weight factor, the weight value corresponding to each of the preset indicators is determined.

[0019] Specifically, since the fluctuation levels of different preset indicators have varying impacts on the operational stability of the target equipment, the degree of influence of each preset indicator on the operational stability of the target equipment can be determined based on the weighting factors of all preset indicators. This allows for a more accurate determination of the operational stability of the target equipment during subsequent processing.

[0020] In an optional implementation, before scaling the metric values ​​in each performance data set in the first data set to obtain the second data set, the method further includes:

[0021] Traverse the first data set to determine whether any performance data corresponding to the target time point is missing from the first data set;

[0022] If the performance data corresponding to the target time point is missing in the first data set, the performance data corresponding to the target time point is obtained.

[0023] Based on the performance data corresponding to the target time point, the first data set is updated to obtain a fourth data set, so that the indicator values ​​in each performance data in the fourth data set can be scaled to obtain the second data set.

[0024] Specifically, due to potential network instability on the target device or task server, some performance data in the dataset may be lost. Using a dataset with this missing performance data for subsequent processing would result in lower accuracy in determining the target device's operational stability. Therefore, when performance data loss is detected, it's necessary to retrieve the missing data and replenish the dataset. This ensures the integrity of the performance data in the dataset, further enhancing the accuracy of the determination of the target device's operational stability based on complete performance data.

[0025] In one optional implementation, the step of acquiring the performance data corresponding to the target time point when the first data set is missing performance data corresponding to the target time point includes:

[0026] Determine the previous and next time points corresponding to the target time point;

[0027] Extract the first performance data corresponding to the previous time point and the second performance data corresponding to the next time point from the first data set;

[0028] Based on the number of target time points between the previous time point and the next time point, determine the time interpolation corresponding to the target time point;

[0029] Based on the index values ​​included in the first performance data, the index values ​​included in the second performance data, the time interpolation corresponding to the target time point, and the pre-acquired interpolation model, determine the index values ​​included in the performance data corresponding to the target time point.

[0030] The performance data corresponding to the target time point is constructed by using the index values ​​included in the performance data corresponding to the target time point and the target time point;

[0031] or,

[0032] Extract the performance data corresponding to the target time point from the performance data stored in the target device.

[0033] Specifically, in some scenarios, the target device stores its own performance data long-term. Therefore, when missing performance data is identified, the original performance data can be retrieved from the target device. In other scenarios, the target device does not store its own performance data, or, for some reason, some performance data is not recorded. Therefore, to identify this type of missing performance data, this solution employs an interpolation calculation method. Since the performance data at the target time point is relatively similar to the performance data at nearby time points, the performance data at the target time point can be determined using the performance data at the nearest previous and next time points. By setting different methods to complete missing performance data, this solution ensures the integrity of the performance data in the dataset, further leading to a more accurate determination of the target device's operational stability.

[0034] In one optional implementation, scaling the metric values ​​included in each performance data point in the first data set to obtain the second data set includes:

[0035] Get the scaling range;

[0036] In the first data set, extract the maximum and minimum indicator values ​​corresponding to the first preset indicator;

[0037] Determine the absolute value of the difference between the maximum index value and the minimum index value;

[0038] The scaling factor is determined based on the absolute value of the difference and the length of the scaling interval.

[0039] Based on the scaling factor, any boundary interval value of the scaling interval, and the minimum index value, the performance data included in the first data set is scaled to obtain the second data set.

[0040] Specifically, when performing time-frequency conversion on time-series data, the performance data needs to meet certain conditions, namely, it must be within a specified symmetrical interval. Therefore, by scaling the performance data indicators in the dataset, the scaled indicators can satisfy the time-frequency conversion conditions for subsequent time-frequency conversion processing.

[0041] In one optional implementation, the step of inputting the second data set into the time-frequency conversion model to obtain the third data set includes:

[0042] Count the number of performance data in the second dataset;

[0043] Determine whether each preset frequency value among multiple preset frequency values ​​is equal to zero;

[0044] When the preset frequency value is equal to zero, a compensation coefficient corresponding to the preset frequency value is determined based on the first preset coefficient and the number of performance data in the second data set.

[0045] or,

[0046] When the preset frequency value is not equal to zero, a compensation coefficient corresponding to the preset frequency value is determined based on the second preset coefficient and the number of performance data in the second data set.

[0047] Based on the index values ​​included in the i-th performance data in the second data set, the number of performance data in the second data set, and the value corresponding to i, determine the cosine transformation value corresponding to the i-th performance data, where i is a positive integer;

[0048] The target cosine transform value is obtained by summing the cosine transform values ​​corresponding to each performance data point.

[0049] Based on the target cosine transform value and the compensation coefficient corresponding to the preset frequency value, determine the index value corresponding to the preset frequency value;

[0050] The multiple preset frequency values ​​and the index values ​​corresponding to each preset frequency value constitute the third data set.

[0051] Specifically, during the time-frequency conversion of indicator values, the indicator value corresponding to each frequency value is determined based on the indicator value at each time point. This allows us to determine the value of the preset indicator at each frequency, and thus the degree of fluctuation of each preset indicator. Furthermore, based on the degree of fluctuation of each preset indicator, the operational stability of the target equipment can be directly determined without the need for complex analysis by technical personnel, making it more convenient and improving the efficiency of determining the stability of the target equipment.

[0052] In one alternative implementation, the target frequency is the highest frequency value in the third data set, or any frequency value in the third data set.

[0053] Specifically, during the time-frequency conversion of indicator values, the indicator value corresponding to each frequency value is determined based on the indicator values ​​at all time points. Therefore, the indicator value corresponding to each frequency value can represent the fluctuation level of the preset indicator. Since the indicator value corresponding to the highest frequency value can represent the fluctuation level of the preset indicator over a short time span, it can be used as a feature value of the preset indicator to more accurately represent the real-time operational stability of the target equipment. Furthermore, the operational stability of the target equipment determined through this feature value corresponding to each preset indicator is also more accurate.

[0054] In one optional implementation, the second data set is input into the time-frequency conversion model to obtain a third data set.

[0055] It can be represented by the following expression:

[0056]

[0057]

[0058] Wherein, u is any one of the preset frequency values, F(u) is the index value corresponding to the preset frequency value, c(u) is the compensation coefficient corresponding to the preset frequency value, i is the sequence number of any performance data in the second data set, N is the number of performance data in the second data set, and f(i) is the index value included in the i-th performance data in the second data set; the multiple preset frequency values ​​and the index value corresponding to each preset frequency value constitute the third data set.

[0059] Specifically, during the time-frequency conversion of indicator values, the indicator value corresponding to each frequency value is determined based on the indicator value at each time point. This allows us to determine the value of the preset indicator at each frequency, and thus the degree of fluctuation of each preset indicator. Furthermore, based on the degree of fluctuation of each preset indicator, the operational stability of the target equipment can be directly determined without the need for complex analysis by technical personnel, making it more convenient and improving the efficiency of determining the stability of the target equipment.

[0060] Secondly, the present invention provides a device for determining the operational stability of a target device, including an acquisition module, which, when a trigger command is received, acquires a data set corresponding to each of a plurality of preset indicators, wherein the data set includes performance data of the target device corresponding to the preset indicators;

[0061] The determining module is used to determine a weight factor corresponding to the first preset indicator based on a first data set corresponding to the first preset indicator, wherein the first preset indicator is any one of a plurality of preset indicators; and to determine a weight value corresponding to each preset indicator based on the weight factors corresponding to all the preset indicators respectively.

[0062] The scaling module is used to scale the indicator values ​​included in each performance data in the first data set to obtain the second data set.

[0063] The conversion module is used to input the second data set into the time-frequency conversion model to obtain the third data set;

[0064] The determining module is used to determine the index value corresponding to the target frequency in the third data set as the feature value corresponding to the first preset index; and to determine the stability of the target device during operation based on the feature value and the weight value.

[0065] Thirdly, the present invention provides a computer device, comprising: a memory and a processor, the memory and the processor being communicatively connected to each other, the memory storing computer instructions, and the processor executing the computer instructions to perform the method for determining the operational stability of the target device described in the first aspect or any corresponding embodiment.

[0066] Fourthly, the present invention provides a computer-readable storage medium storing computer instructions for causing a computer to execute the method for determining the operational stability of a target device as described in the first aspect or any corresponding embodiment. Attached Figure Description

[0067] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0068] Figure 1 This is a flowchart illustrating a method for determining the operational stability of a target device according to an embodiment of the present invention.

[0069] Figure 2 This is a flowchart illustrating a method for determining the operational stability of another target device according to an embodiment of the present invention;

[0070] Figure 3 This is a flowchart illustrating another method for determining the operational stability of a target device according to an embodiment of the present invention;

[0071] Figure 4 This is a structural block diagram of a device for determining the operational stability of a target device according to an embodiment of the present invention;

[0072] Figure 5 This is a schematic diagram of the hardware structure of a computer device according to an embodiment of the present invention. Detailed Implementation

[0073] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0074] Data centers today typically house a large number of servers, and managing such a vast number of servers through operations and maintenance (O&M) software is a crucial issue the industry needs to address. Current O&M software collects real-time performance, alarms, and logs from server devices via in-band or out-of-band networks, performs simple statistical analysis, and then displays the data to users, lacking intermediate analysis and processing. Specific processing techniques involve statistical analysis of data from both temporal and physical dimensions: First, in the temporal dimension, performance data such as power consumption, temperature, CPU utilization, memory utilization, and input / output (IO) utilization are simply presented as trend graphs over time. Second, in the physical dimension, multiple server devices are simply grouped according to their physical location (rack, server room, data center) for data aggregation and simple calculations.

[0075] Depending on the perspective, the operational status of servers can be simply divided into the overall operational status of the server cluster and the operational status of individual servers. A comprehensive description of the various metrics of the server cluster helps us grasp the overall operational status of the cluster from a macro perspective, thereby guiding physical or business adjustments to the entire server cluster. On the other hand, summarizing and analyzing the various metrics of individual servers allows for the timely detection of anomalies during server operation, thus preventing adverse effects from a single server on the overall cluster.

[0076] Server performance metrics are time-series data, meaning these values ​​change over time. For a single server, it's generally desirable to minimize fluctuations and maintain stability in performance metrics such as power consumption, temperature, CPU utilization, memory utilization, and I / O utilization. Firstly, server stability reflects the health of the server hardware; technicians can monitor server stability to identify abnormal servers and perform timely troubleshooting and repairs. Secondly, server stability also reflects the proper functioning of the services running on the server; similarly, technicians can monitor server stability to adjust services promptly and detect and repair any abnormalities.

[0077] According to an embodiment of the present invention, a method for determining the operational stability of a target device is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0078] This embodiment provides a method for determining the operational stability of a target device, applicable to computer devices such as servers and computers. Specifically, this embodiment can be used to determine the operational stability of multiple servers in a server cluster. The server cluster may include one or more servers specifically dedicated to executing this method (hereinafter referred to as task servers), or each server in the server cluster may determine its own operational stability. This embodiment uses the first scenario as an example for illustration. This embodiment uses a server as the target device for illustration. Figure 1 This is a flowchart of a method for determining the operational stability of a target device according to an embodiment of the present invention, such as... Figure 1 As shown, the process includes the following steps:

[0079] Step S101: When a trigger command is received, obtain the data set corresponding to each of the multiple preset indicators.

[0080] The triggering instruction can be an instruction triggered according to a preset time period, or an instruction triggered after the processing in step S107 is completed. Multiple preset indicators can include the device's power consumption, temperature, CPU utilization, memory utilization, and I / O utilization. The dataset includes performance data from the target device corresponding to the preset indicators. The target device is any device in the server cluster. The performance data includes multiple time points and the corresponding indicator values ​​for each time point.

[0081] Specifically, because different devices in a server cluster handle different types of business, their peak runtimes also vary significantly. During non-peak runtimes, the values ​​of each preset metric change very little, and performance data obtained during non-peak runtimes is not very meaningful for determining the stability of device operation. Therefore, the two trigger commands mentioned above can be set during the device's peak runtime. This can save processing resources on the task server.

[0082] Different trigger commands determine the frequency of data acquisition. When it is necessary to determine the operational stability of a device over a longer time span, the task server can use a command triggered at a preset time period, setting a longer time period according to actual needs. When it is necessary to determine the operational stability of a device over a shorter time span, the task server can use a command triggered after the completion of step S107. These two types of trigger commands allow technicians to select the appropriate type based on actual needs, facilitating monitoring of the task server's operational stability from different time dimensions. Furthermore, selecting appropriate trigger commands based on the characteristics of different task servers ensures that the acquired data set better matches the server's operational characteristics, further leading to more accurate determination of operational stability in subsequent processing.

[0083] Step S102: Determine the weight factor corresponding to the first preset indicator based on the first data set corresponding to the first preset indicator.

[0084] The first preset indicator is any one of multiple preset indicators.

[0085] Specifically, the task server can obtain the data set corresponding to each preset indicator and determine the weight factor corresponding to each preset indicator.

[0086] The following explanation uses a preset indicator as the first preset indicator as an example to illustrate the process of determining the weight factor. The task server can determine the average difference of the first preset indicator based on all indicator values ​​in the first dataset, and then use this average difference as the weight factor for the first preset indicator. Alternatively, the task server can determine the variance of the first preset indicator based on all indicator values ​​in the first dataset, and then use this variance as the weight factor for the first preset indicator. Or, the task server can determine the range of the first preset indicator based on the maximum and minimum indicator values ​​in the first dataset, and then use this range as the weight factor for the first preset indicator.

[0087] Since variance is calculated as the sum of the squares of the differences between each indicator value and the average, it reflects the overall drastic change in the preset indicator. Since the mean difference is calculated as the average of the absolute values ​​of the differences between each indicator value and the average, it reflects the average drastic change in all indicator values ​​in the dataset. Since the range is the difference between the maximum and minimum indicator values, it reflects the span of change in the preset indicator, i.e., the drastic change in the preset indicator. In summary, the weighting factors determined by any of the above methods can represent the drastic change in the preset indicator. Furthermore, using the drastic change as a weighting factor can accurately reflect the impact of the preset indicator on the stability of equipment operation.

[0088] The same method can be used to obtain the weight factors corresponding to all preset indicators.

[0089] Step S103: Determine the weight value corresponding to each preset indicator based on the weight factors corresponding to all preset indicators.

[0090] Specifically, the task server can sum the weight factors corresponding to all preset indicators, and determine the weight value corresponding to each preset indicator based on the summation result and the weight factor corresponding to each preset indicator.

[0091] Step S104: Scale the index values ​​included in each performance data in the first data set to obtain the second data set.

[0092] Specifically, the task server can input the metric values ​​included in each performance data point in the first dataset into a pre-acquired scaling model to obtain a scaled metric value corresponding to each metric value. The scaling model can be a machine learning model. Alternatively, the task server can multiply each metric value included in the first dataset by a preset scaling factor to obtain a scaled metric value for each metric value. The time point corresponding to each metric value and the scaled metric value corresponding to each metric value constitute the second dataset.

[0093] The same method can be used to obtain the data sets corresponding to all preset indicators after scaling operations.

[0094] Step S105: Input the second data set into the time-frequency conversion model to obtain the third data set.

[0095] The time-frequency conversion model can be a machine learning model or a pre-defined mathematical model.

[0096] Specifically, the task server can input the index values ​​included in each performance data in the second data set into the time-frequency conversion model to obtain multiple frequency values ​​and the index values ​​corresponding to each frequency value. Each frequency value and the index value corresponding to that frequency value constitute each performance data, and all the performance data constitute the third data set.

[0097] The same method can be used to obtain the time-frequency converted data sets corresponding to all preset indicators.

[0098] Step S106: Determine the index value corresponding to the target frequency in the third data set as the feature value corresponding to the first preset index.

[0099] Specifically, since the index value corresponding to the target frequency in the third data set is determined based on the index values ​​corresponding to all time points in the second data set, it can represent the characteristics of the first preset index. Therefore, the index value corresponding to the target frequency can be determined as the characteristic value corresponding to the first preset index.

[0100] The same method can be used to obtain the feature values ​​corresponding to all preset indicators.

[0101] Step S107: Determine the stability of the target device during operation based on the feature value and weight value.

[0102] Specifically, the task server can perform a weighted summation based on the feature values ​​and weight values ​​corresponding to each preset indicator to obtain the target fluctuation value of the target device. The task server can store the correspondence between fluctuation value ranges and operational stability levels. In this way, the task service can determine the operational stability level of the target device by determining the target fluctuation value and this correspondence.

[0103] The same method can be used to determine the operational stability of all devices in a server cluster.

[0104] The method for determining the operational stability of a target device provided in this embodiment is based on the fact that the operational stability of the target device is closely related to its performance data at each time point. Therefore, the stability of the target device can be determined through its performance data. Since different preset indicators have different impacts on the operational stability of the target device, the degree of influence of each preset indicator on the operational stability can be determined based on the statistical results of the performance data of all preset indicators. By performing time-frequency conversion processing on the performance data, each converted performance data point can reflect the drastic change of the preset indicator. In summary, by determining the operational stability of the target device through the weight value of each preset indicator and the corresponding characteristic value, both the degree of influence of each preset indicator and the drastic change of each indicator can be considered, making the final determined operational stability more accurate. Furthermore, technicians only need to take corresponding measures based on the final given operational stability, without needing to carefully analyze the statistical results of the performance data to determine whether the target device is operationally stable before taking measures. This greatly improves the efficiency of determining the operational stability of the target device and the efficiency of repairing problematic equipment.

[0105] This embodiment provides a method for determining the operational stability of a target device, applicable to computer devices such as servers and computers. Specifically, this embodiment can be used to determine the operational stability of multiple servers in a server cluster. The server cluster may include one or more servers specifically dedicated to executing this method (hereinafter referred to as task servers), or each server in the server cluster may determine its own operational stability. This embodiment uses the first scenario as an example for illustration. This embodiment uses a server as the target device for illustration. Figure 2 This is a flowchart of a method for determining the operational stability of a target device according to an embodiment of the present invention, such as... Figure 2 As shown, the process includes the following steps:

[0106] Step S201: When a trigger command is received, obtain the data set corresponding to each of the multiple preset indicators.

[0107] Step S202: Determine the weight factor corresponding to the first preset indicator based on the first data set corresponding to the first preset indicator.

[0108] For details on steps S201 to S202, please refer to [link / reference]. Figure 1 Steps S101 to S102 of the illustrated embodiment will not be described again here.

[0109] Step S203: Determine the weight value corresponding to each preset indicator based on the weight factors corresponding to all preset indicators.

[0110] Step S203 above includes:

[0111] Step S2031: Construct the total weight factor based on the weight factors corresponding to all preset indicators.

[0112] Specifically, the task server can sum the weight factors corresponding to all preset indicators to obtain the target total weight factor.

[0113] Step S2032: Based on the weight factor corresponding to each preset indicator and the total weight factor, determine the weight value corresponding to each preset indicator.

[0114] Specifically, for each preset indicator, the weight factor corresponding to that preset indicator is used as the numerator, and the total weight factor is used as the denominator to obtain the weight value corresponding to each preset indicator.

[0115] For example, the weight value corresponding to each preset indicator can be expressed as follows:

[0116]

[0117] Among them, P i F represents the weight value of the i-th preset indicator. i This represents the weight factor corresponding to the i-th preset indicator, and n represents the number of preset indicators.

[0118] Step S204: Traverse the first data set to determine whether any performance data corresponding to the target time point is missing from the first data set.

[0119] Specifically, when the task server acquires a data set, if the data set is obtained from the target device, the task server needs to send an acquisition command to the target device before acquiring the data set. After receiving the acquisition command, the task server sends the corresponding data set to the target device. During this process, data loss at some points in time may occur due to network instability of the task server or the target device. If data loss occurs, the performance data in the data set will be incomplete, further leading to inaccurate operational stability obtained through subsequent processing. Therefore, the task server can iterate through the data sets corresponding to all preset indicators to determine whether performance data corresponding to the target time point is missing in the data set corresponding to each preset indicator. The target time point may be one or multiple. If so, step S205 can be performed to complete the performance data corresponding to the target time point; otherwise, step S206 can be performed. This step ensures that the performance data in the data sets corresponding to all preset indicators is relatively complete, further making the operational stability determined through subsequent processing more accurate.

[0120] Step S205: When the performance data corresponding to the target time point is missing in the first data set, obtain the performance data corresponding to the target time point.

[0121] Specifically, if the target device stores performance data for each time point, then the performance data corresponding to the target time point can be obtained according to step S2056. If the target device does not store performance data for each time point, or if the target device fails to record data for the target time point for other reasons, then the performance data corresponding to the target time point can be obtained according to steps S2051 to S2055.

[0122] Step S205 above includes:

[0123] Step S2051: Determine the previous time point and the next time point corresponding to the target time point.

[0124] Step S2052: Extract the first performance data corresponding to the previous time point and the second performance data corresponding to the next time point from the first data set.

[0125] Specifically, since the original performance data of the target device at the target time point is lost, the performance data of the two time points closest to the target time point can represent the performance data of the target time point to a certain extent. Therefore, the previous time point and the next time point corresponding to the target time point can be determined, and the performance data corresponding to these two time points can be extracted from the first data set.

[0126] In some cases, performance data for multiple consecutive target time points may be missed. Therefore, for each target time point, the task server can determine the nearest previous and next time points that have no missing performance data. Furthermore, by analyzing the performance data at these two time points, the performance data for all target time points in between can be determined.

[0127] Step S2053: Determine the time interpolation corresponding to the target time point based on the number of target time points between the previous time point and the next time point.

[0128] Specifically, first, the task server determines the number of target time points between the previous and next time points and sets the time span between them to one. Then, it increments the number of target time points by one to obtain the number of intervals between the previous and next time points. Finally, it uses the time span of one as the numerator and the number of intervals as the denominator to obtain the interval length. Based on the order of each target time point between the previous and next time points and the interval length, it determines the time interpolation corresponding to each target time point. Specifically, the product of the order of the target time points and the interval length is used as the time interpolation value for that target time point. For example, when the number of target time points between the previous and next time points is 1, the interval length is 1 / 2, and 1 / 2 is used as the time interpolation value for that target time point. When the number of target time points between the previous and next time points is 2, the interval length is 1 / 3, and 1 / 3 is used as the time interpolation value for the first target time point, and 2 / 3 is used as the time interpolation value for the second target time point.

[0129] Step S2054: Based on the index values ​​included in the first performance data, the index values ​​included in the second performance data, the time interpolation corresponding to the target time point, and the pre-acquired interpolation model, determine the index values ​​included in the performance data corresponding to the target time point.

[0130] Specifically, the pre-obtained interpolation model can be expressed as follows:

[0131] lerp(a,b,u)=a+fade(t)*(ba)……(2)

[0132] fade(t) = 6t 5 -15t 4 +10t 3 ……(3)

[0133] Where lerp(a,b,u) represents the index values ​​included in the performance data corresponding to the target time point, a is the index value included in the first performance data, b is the index value included in the second performance data, fade(t) is the nonlinear interpolation function fade replacement, and t is the time interpolation corresponding to the target time point.

[0134] Step S2055: Using the performance data corresponding to the target time point, including the index values ​​and the target time point, construct the performance data corresponding to the target time point.

[0135] or,

[0136] Step S2056: Extract the performance data corresponding to the target time point from the performance data stored in the target device.

[0137] Specifically, the task server can send a missing data retrieval command to the target device, whereby the data retrieval command includes a target time point. Upon receiving the missing data retrieval command, the target device extracts the performance data corresponding to the target time point from the stored performance data based on the target time point in the missing data retrieval command.

[0138] Step S206: Update the first data set based on the performance data corresponding to the target time point to obtain the fourth data set, so that the index values ​​in each performance data in the fourth data set can be scaled to obtain the second data set.

[0139] Specifically, the task server fills the corresponding positions in the first data set with the performance data corresponding to the target time point, thus obtaining the fourth data set. In this way, a fourth data set with complete performance data can be obtained.

[0140] Step S207: Scale the index values ​​included in each performance data in the fourth data set to obtain the second data set.

[0141] Step S208: Input the second data set into the time-frequency conversion model to obtain the third data set.

[0142] Step S209: Determine the index value corresponding to the target frequency in the third data set as the feature value corresponding to the first preset index.

[0143] Step S210: Determine the stability of the target device during operation based on the feature value and weight value.

[0144] For details on steps S207 to S210, please refer to [link / reference]. Figure 1 Steps S104 to S107 of the illustrated embodiment will not be described again here.

[0145] The method for determining the operational stability of a target device provided in this embodiment addresses the issue that network instability may occur on the target device or task server, leading to the loss of some performance data in the dataset. Directly using the dataset with missing performance data for subsequent processing would result in a lower accuracy in determining the operational stability of the target device. Therefore, this solution ensures the integrity of the performance data in the dataset by setting different methods to complete the missing performance data, thereby making the final determination of the operational stability of the target device more accurate.

[0146] This embodiment provides a method for determining the operational stability of a target device, applicable to computer devices such as servers and computers. Specifically, this embodiment can be used to determine the operational stability of multiple servers in a server cluster. The server cluster may include one or more servers specifically dedicated to executing this method (hereinafter referred to as task servers), or each server in the server cluster may determine its own operational stability. This embodiment uses the first scenario as an example for illustration. This embodiment uses a server as the target device for illustration. Figure 3 This is a flowchart of a method for determining the operational stability of a target device according to an embodiment of the present invention, such as... Figure 3 As shown, the process includes the following steps:

[0147] Step S301: When a trigger command is received, obtain the data set corresponding to each of the multiple preset indicators.

[0148] Step S302: Determine the weight factor corresponding to the first preset indicator based on the first data set corresponding to the first preset indicator.

[0149] Step S303: Determine the weight value corresponding to each preset indicator based on the weight factors corresponding to all preset indicators.

[0150] For details on steps S301 to S303, please refer to [link / reference]. Figure 1 Steps S101 to S103 of the illustrated embodiment will not be described again here.

[0151] Step S304: Scale the index values ​​included in each performance data in the first data set to obtain the second data set.

[0152] Step S304 above includes:

[0153] S3041, Get the scaling range.

[0154] Specifically, the scaling interval is a symmetrical interval. The scaling intervals corresponding to different preset indicators can be different or the same. When the scaling intervals corresponding to different preset indicators are different, the task server can obtain the scaling interval corresponding to each preset indicator. This step is to obtain the scaling interval corresponding to the first preset indicator. For example, the scaling interval of the first preset indicator can be [-128, 128].

[0155] S3042, Extract the maximum and minimum index values ​​corresponding to the first preset index from the first data set.

[0156] S3043, determine the absolute value of the difference between the maximum and minimum index values.

[0157] For example, the absolute value of the difference D = MAX - MIN, where MAX is the maximum index value and MIN is the minimum index value.

[0158] S3044, determine the scaling factor based on the absolute value of the difference and the length of the scaling interval.

[0159] Specifically, the task server can use the interval length as the numerator and the absolute value of the difference as the denominator to obtain the scaling factor. For example, when the interval length is 256 and the absolute value of the difference is D, the scaling factor I = 256 / D.

[0160] S3045, based on the scaling factor, any boundary interval value of the scaling interval, and the minimum index value, scale the index values ​​included in the performance data in the first data set, and then obtain the second data set.

[0161] Specifically, the task server can input the scaling factor, any boundary interval value of the scaling interval, the minimum index value, and the index values ​​included in each performance data in the first data set into the scaling model to obtain the scaled index value corresponding to the index value included in each performance data. Each scaled index value and its corresponding time point are used as the performance data constituting the second data set.

[0162] The scaling model can be expressed as follows:

[0163] X=(x-MIN)*I-128……(4)

[0164] Where X is the scaled index value, x is the index value before scaling, I is the scaling factor, and 128 is the right boundary interval value of the scaling interval [-128, 128].

[0165] Step S305: Input the second data set into the time-frequency conversion model to obtain the third data set.

[0166] Step S305 above includes:

[0167] Step S3051: Count the number of performance data in the second data set.

[0168] Step S3052: Determine whether each preset frequency value among the multiple preset frequency values ​​is equal to zero.

[0169] Step S3053: When the preset frequency value is equal to zero, determine the compensation coefficient corresponding to the preset frequency value based on the first preset coefficient and the number of performance data in the second data set.

[0170] or,

[0171] Step S3054: When the preset frequency value is not equal to zero, determine the compensation coefficient corresponding to the preset frequency value based on the second preset coefficient and the number of performance data in the second data set.

[0172] Step S3055: Determine the cosine transform value corresponding to the i-th performance data based on the index values ​​included in the i-th performance data in the second data set, the number of performance data in the second data set, and the value corresponding to i.

[0173] Where i is a positive integer.

[0174] Step S3056: Sum the cosine transformation values ​​corresponding to each performance data to obtain the target cosine transformation value.

[0175] Step S3057: Determine the index value corresponding to the preset frequency value based on the target cosine transformation value and the compensation coefficient corresponding to the preset frequency value.

[0176] Specifically, steps S3051 to S3057 can be expressed as follows:

[0177]

[0178]

[0179] Where F(u) is the index value corresponding to the preset frequency value, c(u) is the compensation coefficient, i is the sequence number of any performance data in the second data set, N is the number of performance data in the second data set, u is the preset frequency value, and f(i) is the index value included in the i-th performance data in the second data set.

[0180] In step S3051, N in the above expression is determined.

[0181] In step S3052, multiple preset frequency values ​​are obtained, for example, u1=0, u2=50, u3=100, u4=150, ...

[0182] In step S3053, when u = 0, determine

[0183] In step S3054, when u≠0, determine

[0184] In step S3055, the cosine transform value corresponding to each performance data is determined. That is, determine the cosine transform value corresponding to the first performance data. The cosine transform value corresponding to the second performance data is The cosine transform value corresponding to the third performance data is And so on.

[0185] In step S3056, the target cosine transform value is determined.

[0186] In step S3057, when u = 0, equation (5) can be transformed into the following expression:

[0187]

[0188] When u = 50, equation (5) can be transformed into the following expression:

[0189]

[0190] When u = 100, equation (5) can be transformed into the following expression:

[0191]

[0192] By analogy, the index values ​​corresponding to all preset frequency values ​​can be obtained.

[0193] Step S306: Determine the index value corresponding to the target frequency in the third data set as the feature value corresponding to the first preset index.

[0194] Step S307: Determine the stability of the target device during operation based on the feature value and weight value.

[0195] The target frequency is either the highest frequency value in the third data set or any frequency value in the third data set.

[0196] For details on steps S306 to S307, please refer to [link / reference]. Figure 1 Steps S106 to S107 of the illustrated embodiment will not be described again here.

[0197] The method for assessing the operational stability of a target device provided in this embodiment requires performance data to meet certain conditions, namely, to be within a specified symmetrical interval, when performing time-frequency conversion on time-series data. Therefore, by scaling the indicator values ​​included in the performance data in the dataset, the scaled indicator values ​​can satisfy the time-frequency conversion conditions. Furthermore, during the time-frequency conversion of indicator values, the indicator value corresponding to each frequency value is determined based on the indicator values ​​corresponding to all time points. Therefore, the indicator value corresponding to each frequency value can represent the fluctuation level of a preset indicator. Since the indicator value corresponding to the highest frequency value can represent the fluctuation level of the preset indicator within a short time span, the indicator value corresponding to the highest frequency value can be used as a feature value of the preset indicator to more accurately represent the real-time operational stability of the target device. Furthermore, the operational stability of the target device determined by using this feature value corresponding to each preset indicator is also more accurate.

[0198] This embodiment also provides a device for determining the operational stability of a target device. This device is used to implement the above embodiments and preferred embodiments, and details already described will not be repeated. As used below, the term "module" can be a combination of software and / or hardware that implements a predetermined function. Although the device described in the following embodiments is preferably implemented in software, hardware implementation, or a combination of software and hardware, is also possible and contemplated.

[0199] This embodiment provides a device for determining the operational stability of a target device, such as... Figure 4 As shown, it includes:

[0200] The acquisition module 401 is used to acquire a data set corresponding to each of the multiple preset indicators when a trigger command is received, wherein the data set includes performance data in the target device corresponding to the preset indicators;

[0201] The determining module 402 is used to determine the weight factor corresponding to the first preset indicator based on the first data set corresponding to the first preset indicator, wherein the first preset indicator is any one of a plurality of preset indicators; and to determine the weight value corresponding to each preset indicator based on the weight factors corresponding to all preset indicators respectively.

[0202] The scaling module 403 is used to scale the indicator values ​​included in each performance data in the first data set to obtain the second data set;

[0203] The conversion module 404 is used to input the second data set into the time-frequency conversion model to obtain the third data set;

[0204] The determination module 402 is used to determine the index value corresponding to the target frequency in the third data set as the feature value corresponding to the first preset index; and to determine the stability of the target device during operation based on the feature value and the weight value.

[0205] In some alternative implementations, the determining module 402 is used for:

[0206] The total weight factor is constructed based on the weight factors corresponding to all preset indicators.

[0207] Based on the weight factor corresponding to each preset indicator and the total weight factor, determine the weight value corresponding to each preset indicator.

[0208] In some alternative implementations, before scaling the metric values ​​in each performance data point in the first dataset to obtain the second dataset, the method further includes:

[0209] Iterate through the first data set to determine if any performance data corresponding to the target time point is missing from the first data set;

[0210] If the performance data corresponding to the target time point is missing in the first dataset, retrieve the performance data corresponding to the target time point.

[0211] Based on the performance data corresponding to the target time point, the first data set is updated to obtain the fourth data set, so that the indicator values ​​in each performance data in the fourth data set can be scaled to obtain the second data set.

[0212] In some alternative implementations, the acquisition module 401 is used for:

[0213] Determine the previous and next time points corresponding to the target time point;

[0214] Extract the first performance data corresponding to the previous time point and the second performance data corresponding to the next time point from the first data set;

[0215] Determine the time interpolation corresponding to the target time point based on the number of target time points between the previous time point and the next time point;

[0216] Based on the index values ​​included in the first performance data, the index values ​​included in the second performance data, the time interpolation corresponding to the target time point, and the pre-acquired interpolation model, determine the index values ​​included in the performance data corresponding to the target time point.

[0217] The performance data corresponding to the target time point is constructed by using the indicator values ​​and the target time point in the performance data corresponding to the target time point.

[0218] or,

[0219] Extract the performance data corresponding to the target time point from the performance data stored on the target device.

[0220] In some alternative implementations, the scaling module 403 is used for:

[0221] Get the scaling range;

[0222] In the first dataset, extract the maximum and minimum index values ​​corresponding to the first preset index;

[0223] Determine the absolute value of the difference between the maximum and minimum indicator values;

[0224] The scaling factor is determined based on the absolute value of the difference and the length of the scaling interval.

[0225] Based on the scaling factor, any boundary interval value of the scaling interval, and the minimum index value, the index values ​​included in the performance data in the first data set are scaled to obtain the second data set.

[0226] In some alternative implementations, the conversion module 404 is used for:

[0227] Count the number of performance data points in the second dataset;

[0228] Determine whether each preset frequency value among multiple preset frequency values ​​is equal to zero;

[0229] When the preset frequency value is equal to zero, the compensation coefficient corresponding to the preset frequency value is determined based on the first preset coefficient and the number of performance data in the second data set.

[0230] or,

[0231] When the preset frequency value is not equal to zero, the compensation coefficient corresponding to the preset frequency value is determined based on the second preset coefficient and the number of performance data in the second data set.

[0232] Based on the index values ​​included in the i-th performance data in the second data set, the number of performance data in the second data set, and the value corresponding to i, determine the cosine transformation value corresponding to the i-th performance data, where i is a positive integer;

[0233] The target cosine transform value is obtained by summing the cosine transform values ​​corresponding to each performance data point.

[0234] Based on the target cosine transform value and the compensation coefficient corresponding to the preset frequency value, determine the index value corresponding to the preset frequency value;

[0235] Multiple preset frequency values ​​and the corresponding index values ​​for each preset frequency value constitute a third data set.

[0236] In some alternative implementations, the target frequency is the highest frequency value in the third data set, or any frequency value in the third data set.

[0237] In some optional implementations, the conversion module 404 is represented by the following expression:

[0238]

[0239]

[0240] Where u is any one of the preset frequency values, F(u) is the index value corresponding to the preset frequency value, c(u) is the compensation coefficient corresponding to the preset frequency value, i is the sequence number of any performance data in the second data set, N is the number of performance data in the second data set, and f(i) is the index value included in the i-th performance data in the second data set.

[0241] The multiple preset frequency values ​​and the index values ​​corresponding to each preset frequency value constitute the third data set.

[0242] Further functional descriptions of the above modules are the same as those in the corresponding embodiments described above, and will not be repeated here.

[0243] In this embodiment, the device for determining the stability of the target device is presented in the form of a functional unit. Here, a unit refers to an ASIC (Application Specific Integrated Circuit) circuit, a processor and memory that execute one or more software or fixed programs, and / or other devices that can provide the above functions.

[0244] This invention also provides a computer device having the above-described features. Figure 4 The device shown is for determining the stability of the target equipment's operation.

[0245] Please see Figure 5 , Figure 5 This is a schematic diagram of the structure of a computer device provided in an optional embodiment of the present invention, such as... Figure 5As shown, the computer device includes one or more processors 10, memory 20, and interfaces for connecting the components, including high-speed interfaces and low-speed interfaces. The components communicate with each other via different buses and can be mounted on a common motherboard or otherwise installed as needed. The processors can process instructions executed within the computer device, including instructions stored in or on memory to display graphical information of a GUI on external input / output devices (such as display devices coupled to the interfaces). In some alternative implementations, multiple processors and / or multiple buses can be used with multiple memories and multiple memory modules, if desired. Similarly, multiple computer devices can be connected, each providing some of the necessary operations (e.g., as a server array, a group of blade servers, or a multiprocessor system). Figure 5 Take a processor 10 as an example.

[0246] Processor 10 may be a central processing unit, a network processor, or a combination thereof. Processor 10 may further include a hardware chip. The hardware chip may be an application-specific integrated circuit (ASIC), a programmable logic device (PLD), or a combination thereof. The programmable logic device may be a complex programmable logic device (CAMP), a field-programmable gate array (FPGA), a general-purpose array logic (GDA), or any combination thereof.

[0247] The memory 20 stores instructions executable by at least one processor 10 to cause the at least one processor 10 to perform the method shown in the above embodiments.

[0248] The memory 20 may include a program storage area and a data storage area. The program storage area may store the operating system and applications required for at least one function; the data storage area may store data created based on the use of the computer device. Furthermore, the memory 20 may include high-speed random access memory and may also include non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid-state storage device. In some alternative embodiments, the memory 20 may optionally include memory remotely located relative to the processor 10, and these remote memories may be connected to the computer device via a network. Examples of such networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.

[0249] The memory 20 may include volatile memory, such as random access memory; the memory may also include non-volatile memory, such as flash memory, hard disk or solid-state drive; the memory 20 may also include a combination of the above types of memory.

[0250] The computer device also includes a communication interface 30 for communicating with other devices or communication networks.

[0251] This invention also provides a computer-readable storage medium. The methods described above according to embodiments of the invention can be implemented in hardware or firmware, or implemented as computer code that can be recorded on a storage medium, or implemented as computer code downloaded via a network and originally stored on a remote storage medium or a non-transitory machine-readable storage medium and then stored on a local storage medium. Thus, the methods described herein can be processed by software stored on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. The storage medium can be a magnetic disk, optical disk, read-only memory, random access memory, flash memory, hard disk, or solid-state drive, etc.; further, the storage medium can also include combinations of the above types of memory. It is understood that computers, processors, microprocessor controllers, or programmable hardware include storage components capable of storing or receiving software or computer code, which, when accessed and executed by the computer, processor, or hardware, implements the methods shown in the above embodiments.

[0252] Although embodiments of the invention have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of the invention, and such modifications and variations all fall within the scope defined by the appended claims.

Claims

1. A method for determining the operational stability of a target device, characterized in that, The method includes: When a trigger command is received, a data set corresponding to each of the multiple preset indicators is obtained, wherein the data set includes performance data in the target device corresponding to the preset indicator; Based on the first data set corresponding to the first preset indicator, a weight factor corresponding to the first preset indicator is determined, wherein the first preset indicator is any one of the plurality of preset indicators; Based on the weight factors corresponding to all the preset indicators, determine the weight value corresponding to each preset indicator; The performance data in the first dataset is scaled to obtain the second dataset. The second data set is input into the time-frequency conversion model to obtain the third data set; The index value corresponding to the target frequency in the third data set is determined as the feature value corresponding to the first preset index; The stability of the target device during operation is determined based on the feature value and the weight value.

2. The method according to claim 1, characterized in that, Based on the weight factors corresponding to all the preset indicators, determine the weight value corresponding to each preset indicator, including: The total weight factor is constructed based on the weight factors corresponding to all the preset indicators; Based on the weight factor corresponding to each of the preset indicators and the total weight factor, the weight value corresponding to each of the preset indicators is determined.

3. The method according to claim 1 or 2, characterized in that, Before scaling the metric values ​​in each performance data set in the first data set to obtain the second data set, the method further includes: Traverse the first data set to determine whether any performance data corresponding to the target time point is missing from the first data set; If the performance data corresponding to the target time point is missing in the first data set, the performance data corresponding to the target time point is obtained. Based on the performance data corresponding to the target time point, the first data set is updated to obtain a fourth data set, so that the indicator values ​​in each performance data in the fourth data set can be scaled to obtain the second data set.

4. The method according to claim 3, characterized in that, When performance data corresponding to the target time point is missing from the first data set, obtaining the performance data corresponding to the target time point includes: Determine the previous and next time points corresponding to the target time point; Extract the first performance data corresponding to the previous time point and the second performance data corresponding to the next time point from the first data set; Based on the number of target time points between the previous time point and the next time point, determine the time interpolation corresponding to the target time point; Based on the index values ​​included in the first performance data, the index values ​​included in the second performance data, the time interpolation corresponding to the target time point, and the pre-acquired interpolation model, determine the index values ​​included in the performance data corresponding to the target time point. The performance data corresponding to the target time point is constructed by using the index values ​​included in the performance data corresponding to the target time point and the target time point; or, Extract the performance data corresponding to the target time point from the performance data stored in the target device.

5. The method according to claim 1, characterized in that, The scaling operation performed on the metric values ​​included in each performance data point in the first data set to obtain the second data set includes: Get the scaling range; In the first data set, extract the maximum and minimum indicator values ​​corresponding to the first preset indicator; Determine the absolute value of the difference between the maximum index value and the minimum index value; The scaling factor is determined based on the absolute value of the difference and the length of the scaling interval. Based on the scaling factor, any boundary interval value of the scaling interval, and the minimum index value, the performance data included in the first data set is scaled to obtain the second data set.

6. The method according to claim 5, characterized in that, The step of inputting the second data set into the time-frequency conversion model to obtain the third data set includes: Count the number of performance data in the second dataset; Determine whether each preset frequency value among multiple preset frequency values ​​is equal to zero; When the preset frequency value is equal to zero, a compensation coefficient corresponding to the preset frequency value is determined based on the first preset coefficient and the number of performance data in the second data set. or, When the preset frequency value is not equal to zero, a compensation coefficient corresponding to the preset frequency value is determined based on the second preset coefficient and the number of performance data in the second data set. Based on the index values ​​included in the i-th performance data in the second data set, the number of performance data in the second data set, and the value corresponding to i, determine the cosine transformation value corresponding to the i-th performance data, where i is a positive integer; The target cosine transform value is obtained by summing the cosine transform values ​​corresponding to each performance data point. Based on the target cosine transform value and the compensation coefficient corresponding to the preset frequency value, determine the index value corresponding to the preset frequency value; The multiple preset frequency values ​​and the index values ​​corresponding to each preset frequency value constitute the third data set.

7. The method according to claim 6, characterized in that, The target frequency is the highest frequency value in the third data set, or any frequency value in the third data set.

8. The method according to claim 6, characterized in that, The second data set is input into the time-frequency conversion model to obtain the third data set. It can be represented by the following expression: Where u is any one of the preset frequency values, F(u) is the index value corresponding to the preset frequency value, c(u) is the compensation coefficient corresponding to the preset frequency value, i is the sequence number of any performance data in the second data set, N is the number of performance data in the second data set, and f(i) is the index value included in the i-th performance data in the second data set. The multiple preset frequency values ​​and the index values ​​corresponding to each preset frequency value constitute the third data set.

9. A device for determining the operational stability of a target device, characterized in that, The device includes: The acquisition module is used to acquire a data set corresponding to each of a plurality of preset indicators when a trigger command is received, wherein the data set includes performance data in the target device corresponding to the preset indicator; The determining module is used to determine a weight factor corresponding to the first preset indicator based on a first data set corresponding to the first preset indicator, wherein the first preset indicator is any one of a plurality of preset indicators; and to determine a weight value corresponding to each preset indicator based on the weight factors corresponding to all the preset indicators respectively. The scaling module is used to scale the indicator values ​​included in each performance data in the first data set to obtain the second data set. The conversion module is used to input the second data set into the time-frequency conversion model to obtain the third data set; The determining module is used to determine the index value corresponding to the target frequency in the third data set as the feature value corresponding to the first preset index; and to determine the stability of the target device during operation based on the feature value and the weight value.

10. A computer device, characterized in that, include: A memory and a processor are communicatively connected, the memory stores computer instructions, and the processor executes the computer instructions to perform the method for determining the operational stability of the target device 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 computer instructions for causing the computer to execute the method for determining the operational stability of the target device as described in any one of claims 1 to 8.

Images