Method, device and equipment for determining equipment fault factors

Abstract

The invention provides an equipment fault factor determination method, device and equipment, which are applied to computing equipment, and the method comprises the following steps: obtaining equipment information of a plurality of target equipment, the equipment information comprising feature information of the target equipment in a plurality of feature dimensions and historical fault information; determining a plurality of target feature dimensions in the plurality of feature dimensions according to the computing performance of the computing device and the device information of each target device; determining a plurality of dimension combinations corresponding to the plurality of target feature dimensions, and determining a plurality of target dimension combinations in the plurality of dimension combinations according to the device information of the plurality of target devices, the dimension combinations comprising at least one target feature dimension, the target dimension combination is a fault factor combination of which the influence degree on the equipment fault is greater than or equal to a preset threshold value. And the accuracy of determining the target dimension combination is improved.

Description

Technical Field

[0001] The present application relates to the field of big data, and in particular to a method, device and equipment for determining equipment failure factors. Background Art

[0002] Equipment may crash during operation. In order to reduce the crash rate, it is necessary to analyze the characteristic dimensions (i.e., influencing factors) of equipment crashes.

[0003] The characteristic dimensions of equipment downtime may be varied. In the related art, the target dimension combination that causes equipment downtime is usually determined based on human experience in multiple characteristic dimensions. However, in the above method, the accuracy of determining the target dimension combination based on human experience is limited, and there may be potential connections between the characteristic dimensions, which further leads to low accuracy in determining the target dimension combination. Summary of the invention

[0004] Multiple aspects of the present application provide a method, apparatus and device for determining equipment failure factors, so as to improve the accuracy of determining target dimension combinations.

[0005] In a first aspect, an embodiment of the present application provides a method for determining a device failure factor, which is applied to a computing device, and the method includes:

[0006] Acquire device information of multiple target devices, wherein the device information includes feature information and historical fault information of the target devices in multiple feature dimensions;

[0007] Determining a plurality of target feature dimensions from the plurality of feature dimensions according to the computing performance of the computing device and the device information of each target device;

[0008] Determine multiple dimensional combinations corresponding to the multiple target feature dimensions, and determine multiple target dimension combinations in the multiple dimensional combinations based on the device information of the multiple target devices, the dimension combinations including at least one of the target feature dimensions, and the target dimension combinations are combinations of fault factors whose impact on device failure is greater than or equal to a preset threshold.

[0009] In a possible implementation, according to the computing performance of the computing device and the device information of each target device, a plurality of target feature dimensions are determined from the plurality of feature dimensions, including:

[0010] Determine a maximum processing capacity corresponding to the computing performance of the computing device, and determine a maximum number of dimensions based on the maximum processing capacity;

[0011] The plurality of target feature dimensions are determined from the plurality of feature dimensions according to the number of the plurality of feature dimensions, the maximum processing amount, the maximum number of dimensions, and device information of each target device.

[0012] In a possible implementation, determining the multiple target feature dimensions from the multiple feature dimensions according to the number of the multiple feature dimensions, the maximum processing amount, the maximum number of dimensions, and device information of each target device includes:

[0013] Determine a combination depth according to the number of the plurality of feature dimensions and the maximum processing amount, wherein the combination depth is used to indicate the number of dimensions included in the dimension combination;

[0014] Determining whether to eliminate the multiple feature dimensions according to the combination depth, the number of the multiple feature dimensions, and the maximum number of dimensions;

[0015] If yes, performing at least one elimination process on the multiple feature dimensions according to the combination depth, the number of the multiple feature dimensions, the maximum number of dimensions, and the device information of each target device to obtain the multiple target feature dimensions;

[0016] If not, the multiple feature dimensions are determined as the multiple target feature dimensions.

[0017] In a possible implementation, according to the combination depth, the number of the multiple feature dimensions, the maximum number of dimensions, and the device information of each target device, the multiple feature dimensions are eliminated at least once to obtain the multiple target feature dimensions, including:

[0018] Determine a first elimination number according to the number of the multiple feature dimensions, and perform a first elimination process on the multiple feature dimensions according to the first elimination number, the combination depth, and the device information of each target device to obtain multiple first remaining dimensions;

[0019] According to the number of the multiple i-th remaining dimensions, the combination depth is updated to obtain the i+1th updated combination depth, and according to the i+1th updated combination depth, the number of the multiple i-th remaining dimensions, i+1, and the maximum number of dimensions, it is determined whether to perform the i+1th elimination process, and if so, the i+1th elimination number is determined according to the number of the i-th remaining dimensions and i+1, and according to the i+1th elimination number, the i+1th updated combination depth and the device information of each target device, the multiple i-th remaining dimensions are eliminated to obtain multiple i+1th remaining dimensions; if not, the multiple i-th remaining dimensions are determined as the multiple target feature dimensions;

[0020] Here, the i is an integer starting from 1 in sequence until the multiple target feature dimensions are obtained.

[0021] In a possible implementation, the plurality of feature dimensions are first eliminated according to the first elimination quantity, the combination depth, and the device information of each target device to obtain a plurality of first remaining dimensions, including:

[0022] Determining the influence of each feature dimension on the device failure according to the combination depth and the device information of each target device;

[0023] Sorting the multiple feature dimensions in order of influence from small to large;

[0024] Among the multiple feature dimensions, the first elimination number of feature dimensions is deleted to obtain the multiple first remaining dimensions.

[0025] In a possible implementation manner, for any feature dimension, determining the influence of the feature dimension on the device failure according to the combination depth and the device information of each target device includes:

[0026] Determine, according to the device information of each target device, a plurality of feature sets corresponding to the feature dimension, each feature set respectively including a plurality of combined feature information, the combined feature information including the feature information corresponding to the feature dimension, or the combined feature information including the feature information corresponding to the feature dimension and feature information corresponding to other feature dimensions combined with the feature dimension;

[0027] For each feature set, determining the influence of each combination of feature information in the feature set on the equipment failure, and determining the statistical value of the influence of each combination of feature information on the equipment failure as the influence of the feature set on the equipment failure;

[0028] According to the influence of each feature set on the device failure, the influence of the feature dimension on the device failure is determined.

[0029] In a possible implementation manner, for any one first combination of feature information in the feature set; determining the influence of the first combination of feature information on the device failure includes:

[0030] Determine at least one first target device from the multiple target devices according to the first combined feature information, wherein device information of the first target device includes the first combined feature information;

[0031] Determining a first number of failed devices in the at least one first target device, and determining a second number of the at least one first target device;

[0032] The ratio of the first quantity to the second quantity is determined as the influence of the first combination feature information on the equipment failure.

[0033] In a possible implementation, determining the influence of the feature dimension on the device failure according to the influence of each feature set on the device failure includes:

[0034] For any feature set, determining a third number of feature dimensions in the feature set;

[0035] Determine the initial influence of the feature dimension in the feature set by taking the ratio of the influence of the feature set on the device failure and the third quantity;

[0036] The statistical value of the initial influence of the feature dimension in each feature set is determined as the influence of the feature dimension on the device failure.

[0037] In a possible implementation, judging whether to eliminate the multiple feature dimensions according to the combination depth, the number of the multiple feature dimensions, and the maximum number of dimensions includes:

[0038] If the combination depth is equal to the number of the plurality of feature dimensions, determining not to eliminate the plurality of feature dimensions;

[0039] If the combination depth is less than the number of the plurality of feature dimensions, determining whether one half of the number of the plurality of feature dimensions is greater than or equal to the maximum number of dimensions;

[0040] If yes, determine to eliminate the multiple feature dimensions;

[0041] If not, it is determined that the multiple feature dimensions will not be eliminated.

[0042] In a possible implementation, determining multiple target dimension combinations from the multiple dimension combinations according to the device information of the multiple target devices includes:

[0043] Determining, based on the device information of the multiple target devices, a target impact degree of each dimension combination on the device failure;

[0044] Among the multiple dimension combinations, the dimension combinations whose target impact on the equipment failure is greater than or equal to the preset threshold are determined as the multiple target dimension combinations.

[0045] In a possible implementation manner, for any dimension combination, determining the target impact of the dimension combination on the device failure according to the device information of the multiple target devices includes:

[0046] Determine that the dimension combination corresponds to a plurality of combination feature information, wherein the combination feature information includes feature information corresponding to each feature dimension in the dimension combination;

[0047] Determine the impact of each combination of characteristic information on equipment failure;

[0048] Determine the statistical value of the influence of each combination of characteristic information on the equipment failure as the initial influence of the dimension combination on the equipment failure;

[0049] According to the initial impacts of the multiple dimensional combinations, a target impact of each dimensional combination on the device failure is determined.

[0050] In a possible implementation manner, for any second piece of combined feature information among the plurality of combined feature information, determining the influence of the second piece of combined feature information on the device failure includes:

[0051] determining at least one second target device from the plurality of target devices according to the second combined feature information, wherein the device information of the second target device includes the second combined feature information;

[0052] determining a fourth number of failed devices in the at least one second target device, determining a fifth number of the at least one second target device;

[0053] The ratio of the fourth number to the fifth number is determined as the influence of the second combination feature information on the equipment failure.

[0054] In a possible implementation manner, for any first dimensional combination among the multiple dimensional combinations, determining a target impact degree of the first dimensional combination on a device failure according to the initial impact degrees of the multiple dimensional combinations includes:

[0055] Determine, among the multiple dimension combinations, multiple second dimension combinations corresponding to the first dimension combination, wherein the multiple feature dimensions included in the second dimension combination include the multiple feature dimensions in the first dimension combination;

[0056] For any second dimensional combination, a ratio of a sixth number of characteristic dimensions in the first dimensional combination to a seventh number of characteristic dimensions in the second dimensional combination is determined as a characteristic ratio of the first dimensional combination in the second dimensional combination;

[0057] Determine the product of the initial impact of the second dimensional combination on the equipment failure and the characteristic ratio as the intermediate impact of the first dimensional combination in the second dimensional combination;

[0058] The statistical value of the intermediate influence of the first dimensional combination in each second dimensional combination is determined as the target influence of the first dimensional combination on the equipment failure.

[0059] In a second aspect, an embodiment of the present application provides a device for determining a device failure factor, which is applied to a computing device. The device includes: an acquisition module, a first determination module, and a second determination module, wherein:

[0060] The acquisition module is used to acquire device information of multiple target devices, wherein the device information includes feature information and historical fault information of the target devices in multiple feature dimensions;

[0061] The first determining module is used to determine a plurality of target feature dimensions from the plurality of feature dimensions according to the computing performance of the computing device and the device information of each target device;

[0062] The second determination module is used to determine multiple dimension combinations corresponding to the multiple target feature dimensions, and determine multiple target dimension combinations in the multiple dimension combinations according to the device information of the multiple target devices, wherein the dimension combinations include at least one of the target feature dimensions, and the target dimension combinations are combinations of fault factors whose impact on the equipment failure is greater than or equal to a preset threshold.

[0063] In a possible implementation manner, the first determining module is specifically configured to:

[0064] Determine a maximum processing capacity corresponding to the computing performance of the computing device, and determine a maximum number of dimensions based on the maximum processing capacity;

[0065] The plurality of target feature dimensions are determined from the plurality of feature dimensions according to the number of the plurality of feature dimensions, the maximum processing amount, the maximum number of dimensions, and device information of each target device.

[0066] In a possible implementation manner, the first determining module is specifically configured to:

[0067] Determine a combination depth according to the number of the plurality of feature dimensions and the maximum processing amount, wherein the combination depth is used to indicate the number of dimensions included in the dimension combination;

[0068] Determining whether to eliminate the multiple feature dimensions according to the combination depth, the number of the multiple feature dimensions, and the maximum number of dimensions;

[0069] If yes, performing at least one elimination process on the multiple feature dimensions according to the combination depth, the number of the multiple feature dimensions, the maximum number of dimensions, and the device information of each target device to obtain the multiple target feature dimensions;

[0070] If not, the multiple feature dimensions are determined as the multiple target feature dimensions.

[0071] In a possible implementation manner, the first determining module is specifically configured to:

[0072] Determine a first elimination number according to the number of the multiple feature dimensions, and perform a first elimination process on the multiple feature dimensions according to the first elimination number, the combination depth, and the device information of each target device to obtain multiple first remaining dimensions;

[0073] According to the number of the multiple i-th remaining dimensions, the combination depth is updated to obtain the i+1th updated combination depth, and according to the i+1th updated combination depth, the number of the multiple i-th remaining dimensions, i+1, and the maximum number of dimensions, it is determined whether to perform the i+1th elimination process, and if so, the i+1th elimination number is determined according to the number of the i-th remaining dimensions and i+1, and according to the i+1th elimination number, the i+1th updated combination depth and the device information of each target device, the multiple i-th remaining dimensions are eliminated to obtain multiple i+1th remaining dimensions; if not, the multiple i-th remaining dimensions are determined as the multiple target feature dimensions;

[0074] Here, the i is an integer starting from 1 in sequence until the multiple target feature dimensions are obtained.

[0075] In a possible implementation manner, the first determining module is specifically configured to:

[0076] Determining the influence of each feature dimension on the device failure according to the combination depth and the device information of each target device;

[0077] Sorting the multiple feature dimensions in order of influence from small to large;

[0078] Among the multiple feature dimensions, the first elimination number of feature dimensions is deleted to obtain the multiple first remaining dimensions.

[0079] In a possible implementation manner, for any feature dimension, the first determining module is specifically configured to:

[0080] Determine, according to the device information of each target device, a plurality of feature sets corresponding to the feature dimension, each feature set respectively including a plurality of combined feature information, the combined feature information including the feature information corresponding to the feature dimension, or the combined feature information including the feature information corresponding to the feature dimension and feature information corresponding to other feature dimensions combined with the feature dimension;

[0081] For each feature set, determining the influence of each combination of feature information in the feature set on the equipment failure, and determining the statistical value of the influence of each combination of feature information on the equipment failure as the influence of the feature set on the equipment failure;

[0082] According to the influence of each feature set on the device failure, the influence of the feature dimension on the device failure is determined.

[0083] In a possible implementation manner, the first determining module is specifically configured to:

[0084] Determine at least one first target device from the multiple target devices according to the first combined feature information, wherein device information of the first target device includes the first combined feature information;

[0085] Determining a first number of failed devices in the at least one first target device, and determining a second number of the at least one first target device;

[0086] The ratio of the first quantity to the second quantity is determined as the influence of the first combination feature information on the equipment failure.

[0087] In a possible implementation manner, the first determining module is specifically configured to:

[0088] For any feature set, determining a third number of feature dimensions in the feature set;

[0089] Determine the initial influence of the feature dimension in the feature set by taking the ratio of the influence of the feature set on the device failure and the third quantity;

[0090] The statistical value of the initial influence of the feature dimension in each feature set is determined as the influence of the feature dimension on the device failure.

[0091] In a possible implementation manner, the first determining module is specifically configured to:

[0092] If the combination depth is equal to the number of the plurality of feature dimensions, determining not to eliminate the plurality of feature dimensions;

[0093] If the combination depth is less than the number of the plurality of feature dimensions, determining whether one half of the number of the plurality of feature dimensions is greater than or equal to the maximum number of dimensions;

[0094] If yes, determine to eliminate the multiple feature dimensions;

[0095] If not, it is determined that the multiple feature dimensions will not be eliminated.

[0096] In a possible implementation manner, the second determining module is specifically configured to:

[0097] Determining, based on the device information of the multiple target devices, a target impact degree of each dimension combination on the device failure;

[0098] Among the multiple dimension combinations, the dimension combinations whose target impact on the equipment failure is greater than or equal to the preset threshold are determined as the multiple target dimension combinations.

[0099] In a possible implementation manner, for any combination of dimensions, the second determining module is specifically configured to:

[0100] Determine that the dimension combination corresponds to a plurality of combination feature information, wherein the combination feature information includes feature information corresponding to each feature dimension in the dimension combination;

[0101] Determine the impact of each combination of characteristic information on equipment failure;

[0102] Determine the statistical value of the influence of each combination of characteristic information on the equipment failure as the initial influence of the dimension combination on the equipment failure;

[0103] According to the initial impacts of the multiple dimensional combinations, a target impact of each dimensional combination on the device failure is determined.

[0104] In a possible implementation manner, for any second piece of combined feature information among the plurality of combined feature information, the second determining module is specifically configured to:

[0105] determining at least one second target device from the plurality of target devices according to the second combined feature information, wherein the device information of the second target device includes the second combined feature information;

[0106] determining a fourth number of failed devices in the at least one second target device, determining a fifth number of the at least one second target device;

[0107] The ratio of the fourth number to the fifth number is determined as the influence of the second combination feature information on the equipment failure.

[0108] In a possible implementation manner, for any first dimension combination among the multiple dimension combinations, the second determining module is specifically configured to:

[0109] Determine, among the multiple dimension combinations, multiple second dimension combinations corresponding to the first dimension combination, wherein the multiple feature dimensions included in the second dimension combination include the multiple feature dimensions in the first dimension combination;

[0110] For any second dimensional combination, a ratio of a sixth number of characteristic dimensions in the first dimensional combination to a seventh number of characteristic dimensions in the second dimensional combination is determined as a characteristic ratio of the first dimensional combination in the second dimensional combination;

[0111] Determine the product of the initial impact of the second dimensional combination on the equipment failure and the characteristic ratio as the intermediate impact of the first dimensional combination in the second dimensional combination;

[0112] The statistical value of the intermediate influence of the first dimensional combination in each second dimensional combination is determined as the target influence of the first dimensional combination on the equipment failure.

[0113] In a third aspect, an embodiment of the present application provides a computing device, including: a memory and a processor;

[0114] The memory stores computer-executable instructions;

[0115] The processor executes the computer-executable instructions stored in the memory, so that the processor performs the method described in any one of the first aspects.

[0116] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores computer-executable instructions, and when the computer-executable instructions are executed by a processor, they are used to implement the method described in any one of the first aspects.

[0117] In a fifth aspect, an embodiment of the present application provides a computer program product, including a computer program, which implements the method shown in any one of the first aspects when executed by a processor.

[0118] The embodiments of the present application provide a method, apparatus and device for determining equipment failure factors. The computing device can obtain the equipment information of multiple target devices, and determine multiple target feature dimensions in multiple feature dimensions based on the computing performance of the computing device and the equipment information of each target device, and then determine multiple dimension combinations corresponding to the multiple target feature dimensions, and determine multiple target dimension combinations in multiple dimension combinations based on the equipment information of the multiple target devices. Through the technical solution of the present application, the combination depth can be adaptively adjusted to prevent the dimension explosion problem; multiple feature dimensions can be eliminated at least once to determine multiple target feature dimensions; and the target influence of multiple dimension combinations on equipment failure can be calculated twice by iteration, eliminating the correlation between feature dimensions, so as to determine the target dimension combination in multiple dimension combinations, which improves the accuracy of determining the target dimension combination compared to manually determining multiple target dimension combinations. BRIEF DESCRIPTION OF THE DRAWINGS

[0119] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0120] Figure 1 A schematic diagram of a scenario provided for an exemplary embodiment of the present application;

[0121] Figure 2 A flowchart of a method for determining a device failure factor provided by an exemplary embodiment of the present application;

[0122] Figure 3 A flowchart of another method for determining a device failure factor provided by an exemplary embodiment of the present application;

[0123] Figure 4 A schematic diagram of a flow chart of determining multiple target feature dimensions from multiple feature dimensions provided for an exemplary embodiment of the present application;

[0124] Figure 5 A schematic diagram of the structure of a device for determining equipment failure factors provided by an exemplary embodiment of the present application;

[0125] Figure 6 A schematic diagram of the structure of a computing device provided for an exemplary embodiment of the present application. DETAILED DESCRIPTION

[0126] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant laws, regulations and standards, and provide corresponding operation entrances for users to choose to authorize or refuse.

[0127] In order to make the purpose, technical solution and advantages of the present application clearer, the technical solution of the present application will be clearly and completely described below in combination with the specific embodiments of the present application and the corresponding drawings. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present application.

[0128] Figure 1 A schematic diagram of a scenario provided for an exemplary embodiment of the present application. Figure 1 , the computing device can obtain device information of multiple target devices. The device information may include feature information and historical fault information of the target device in multiple feature dimensions.

[0129] For example, the device information of multiple target devices may include device information 1 of target device 1, device information 2 of target device 2, ..., device information k of target device k. Among them, device information 1 may include processor model 1, device model 2, disk model 1, usage time of 10,000 hours, and downtime of target device 1.

[0130] The computing device may determine multiple dimension combinations according to computing performance and device information of multiple target devices, and then may determine multiple target dimension combinations from the multiple dimension combinations. The multiple target dimension combinations are combinations that are likely to cause target device failures.

[0131] For example, the computing device may determine 1000 dimension combinations according to the computing performance and the device information of the k target devices, namely dimension combination 1, dimension combination 2, ..., dimension combination 1000, and then determine 3 target dimension combinations from the 1000 dimension combinations, namely dimension combination 2, dimension combination 50, and dimension combination 112. For example, dimension combination 2 may include a processor model, dimension combination 50 may be a combination of a device model and a disk model, and dimension combination 112 may be a combination of a processor model, a device model, and a disk model.

[0132] In the related art, the target dimension combination that causes equipment downtime is usually determined from multiple feature dimensions based on human experience. However, in the above method, the accuracy of determining the target dimension combination based on human experience is limited, and there may be potential connections between feature dimensions, which further leads to low accuracy in determining the target dimension combination.

[0133] In an embodiment of the present application, the computing device can determine multiple target feature dimensions from multiple feature dimensions based on computing performance and device information of multiple target devices, and determine multiple dimension combinations corresponding to the multiple target feature dimensions, and then determine multiple target dimension combinations from the multiple dimension combinations. Since multiple target feature dimensions can be determined iteratively multiple times in multiple feature dimensions, a target dimension combination whose target impact on device failure is greater than or equal to a preset threshold can be determined from the multiple dimension combinations corresponding to the multiple target feature dimensions, the accuracy of determining the target dimension combination is improved compared to manually determining multiple target dimension combinations.

[0134] The technical solutions shown in the present application are described in detail below through specific embodiments. It should be noted that the following embodiments can exist independently or in combination with each other, and the same or similar contents will not be described repeatedly in different embodiments.

[0135] Figure 2A flowchart of a method for determining a device failure factor provided by an exemplary embodiment of the present application. Figure 2 , the method may include:

[0136] S201. Obtain device information of multiple target devices.

[0137] The execution subject of the embodiment of the present application may be a computing device, or may be a device for determining a device failure factor set in a computing device. The device for determining a device failure factor may be implemented by software, or may be implemented by a combination of software and hardware. The device for determining a device failure factor may be a processor in a computing device. For ease of understanding, the following description is made by taking the execution subject as a computing device as an example.

[0138] The device information may include feature information and historical fault information of the target device in multiple feature dimensions.

[0139] Optionally, multiple feature dimensions may be pre-selected manually. For any feature dimension, the target device has corresponding feature information under the feature dimension. For example, multiple feature dimensions may include processor model, device model, disk model, and usage time. Then, the feature information of the target device in multiple feature dimensions may include processor model 1, device model 1, disk model 3, and usage time of 10,000 hours.

[0140] The historical fault information refers to whether the device has experienced a downtime. Optionally, the historical fault information is represented by 0 or 1, 0 represents no downtime, and 1 represents a downtime.

[0141] For example, if there are 500 target devices, and if 20 feature dimensions are preset, including processor model, device model, disk model, usage time, and other 16 feature dimensions, the computing device can obtain device information of the 500 target devices, as shown in Table 1:

[0142] Table 1

[0143]

[0144] S202: Determine multiple target feature dimensions from multiple feature dimensions according to the computing performance of the computing device and the device information of each target device.

[0145] In an optional embodiment, multiple target feature dimensions can be determined among multiple feature dimensions based on the computing performance of the computing device and the device information of each target device in the following manner: determine the maximum processing capacity corresponding to the computing performance of the computing device, and determine the maximum number of dimensions based on the maximum processing capacity; determine multiple target feature dimensions among multiple feature dimensions based on the number of multiple feature dimensions, the maximum processing capacity, the maximum number of dimensions, and the device information of each target device.

[0146] The computing performance of a computing device is limited, and the computing performance can be characterized by the maximum processing volume that the computing device can handle. Optionally, the maximum processing volume can be the maximum processing volume corresponding to the number of calculations, which can be represented by x. x can be a manually preset constant. For example, x can be 10,000.

[0147] The maximum number of dimensions that can be calculated by the computing device is limited by the maximum processing capacity. Optionally, if the maximum processing capacity of the computing device is represented by x, the maximum number of dimensions can be determined as For example, if x is 10,000, the maximum number of dimensions can be 13.

[0148] Optionally, the number of the plurality of feature dimensions may be represented by n. For example, if 20 feature dimensions are preset, then n is equal to 20.

[0149] The computing device can be based on the number of feature dimensions n, the maximum processing volume x, the maximum number of dimensions and device information of each target device, and determining a plurality of target feature dimensions in the plurality of feature dimensions.

[0150] Optionally, multiple target feature dimensions can be determined from multiple feature dimensions in the following manner: determine the combination depth based on the number of multiple feature dimensions and the maximum processing volume; determine whether to eliminate the multiple feature dimensions based on the combination depth, the number of multiple feature dimensions, and the maximum number of dimensions; if so, eliminate the multiple feature dimensions at least once based on the combination depth, the number of multiple feature dimensions, the maximum number of dimensions, and the device information of each target device to obtain multiple target feature dimensions; if not, determine the multiple feature dimensions as multiple target feature dimensions.

[0151] In order to prevent dimension explosion, a combination depth may be determined. The combination depth may be used to indicate the maximum number of dimensions that may be included in a dimension combination. For example, if the combination depth is 10, it may be determined that a dimension combination may include at most 10 dimensions.

[0152] Optionally, if the combination depth can be represented by d, where d is a positive integer, the combination depth d can be determined according to the number n of multiple feature dimensions and the maximum processing amount x, in combination with the following formula (1) and formula (2):

[0153]

[0154] Max(d) Formula (2)

[0155] The value range of m is 1, 2, ..., d. The constraint of formula (1) means to increase the calculation depth as much as possible under the condition of fixed resources, in order to prevent excessive information loss.

[0156] For example, if x is 10,000, then it can be determined that the combination depth d is 3.

[0157] After determining the combination depth, the computing device may calculate the combination depth d, the number of feature dimensions n, the maximum number of dimensions, and the Determine whether to perform elimination processing on the multiple feature dimensions. If elimination processing is performed, the multiple feature dimensions can be eliminated at least once according to the combination depth, the number of multiple feature dimensions, the maximum number of dimensions, and the device information of each target device to obtain multiple target feature dimensions.

[0158] For example, if the combination depth d is 3, the number of multiple feature dimensions is 20, and the maximum number of dimensions is 13, then it can be determined whether to eliminate the multiple feature dimensions based on the combination depth 3, the number of multiple feature dimensions 20, and the maximum number of dimensions 13. If eliminated, then the 20 feature dimensions can be eliminated at least once based on the combination depth 3, the number of multiple feature dimensions 20, the maximum number of dimensions 13, and the device information of each target device to obtain multiple target feature dimensions, assuming that 3 target feature dimensions can be obtained; if not, the 20 feature dimensions are determined as 20 target feature dimensions.

[0159] S203: Determine multiple dimension combinations corresponding to multiple target feature dimensions, and determine multiple target dimension combinations from the multiple dimension combinations according to device information of multiple target devices.

[0160] The dimension combination may include at least one target feature dimension. For example, dimension combination 1 may include one feature dimension, which is a processor model, and dimension combination 2 may include two feature dimensions, which are a processor model and a device model.

[0161] The target dimension combination is a combination of fault factors whose impact on the device failure is greater than or equal to a preset threshold. The preset threshold can be preset manually. For example, if the preset threshold is 0.8, if dimension combination 2 can include two feature dimensions, namely, processor model and device model, and if the target impact of dimension combination 2 is 0.9, then dimension combination 2 can be determined as the target dimension combination, that is, dimension combination 2 is a combination of fault factors that has a greater impact on the device failure.

[0162] For example, if the computing device is determined to have three target feature dimensions, namely, processor model, device model, and disk model, then seven dimension combinations corresponding to the three target feature dimensions can be determined, namely, dimension combination 1, dimension combination 2, dimension combination 3, dimension combination 4, dimension combination 5, dimension combination 6, and dimension combination 7, as shown in Table 2:

[0163] Table 2

[0164]

[0165]

[0166] Optionally, multiple target dimension combinations can be determined from multiple dimension combinations based on the device information of multiple target devices in the following manner: determine the target impact of each dimension combination on device failure based on the device information of multiple target devices; and determine the dimension combinations among the multiple dimension combinations whose target impact on device failure is greater than or equal to a preset threshold as the multiple target dimension combinations.

[0167] Optionally, the target impact of each dimension combination on the equipment failure can be determined through secondary iterative calculation, thereby eliminating the correlation between the characteristic dimensions.

[0168] For example, if there are 7 dimension combinations, the target impact of the 7 dimension combinations on equipment failure can be determined through 2 iterative calculations. It is assumed that the target impact of the 7 dimension combinations on equipment failure is determined as shown in Table 3:

[0169] Table 3

[0170] Dimension combination Target influence Dimension combination 1 0.83 Dimension combination 2 0.81 Dimension combination 3 0.92 Dimension combination 4 0.22 Dimension combination 5 0.91 Dimension combination 6 0.75 Dimension combination 7 0.86

[0171] If the preset threshold is 0.9, then the dimension combinations among the seven dimension combinations whose target impact on equipment failure is greater than or equal to the preset threshold, namely, dimension combination 3 and dimension combination 5, can be determined as target dimension combinations.

[0172] In an embodiment of the present application, a computing device can obtain device information of multiple target devices, and determine multiple target feature dimensions in multiple feature dimensions based on the computing performance of the computing device and the device information of each target device, and then determine multiple dimension combinations corresponding to the multiple target feature dimensions, and determine multiple target dimension combinations in multiple dimension combinations based on the device information of the multiple target devices. Through the technical solution of the present application, the combination depth can be adaptively adjusted to prevent the dimension explosion problem; multiple feature dimensions can be eliminated at least once to determine multiple target feature dimensions; and the target impact of multiple dimensional combinations on device failures can be calculated twice in an iterative manner, eliminating the correlation between feature dimensions, so as to determine the target dimension combination in multiple dimensional combinations, which improves the accuracy of determining the target dimension combination compared to manually determining multiple target dimension combinations.

[0173] Figure 3 A flowchart of another method for determining a device failure factor provided by an exemplary embodiment of the present application. Figure 3 , methods may include:

[0174] S301. Obtain device information of multiple target devices.

[0175] It should be noted that the execution process of step S301 can refer to the execution process of step S201, which will not be repeated here.

[0176] S302. Determine the maximum processing capacity corresponding to the computing performance of the computing device, and determine the maximum number of dimensions based on the maximum processing capacity.

[0177] Alternatively, if the maximum processing volume is x, the maximum number of dimensions can be For example, if x is 10,000, the maximum number of dimensions can be 13.

[0178] S303: Determine the combination depth according to the number of multiple feature dimensions and the maximum processing amount.

[0179] For example, if the number of feature dimensions is 20 and the maximum processing volume is 10,000, the combination depth d can be determined to be 3 by using formula (1) and formula (2).

[0180] S304: Determine whether the combination depth is equal to the number of multiple feature dimensions.

[0181] If the combination depth is equal to the number of multiple feature dimensions, it can be determined that the multiple feature dimensions are not to be eliminated, and S307 can be executed; if the combination depth is less than the number of multiple feature dimensions, S305 can be executed.

[0182] For example, if the combination depth is 3 and the number n of multiple feature dimensions is 3, it can be determined that the three feature dimensions are not to be eliminated, and S307 can be executed; if the combination depth is 3 and the number n of multiple feature dimensions is 20, it can be determined that the combination depth 3 is less than the number 20 of multiple feature dimensions, and S305 can be executed.

[0183] S305 . Determine whether half of the number of multiple feature dimensions is greater than or equal to the maximum number of dimensions.

[0184] If so, it can be determined that the multiple feature dimensions are to be eliminated, and S306 can be executed; if not, it is determined that the multiple feature dimensions are not to be eliminated, and S307 can be executed.

[0185] For example, if the number n of multiple feature dimensions is 20 and the maximum number of dimensions is 8, it can be determined that 20 / 2=10 is greater than the maximum number of dimensions 8, and it can be determined that the 20 feature dimensions are eliminated, and S306 can be executed; if the maximum number of dimensions is 13, it can be determined that 20 / 2=10 is less than the maximum number of dimensions 13, and it can be determined that the 20 feature dimensions are not eliminated, and S307 can be executed.

[0186] S306 : Determine a plurality of target feature dimensions from the plurality of feature dimensions according to the combination depth, the number of the plurality of feature dimensions, the maximum number of dimensions, and the device information of each target device.

[0187] In an optional embodiment, multiple target feature dimensions may be determined from multiple feature dimensions in the following manner: determine a first elimination number according to the number of multiple feature dimensions, and perform a first elimination process on the multiple feature dimensions according to the first elimination number, the combination depth, and the device information of each target device to obtain multiple first remaining dimensions; update the combination depth according to the number of multiple i-th remaining dimensions to obtain the i+1th updated combination depth, and determine whether to perform the i+1th elimination process according to the i+1th updated combination depth, the number of multiple i-th remaining dimensions, i+1, and the maximum number of dimensions; if so, determine the i+1th elimination number according to the number of i-th remaining dimensions and i+1, and perform an elimination process on the multiple i-th remaining dimensions according to the i+1th elimination number, the i+1th updated combination depth, and the device information of each target device to obtain multiple i+1th remaining dimensions; if not, determine the multiple i-th remaining dimensions as multiple target feature dimensions; wherein i takes integers starting from 1 in sequence until multiple target feature dimensions are obtained.

[0188] Optionally, if the number of eliminations in the i-th round can be calculated by T i The elimination quantity T can be determined by the following formula (3): i :

[0189] T i =n*2 -i Formula (3)

[0190] Among them, n represents the number of multiple feature dimensions, and i represents the i-th elimination process.

[0191] For example, if i is 1 and n is 20, then T1 = 20*2 -1 =10, then it can be determined that the first elimination number is 10.

[0192] For example, when i is equal to 1, if the number of multiple feature dimensions is 20, the first elimination number is 10, and if the combination depth d is 3, the 20 feature dimensions can be eliminated for the first time according to the first elimination number 10, the combination depth 3 and the device information of each target device to obtain 10 first remaining dimensions.

[0193] When i+1 is equal to 2, the computing device can update the combination depth d according to the number 10 of the 10 first remaining dimensions through formulas (1) and (2) to obtain the second updated combination depth. Assuming that the second updated combination depth is 4, it can be determined whether to perform the second elimination process according to the second updated combination depth, the number of multiple first remaining dimensions 10, 2, and the maximum number of dimensions. If so, the second elimination number can be determined according to the number of multiple first remaining dimensions 10 and 2, and the 10 first remaining dimensions can be eliminated according to the second elimination number, the second updated combination depth, and the device information of each target device to obtain multiple second remaining dimensions; if not, the 10 first remaining dimensions can be determined as multiple target feature dimensions. i can be 1, 2, ... in sequence until multiple target feature dimensions are obtained.

[0194] S307: Determine the multiple feature dimensions as multiple target feature dimensions.

[0195] If it is determined that the multiple feature dimensions are not to be eliminated, the multiple feature dimensions may be determined as multiple target feature dimensions.

[0196] For example, if the combination depth is 3 and the number n of multiple feature dimensions is 3, it can be determined that the three feature dimensions will not be eliminated, and the three feature dimensions can be determined as the three target feature dimensions; if the number n of multiple feature dimensions is 20 and the maximum number of dimensions is 13, since it can be determined that 20 / 2=10 is less than the maximum number of dimensions 13, it can be determined that the 20 feature dimensions will not be eliminated, and the 20 feature dimensions can be determined as 20 target feature dimensions.

[0197] S308: Determine multiple dimension combinations corresponding to multiple target feature dimensions.

[0198] For example, if the computing device is determined to have three target feature dimensions, namely, processor model, device model, and disk model, then seven dimension combinations corresponding to the three target feature dimensions can be determined, as shown in Table 2.

[0199] S309: Determine the target impact of each dimension combination on the device failure according to the device information of the multiple target devices.

[0200] Optionally, the computing device may perform a second iteration to calculate the target impact of each dimension combination based on the device information of multiple target devices. When calculating the impact for the first time, the computing device may perform the impact calculation of different dimensions with the same number of dimensions on each dimension combination based on the device information of multiple target devices to obtain the initial impact of each dimension combination on the device failure; when calculating the impact for the second time, the computing device may perform the impact calculation of different dimensions on each dimension combination based on the initial impact of each dimension combination on the device failure to obtain the target impact of each dimension combination on the device failure.

[0201] In an optional embodiment, for any dimension combination, the target impact of the dimension combination on the equipment failure can be determined according to the equipment information of multiple target devices in the following manner: determine that the dimension combination corresponds to multiple combination feature information; determine the impact of each combination feature information on the equipment failure; determine the statistical value of the impact of each combination feature information on the equipment failure as the initial impact of the dimension combination on the equipment failure; and determine the target impact of each dimension combination on the equipment failure based on the initial impacts of the multiple dimension combinations.

[0202] Optionally, for any dimension combination, the dimension combination may have corresponding multiple combination feature information. The combination feature information may include feature information corresponding to each feature dimension in the dimension combination.

[0203] For example, if dimension combination 4 includes processor model and device model, if there are 2 processor models, namely processor model 1 and processor model 2, and if there are 2 device models, namely device model 1 and device model 2, then it can be determined that the combination feature information 1 corresponding to dimension combination 4 includes processor model 1 and device model 1, combination feature information 2 can include processor model 1 and device model 2, combination feature information 3 can include processor model 2 and device model 1, and combination feature information 4 can include processor model 2 and device model 2.

[0204] After the computing device determines multiple combined feature information corresponding to the dimensional combination, for any one second combined feature information among the multiple combined feature information corresponding to the dimensional combination, the influence of the second combined feature information on the device failure can be determined in the following manner: based on the second combined feature information, at least one second target device is determined from multiple target devices; a fourth number of failed devices in the at least one second target device is determined, and a fifth number of the at least one second target device is determined; and a ratio of the fourth number to the fifth number is determined as the influence of the second combined feature information on the device failure.

[0205] The device information of the second target device may include the second combined feature information.

[0206] For example, if the second combined feature information is combined feature information 1, and combined feature information 1 includes processor model 1 and device model 1, then 10 second target devices can be determined from multiple target devices based on combined feature information 1, and the processor models of the 10 second target devices are all processor model 2, and the device models are all device model 2. The computing device can determine the fourth number of failed devices from the 10 second target devices. Assuming that the fourth number is 3 and the fifth number of the 10 second target devices is 10, 3 / 10=0.3 can be determined as the influence of the second combined feature information on the device failure; similarly, for combined feature information 2, combined feature information 3, and combined feature information 4 in the dimensional combination, the influence of each combined feature information on the device failure can be determined respectively as shown in Table 4:

[0207] Table 4

[0208] Dimension combination 4 (processor model, device model) The fourth quantity The fifth quantity Impact Combined feature information 1 (processor model 1, device model 1) 3 10 0.3 Combined feature information 2 (processor model 1, device model 2) 2 11 0.18 Combined feature information 3 (processor model 2, device model 1) 1 15 0.07 Combined feature information 4 (processor model 2, device model 2) 4 13 0.31

[0209] After the computing device determines the influence of each combination of feature information on the device failure, the statistical value of the influence of each combination of feature information on the device failure can be determined as the initial influence of the dimension combination on the device failure.

[0210] Optionally, the statistical value may be the sum, mean, mean square error, etc. of the influence of each combination of characteristic information on the equipment failure.

[0211] For example, if the statistical value is the mean square error, the mean square error can be calculated to be 0.098 based on the influence of the four combined feature information on the equipment failure in Table 3, and it can be determined that the initial influence of the dimension combination 4 on the equipment failure is 0.098.

[0212] Since there are multiple dimensional combinations, the computing device can obtain the initial impact of the multiple dimensional combinations on the device failure in the above manner.

[0213] For example, if there are 7 dimensional combinations, as shown in Table 2, the computing device can determine the initial impact of the 7 dimensional combinations on the device failure. Assume that the initial impact of the 7 dimensional combinations on the device failure is as shown in Table 5:

[0214] Table 5

[0215] Dimension combination Initial Impact Dimension combination 1 0.083 Dimension combination 2 0.081 Dimension combination 3 0.092 Dimension combination 4 0.098 Dimension combination 5 0.124 Dimension combination 6 0.275 Dimension combination 7 0.186

[0216] After the computing device determines the initial impact of each dimensional combination on the device failure, it can perform a global calculation of the impact of different dimensional combinations on each dimensional combination based on the initial impact of each dimensional combination on the device failure to obtain the target impact of each dimensional combination on the device failure.

[0217] Optionally, the following method can be used to determine the target influence of the first dimensional combination on equipment failure for any first dimensional combination among multiple dimensional combinations; based on the initial influences of the multiple dimensional combinations: among the multiple dimensional combinations, determine multiple second dimensional combinations corresponding to the first dimensional combination; for any second dimensional combination, determine the ratio of the sixth number of feature dimensions in the first dimensional combination to the seventh number of feature dimensions in the second dimensional combination as the feature ratio of the first dimensional combination in the second dimensional combination; determine the product of the initial influence of the second dimensional combination on equipment failure and the feature ratio as the intermediate influence of the first dimensional combination in the second dimensional combination; and determine the statistical value of the intermediate influence of the first dimensional combination in each second dimensional combination as the target influence of the first dimensional combination on equipment failure.

[0218] The multiple feature dimensions included in the second dimension combination include the multiple feature dimensions in the first dimension combination. For example, if the first dimension combination includes the processor model, the second dimension combination may include the processor model and the disk model.

[0219] For example, if there are 7 dimension combinations, as shown in Table 2, if the first dimension combination is dimension combination 4, then the 2 second dimension combinations corresponding to dimension combination 4 can be determined, which are dimension combination 4 and dimension combination 7. If the second dimension combination is dimension combination 4, since the first dimension combination and the second dimension combination are the same, both are dimension combination 4, the initial impact of dimension combination 4 on equipment failure, 0.098, can be directly determined as the intermediate impact 1; if the second dimension combination is dimension combination 7, then the sixth number of characteristic dimensions in the first dimension combination, i.e., dimension combination 4, can be determined to be 2, and the number of characteristic dimensions in the second dimension combination, i.e., dimension combination 7, can be determined to be 3, and then the characteristic ratio of dimension combination 4 in dimension combination 7 can be determined to be 2 / 3. Assuming that the initial impact of dimension combination 7 on equipment failure is 0.186, 0.186*2 / 3=0.124 can be determined as the intermediate impact 2 of dimension combination 4. If the statistical value is the sum, the computing device can determine the sum of the intermediate impact 1 and the intermediate impact 2, that is, 0.098+0.124=0.22, as the target impact of the dimension combination 4 on the device failure. The computing device can determine the target impact of the 7 dimension combinations on the device failure respectively, assuming that the target impact of the 7 dimension combinations on the device failure can be shown in Table 3.

[0220] By calculating the influence twice, the initial influence and target influence of multiple dimension combinations can be obtained, which can eliminate the coupling between feature dimensions.

[0221] S310: Determine, among the multiple dimension combinations, dimension combinations whose target impact on device failure is greater than or equal to a preset threshold as multiple target dimension combinations.

[0222] For example, if there are 7 dimension combinations, the target impact of the 7 dimension combinations on equipment failure can be determined through 2 iterative calculations as shown in Table 3. If the preset threshold is 0.9, the dimension combinations among the 7 dimension combinations whose impact on equipment failure is greater than or equal to the preset threshold, namely dimension combination 3 and dimension combination 5, can be determined as target dimension combinations.

[0223] Optionally, after determining the target dimension combination, the computing device may also monitor multiple target devices according to the target dimension combination, obtain characteristic information of the multiple target devices under the target dimension combination, and perform timing change intensity detection, timing change anomaly detection, and timing change consistency detection on the multiple target devices to determine the changes in device failures under the target dimension combination.

[0224] In an embodiment of the present application, a computing device may obtain device information of multiple target devices, determine the maximum processing volume corresponding to the computing performance of the computing device, and determine the maximum number of dimensions according to the maximum processing volume, and then determine the combination depth according to the number of multiple feature dimensions and the maximum processing volume. The computing device may determine whether the combination depth is equal to the number of multiple feature dimensions. If so, it may further determine whether one-half of the number of multiple feature dimensions is greater than or equal to the maximum number of dimensions. If so, the multiple feature dimensions may be eliminated at least once according to the combination depth, the number of multiple feature dimensions, the maximum number of dimensions, and the device information of each target device to obtain multiple target feature dimensions; if not, the multiple feature dimensions may be determined as multiple target feature dimensions. The computing device may determine multiple dimension combinations corresponding to the multiple target feature dimensions, and determine the target impact of each dimension combination on the device failure according to the device information of the multiple target devices. The computing device may determine the dimension combination whose target impact on the device failure is greater than or equal to the preset threshold value among the multiple dimension combinations as the multiple target dimension combinations. Through the technical solution of the present application, the combination depth can be adaptively adjusted to prevent the problem of dimensional explosion; multiple target feature dimensions can be iteratively determined in multiple feature dimensions; and the target impact of multiple dimensional combinations on equipment failures can be calculated by secondary iteration, eliminating the correlation between feature dimensions, so as to determine the target dimension combination in the multiple dimensional combinations. Compared with manually determining multiple target dimension combinations, the accuracy of determining the target dimension combination is improved.

[0225] Next, combine Figure 4 ,right Figure 3 In the embodiment, step S306, the process of determining multiple target feature dimensions in multiple feature dimensions is described in detail.

[0226] Figure 4 A flowchart of determining multiple target feature dimensions in multiple feature dimensions provided by an exemplary embodiment of the present application. Figure 4 , the method may include:

[0227] S401. Initialize i to 1.

[0228] Optionally, i is initialized to 1, indicating the first elimination process. i can be integers starting from 1.

[0229] S402: Determine the number of first eliminations according to the number of multiple feature dimensions.

[0230] Optionally, if the number of multiple feature dimensions is represented by n, and i is 1, the first elimination number can be determined by formula (3): T i =n*2 -i .

[0231] For example, if i is 1 and n is 20, then T1=20*2-1=10, that is, the number of eliminations in the first time is 10.

[0232] S403: Determine the influence of each feature dimension on the device failure according to the combination depth and the device information of each target device.

[0233] In an optional embodiment, for any feature dimension, the influence of the feature dimension on the equipment failure can be determined in the following manner: based on the equipment information of each target device, multiple feature sets corresponding to the feature dimension are determined; for each feature set, the influence of each combination of feature information in the feature set on the equipment failure is determined, and the statistical value of the influence of each combination of feature information on the equipment failure is determined as the influence of the feature set on the equipment failure; based on the influence of each feature set on the equipment failure, the influence of the feature dimension on the equipment failure is determined.

[0234] Each feature set may include a plurality of combined feature information. The combined feature information may include feature information corresponding to a feature dimension, or the combined feature information may include feature information corresponding to a feature dimension and feature information corresponding to other feature dimensions combined with the feature dimension.

[0235] For example, if there are 20 feature dimensions, namely feature dimension 1, feature dimension 2, ..., feature dimension 20, where feature dimension 1 is the processor model. For the processor model, multiple feature sets corresponding to the processor model can be determined. Assuming that there are 96 feature sets corresponding to the processor model, and the combined feature information respectively included in the multiple feature sets, can be shown in Table 6:

[0236] Table 6

[0237]

[0238] For example, for feature set 2, feature set 2 may include 4 combined feature information, and the 4 combined feature information may include feature dimension, that is, feature information corresponding to the processor model.

[0239] For each feature set, the computing device may determine the influence of each combination of feature information in the feature set on the device failure.

[0240] Optionally, for any first combination feature information in the feature set, the influence of the first combination feature information on the device failure can be determined in the following manner: based on the first combination feature information, at least one first target device is determined from multiple target devices; a first number of failed devices in the at least one first target device is determined, and a second number of the at least one first target device is determined; and the ratio of the first number to the second number is determined as the influence of the first combination feature information on the device failure.

[0241] The device information of the first target device may include the first combined feature information.

[0242] For example, if there are 500 target devices, if feature set 1 corresponds to combined feature information 1 and combined feature information 2, if the first combined feature information is combined feature information 1, i.e., processor model 1, then 300 first target devices can be determined from the 500 target devices based on processor model 1, and the processor model of the 300 first target devices is processor model 1. Assuming that the computing device can determine that the first number of failed devices is 50 among the 300 first target devices, and the second number of the plurality of first target devices is 300, then 50 / 300=0.17 can be determined as the influence of combined feature information 1 on device failure; similarly, assuming that the influence of combined feature information 2 on device failure can be determined to be 0.21.

[0243] Optionally, for any feature set, after determining the influence of each combination of feature information in the feature set on the equipment failure, the statistical value of the influence of each combination of feature information on the equipment failure can be determined as the influence of the feature set on the equipment failure.

[0244] Optionally, the statistical value may be a mean, a mean square error, a sum, a weighted sum, or the like.

[0245] For example, if feature set 1 corresponds to combined feature information 1 and combined feature information 2, the influence of combined feature information 1 on equipment failure is 0.17, and the influence of combined feature information 2 on equipment failure is 0.21. If the statistical value is the mean square error, then based on 0.17 and 0.21, it can be determined that the influence of feature set 1 on equipment failure is 0.063.

[0246] After the computing device determines the influence of each feature set on the device failure, the influence of the feature dimension on the device failure can be determined according to the influence of each feature set on the device failure in the following manner: for any feature set, determine the third number of feature dimensions in the feature set; determine the initial influence of the feature dimension in the feature set by the ratio of the influence of the feature set on the device failure to the third number; determine the statistical value of the initial influence of the feature dimension in each feature set as the influence of the feature dimension on the device failure.

[0247] Optionally, the statistical value may be a sum, a mean, a mean square error, a weighted sum, or the like.

[0248] For example, if there are 96 feature sets corresponding to the processor model, as shown in Table 6, then for feature set 1 in Table 6, if the influence of feature set 1 on device failure is 0.063, then since the third number of feature dimensions in feature set 1 is 1, 0.063 / 1 can be determined as the initial influence of the processor model in feature set 1 is 0.063; for feature set 2, if the influence of feature set 2 on device failure is 0.098, since the third number of feature dimensions in feature set 2 is 2, 0.098 / 2=0.049 can be determined as the initial influence of the processor model in feature set 2; ...; the computing device can determine the initial influence of the processor model in the corresponding 96 feature sets and obtain 96 initial influences. Assuming that the statistical value is the sum, the sum of the 96 initial influences corresponding to the processor model can be determined as the influence of the processor model on the device failure. Assume that the influence of the processor model on the device failure is 0.83.

[0249] For example, if there are 20 characteristic dimensions, the computing device may determine the influence of the 20 characteristic dimensions on the device failure respectively in the above manner.

[0250] S404: Sort the multiple feature dimensions in order of influence from small to large.

[0251] For example, if there are 20 feature dimensions, namely feature dimension 1, feature dimension 2, ..., feature dimension 20, where feature dimension 1 is the processor model, the computing device can sort the influence of the 20 feature dimensions on the device failure in order from small to large, and the sorting results are shown in Table 7:

[0252] Table 7

[0253] Feature Dimension Impact Feature Dimension 18 0.001 Feature Dimension 20 0.003 Feature Dimension 19 0.004 Feature Dimension 16 0.006 Feature Dimension 17 0.007 Feature Dimension 9 0.008 Feature Dimension 12 0.008 Feature Dimension 8 0.009 Feature Dimension 15 0.009 Feature Dimension 11 0.012 …… …… Feature Dimension 3 0.76 Feature Dimension 1 0.83

[0254] S405 . Delete the feature dimensions that are eliminated the first time from the multiple feature dimensions to obtain multiple first remaining dimensions.

[0255] For example, if the first elimination number is 10, and if there are 20 feature dimensions whose impact on equipment failure is ranked as shown in Table 7, then feature dimension 18, feature dimension 20, feature dimension 19, feature dimension 16, feature dimension 17, feature dimension 9, feature dimension 12, feature dimension 8, feature dimension 15 and feature dimension 11 can be deleted from the 20 feature dimensions to obtain 10 remaining dimensions.

[0256] S406 . Update the combination depth according to the number of the plurality of i-th remaining dimensions to obtain an (i+1)-th updated combination depth.

[0257] Optionally, the second updated combination depth may be determined by formula (1) and formula (2).

[0258] For example, if i is 1 and the number of the first remaining dimensions is 10, the second combination depth can be determined to be 4 by using formula (1) and formula (2).

[0259] S407 , determining whether to perform the i+1th elimination process according to the i+1th updated combination depth, the number of multiple i-th remaining dimensions, i+1, and the maximum number of dimensions.

[0260] If the i+1th update combination depth is equal to the number of multiple i-th remaining dimensions, it is determined that the multiple i-th remaining dimensions will not be eliminated; if the i+1th update combination depth is less than the number of multiple i-th remaining dimensions, it is determined whether half of the number of multiple i-th remaining dimensions is greater than or equal to the maximum number of dimensions; if so, it is determined that the multiple i-th remaining dimensions will be eliminated, and step S408 can be executed; if not, it is determined that the multiple i-th remaining dimensions will not be eliminated, and step S409 can be executed.

[0261] For example, if i+1 is 2, the second update combination depth, the number of the 10 first remaining dimensions, 2, and the maximum number of dimensions is 13, it is determined whether to perform the second elimination process. Since the second update combination depth is 4, which is less than the number of the plurality of first remaining dimensions, 10, it can be further determined that 10 / 2 is less than the maximum number of dimensions, 13, and it is determined that the 10 first remaining dimensions are not to be eliminated, and step S408 can be executed.

[0262] S408. Determine the i+1th elimination number according to the number of the i-th remaining dimensions and i+1, and eliminate multiple i-th remaining dimensions according to the i+1th elimination number, the i+1th updated combination depth and the device information of each target device to obtain multiple i+1th remaining dimensions.

[0263] For example, the second elimination number can be determined based on the number of the first remaining dimensions being 10 and i+1 being 2 using formula (3), and the 10 first remaining dimensions can be eliminated based on the second elimination number, the second update combination depth, and the device information of each target device to obtain multiple second remaining dimensions.

[0264] S409: Determine multiple i-th remaining dimensions as multiple target feature dimensions.

[0265] If it is determined that the multiple i-th remaining dimensions are not to be eliminated, the multiple i-th remaining dimensions may be determined as multiple target feature dimensions.

[0266] For example, if there are 10 first remaining dimensions, the 10 first remaining dimensions may be determined as 10 target feature dimensions.

[0267] S410, update i to i+1.

[0268] When i is equal to 1 and i+1 is equal to 2, after executing the second elimination process, multiple second remaining dimensions can be obtained. The computing device can update i to i+1, that is, i can be updated to 1, i to 2, i+1 to 3, and then the third updated combination depth can be determined, and then it can be determined whether to perform the third elimination process.

[0269] By eliminating multiple feature dimensions and determining multiple target feature dimensions, we can effectively prevent the problem of too shallow calculation depth and information loss caused by too many feature dimensions.

[0270] In the embodiment of the present application, when i is 1, the computing device can determine the first elimination number according to the number of multiple feature dimensions, and determine the influence of each feature dimension on the device failure according to the combination depth and the device information of each target device, and then sort the multiple feature dimensions in order of influence from small to large, and delete the first elimination number of feature dimensions in the multiple feature dimensions to obtain multiple first remaining dimensions. The computing device can update the combination depth according to the number of multiple i-th remaining dimensions to obtain the i+1th updated combination depth, and determine whether to perform the i+1th elimination process according to the i+1th updated combination depth, the number of multiple i-th remaining dimensions, i+1, and the maximum number of dimensions. If yes, the i+1th elimination number can be determined according to the number of i-th remaining dimensions and i+1, and the multiple i-th remaining dimensions can be eliminated according to the i+1th elimination number, the i+1th updated combination depth and the device information of each target device to obtain multiple i+1th remaining dimensions; if not, the multiple i-th remaining dimensions can be determined as multiple target feature dimensions. Through the technical solution of the present application, the combination depth can be adaptively adjusted to prevent the dimensionality explosion problem; multiple target feature dimensions can be iteratively determined multiple times in multiple feature dimensions, thereby improving the accuracy of determining the target feature dimensions.

[0271] Figure 5 A schematic diagram of a device for determining a device failure factor provided by an exemplary embodiment of the present application. Applicable to computing devices, see Figure 5 The device 10 for determining equipment failure factors includes: an acquisition module 11, a first determination module 12, and a second determination module 13, wherein:

[0272] The acquisition module 11 is used to acquire device information of multiple target devices, where the device information includes feature information and historical fault information of the target devices in multiple feature dimensions;

[0273] The first determining module 12 is used to determine a plurality of target feature dimensions from the plurality of feature dimensions according to the computing performance of the computing device and the device information of each target device;

[0274] The second determination module 13 is used to determine multiple dimension combinations corresponding to the multiple target feature dimensions, and determine multiple target dimension combinations in the multiple dimension combinations according to the device information of the multiple target devices, wherein the dimension combinations include at least one of the target feature dimensions, and the target dimension combinations are combinations of fault factors whose impact on the equipment failure is greater than or equal to a preset threshold.

[0275] In a possible implementation manner, the first determining module 12 is specifically configured to:

[0276] Determine a maximum processing capacity corresponding to the computing performance of the computing device, and determine a maximum number of dimensions based on the maximum processing capacity;

[0277] The plurality of target feature dimensions are determined from the plurality of feature dimensions according to the number of the plurality of feature dimensions, the maximum processing amount, the maximum number of dimensions, and device information of each target device.

[0278] The device for determining equipment failure factors provided in the embodiment of the present application can execute the technical solution shown in the above method embodiment, and its implementation principle and beneficial effects are similar, which will not be repeated here.

[0279] In a possible implementation manner, the first determining module 12 is specifically configured to:

[0280] Determine a combination depth according to the number of the plurality of feature dimensions and the maximum processing amount, wherein the combination depth is used to indicate the number of dimensions included in the dimension combination;

[0281] Determining whether to eliminate the multiple feature dimensions according to the combination depth, the number of the multiple feature dimensions, and the maximum number of dimensions;

[0282] If yes, performing at least one elimination process on the multiple feature dimensions according to the combination depth, the number of the multiple feature dimensions, the maximum number of dimensions, and the device information of each target device to obtain the multiple target feature dimensions;

[0283] If not, the multiple feature dimensions are determined as the multiple target feature dimensions.

[0284] In a possible implementation manner, the first determining module 12 is specifically configured to:

[0285] Determine a first elimination number according to the number of the multiple feature dimensions, and perform a first elimination process on the multiple feature dimensions according to the first elimination number, the combination depth, and the device information of each target device to obtain multiple first remaining dimensions;

[0286] According to the number of the multiple i-th remaining dimensions, the combination depth is updated to obtain the i+1th updated combination depth, and according to the i+1th updated combination depth, the number of the multiple i-th remaining dimensions, i+1, and the maximum number of dimensions, it is determined whether to perform the i+1th elimination process, and if so, the i+1th elimination number is determined according to the number of the i-th remaining dimensions and i+1, and according to the i+1th elimination number, the i+1th updated combination depth and the device information of each target device, the multiple i-th remaining dimensions are eliminated to obtain multiple i+1th remaining dimensions; if not, the multiple i-th remaining dimensions are determined as the multiple target feature dimensions;

[0287] Here, the i is an integer starting from 1 in sequence until the multiple target feature dimensions are obtained.

[0288] In a possible implementation manner, the first determining module 12 is specifically configured to:

[0289] Determining the influence of each feature dimension on the device failure according to the combination depth and the device information of each target device;

[0290] Sorting the multiple feature dimensions in order of influence from small to large;

[0291] Among the multiple feature dimensions, the first elimination number of feature dimensions is deleted to obtain the multiple first remaining dimensions.

[0292] In a possible implementation manner, for any feature dimension, the first determining module 12 is specifically configured to:

[0293] Determine, according to the device information of each target device, a plurality of feature sets corresponding to the feature dimension, each feature set respectively including a plurality of combined feature information, the combined feature information including the feature information corresponding to the feature dimension, or the combined feature information including the feature information corresponding to the feature dimension and feature information corresponding to other feature dimensions combined with the feature dimension;

[0294] For each feature set, determining the influence of each combination of feature information in the feature set on the equipment failure, and determining the statistical value of the influence of each combination of feature information on the equipment failure as the influence of the feature set on the equipment failure;

[0295] According to the influence of each feature set on the device failure, the influence of the feature dimension on the device failure is determined.

[0296] In a possible implementation manner, the first determining module 12 is specifically configured to:

[0297] Determine at least one first target device from the multiple target devices according to the first combined feature information, wherein device information of the first target device includes the first combined feature information;

[0298] Determining a first number of failed devices in the at least one first target device, and determining a second number of the at least one first target device;

[0299] The ratio of the first quantity to the second quantity is determined as the influence of the first combination feature information on the equipment failure.

[0300] In a possible implementation manner, the first determining module 12 is specifically configured to:

[0301] For any feature set, determining a third number of feature dimensions in the feature set;

[0302] Determine the initial influence of the feature dimension in the feature set by taking the ratio of the influence of the feature set on the device failure and the third quantity;

[0303] The statistical value of the initial influence of the feature dimension in each feature set is determined as the influence of the feature dimension on the device failure.

[0304] In a possible implementation manner, the first determining module 12 is specifically configured to:

[0305] If the combination depth is equal to the number of the plurality of feature dimensions, determining not to eliminate the plurality of feature dimensions;

[0306] If the combination depth is less than the number of the plurality of feature dimensions, determining whether one half of the number of the plurality of feature dimensions is greater than or equal to the maximum number of dimensions;

[0307] If yes, determine to eliminate the multiple feature dimensions;

[0308] If not, it is determined that the multiple feature dimensions will not be eliminated.

[0309] In a possible implementation manner, the second determining module 13 is specifically configured to:

[0310] Determining, based on the device information of the multiple target devices, a target impact degree of each dimension combination on the device failure;

[0311] Among the multiple dimension combinations, the dimension combinations whose target impact on the equipment failure is greater than or equal to the preset threshold are determined as the multiple target dimension combinations.

[0312] In a possible implementation manner, for any combination of dimensions, the second determining module 13 is specifically configured to:

[0313] Determine that the dimension combination corresponds to a plurality of combination feature information, wherein the combination feature information includes feature information corresponding to each feature dimension in the dimension combination;

[0314] Determine the impact of each combination of characteristic information on equipment failure;

[0315] Determine the statistical value of the influence of each combination of characteristic information on the equipment failure as the initial influence of the dimension combination on the equipment failure;

[0316] According to the initial impacts of the multiple dimensional combinations, a target impact of each dimensional combination on the device failure is determined.

[0317] In a possible implementation manner, for any second piece of combined feature information among the plurality of combined feature information, the second determining module 13 is specifically configured to:

[0318] determining at least one second target device from the plurality of target devices according to the second combined feature information, wherein the device information of the second target device includes the second combined feature information;

[0319] determining a fourth number of failed devices in the at least one second target device, determining a fifth number of the at least one second target device;

[0320] The ratio of the fourth number to the fifth number is determined as the influence of the second combination feature information on the equipment failure.

[0321] In a possible implementation manner, for any first dimension combination among the multiple dimension combinations, the second determining module 13 is specifically configured to:

[0322] Determine, among the multiple dimension combinations, multiple second dimension combinations corresponding to the first dimension combination, wherein the multiple feature dimensions included in the second dimension combination include the multiple feature dimensions in the first dimension combination;

[0323] For any second dimensional combination, a ratio of a sixth number of characteristic dimensions in the first dimensional combination to a seventh number of characteristic dimensions in the second dimensional combination is determined as a characteristic ratio of the first dimensional combination in the second dimensional combination;

[0324] Determine the product of the initial impact of the second dimensional combination on the equipment failure and the characteristic ratio as the intermediate impact of the first dimensional combination in the second dimensional combination;

[0325] The statistical value of the intermediate influence of the first dimensional combination in each second dimensional combination is determined as the target influence of the first dimensional combination on the equipment failure.

[0326] The device for determining equipment failure factors provided in the embodiment of the present application can execute the technical solution shown in the above method embodiment, and its implementation principle and beneficial effects are similar, which will not be repeated here.

[0327] The exemplary embodiment of the present application provides a schematic diagram of the structure of a computing device, see Figure 6 The computing device 20 may include a processor 21 and a memory 22. Exemplarily, the processor 21 and the memory 22 are interconnected via a bus 23.

[0328] The memory 22 stores computer-executable instructions;

[0329] The processor 21 executes the computer-executable instructions stored in the memory 22, so that the processor 21 executes the method shown in the above method embodiment.

[0330] Accordingly, an embodiment of the present application provides a computer-readable storage medium, in which computer-executable instructions are stored. When the computer-executable instructions are executed by a processor, they are used to implement the method described in the above method embodiment.

[0331] Accordingly, an embodiment of the present application may also provide a computer program product, including a computer program, which, when executed by a processor, can implement the method shown in the above method embodiment.

[0332] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0333] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0334] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0335] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0336] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0337] The memory may include non-permanent storage in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.

[0338] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.

[0339] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0340] The above is only an embodiment of the present application and is not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included in the scope of the claims of the present application.

Claims

1. A method for determining equipment failure factors, characterized in that: Applied to a computing device, the method comprises: Acquire device information of multiple target devices, wherein the device information includes feature information and historical fault information of the target devices in multiple feature dimensions; Determining a plurality of target feature dimensions from the plurality of feature dimensions according to the computing performance of the computing device and the device information of each target device; Determine multiple dimensional combinations corresponding to the multiple target feature dimensions, and determine multiple target dimension combinations in the multiple dimensional combinations based on the device information of the multiple target devices, the dimension combinations including at least one of the target feature dimensions, and the target dimension combinations are combinations of fault factors whose impact on device failure is greater than or equal to a preset threshold.

2. The method according to claim 1, characterized in that Determining multiple target feature dimensions from the multiple feature dimensions according to the computing performance of the computing device and the device information of each target device, including: Determine a maximum processing capacity corresponding to the computing performance of the computing device, and determine a maximum number of dimensions based on the maximum processing capacity; The plurality of target feature dimensions are determined from the plurality of feature dimensions according to the number of the plurality of feature dimensions, the maximum processing amount, the maximum number of dimensions, and device information of each target device.

3. The method according to claim 2, characterized in that Determining the plurality of target feature dimensions from the plurality of feature dimensions according to the number of the plurality of feature dimensions, the maximum processing amount, the maximum number of dimensions, and device information of each target device includes: Determine a combination depth according to the number of the plurality of feature dimensions and the maximum processing amount, wherein the combination depth is used to indicate the number of dimensions included in the dimension combination; Determining whether to eliminate the multiple feature dimensions according to the combination depth, the number of the multiple feature dimensions, and the maximum number of dimensions; If yes, performing at least one elimination process on the multiple feature dimensions according to the combination depth, the number of the multiple feature dimensions, the maximum number of dimensions, and the device information of each target device to obtain the multiple target feature dimensions; If not, the multiple feature dimensions are determined as the multiple target feature dimensions.

4. The method according to claim 3, characterized in that According to the combination depth, the number of the multiple feature dimensions, the maximum number of dimensions, and the device information of each target device, the multiple feature dimensions are eliminated at least once to obtain the multiple target feature dimensions, including: Determine a first elimination number according to the number of the multiple feature dimensions, and perform a first elimination process on the multiple feature dimensions according to the first elimination number, the combination depth, and the device information of each target device to obtain multiple first remaining dimensions; According to the number of the multiple i-th remaining dimensions, the combination depth is updated to obtain the i+1th updated combination depth, and according to the i+1th updated combination depth, the number of the multiple i-th remaining dimensions, i+1, and the maximum number of dimensions, it is determined whether to perform the i+1th elimination process, and if so, the i+1th elimination number is determined according to the number of the i-th remaining dimensions and i+1, and according to the i+1th elimination number, the i+1th updated combination depth and the device information of each target device, the multiple i-th remaining dimensions are eliminated to obtain multiple i+1th remaining dimensions; if not, the multiple i-th remaining dimensions are determined as the multiple target feature dimensions; Here, the i is an integer starting from 1 in sequence until the multiple target feature dimensions are obtained.

5. The method according to claim 4, characterized in that According to the first elimination quantity, the combination depth and the device information of each target device, the plurality of feature dimensions are first eliminated to obtain a plurality of first remaining dimensions, including: Determining the influence of each feature dimension on the device failure according to the combination depth and the device information of each target device; Sorting the multiple feature dimensions in order of influence from small to large; Among the multiple feature dimensions, the first elimination number of feature dimensions is deleted to obtain the multiple first remaining dimensions.

6. The method according to claim 5, characterized in that For any feature dimension; Determining the influence of the feature dimension on the device failure according to the combination depth and the device information of each target device includes: Determine, according to the device information of each target device, a plurality of feature sets corresponding to the feature dimension, each feature set respectively including a plurality of combined feature information, the combined feature information including the feature information corresponding to the feature dimension, or the combined feature information including the feature information corresponding to the feature dimension and feature information corresponding to other feature dimensions combined with the feature dimension; For each feature set, determining the influence of each combination of feature information in the feature set on the equipment failure, and determining the statistical value of the influence of each combination of feature information on the equipment failure as the influence of the feature set on the equipment failure; According to the influence of each feature set on the device failure, the influence of the feature dimension on the device failure is determined.

7. The method according to claim 6, characterized in that For any first combination of feature information in the feature set; determining the influence of the first combination of feature information on the device failure, including: Determine at least one first target device from the multiple target devices according to the first combined feature information, wherein device information of the first target device includes the first combined feature information; Determining a first number of failed devices in the at least one first target device, and determining a second number of the at least one first target device; The ratio of the first quantity to the second quantity is determined as the influence of the first combination feature information on the equipment failure.

8. The method according to claim 6, characterized in that Determining the influence of the feature dimension on the device failure according to the influence of each feature set on the device failure includes: For any feature set, determining a third number of feature dimensions in the feature set; Determine the initial influence of the feature dimension in the feature set by taking the ratio of the influence of the feature set on the device failure and the third quantity; The statistical value of the initial influence of the feature dimension in each feature set is determined as the influence of the feature dimension on the device failure.

9. The method according to any one of claims 3 to 8, characterized in that: According to the combination depth, the number of the multiple feature dimensions, and the maximum number of dimensions, determining whether to eliminate the multiple feature dimensions includes: If the combination depth is equal to the number of the plurality of feature dimensions, determining not to eliminate the plurality of feature dimensions; If the combination depth is less than the number of the plurality of feature dimensions, determining whether one half of the number of the plurality of feature dimensions is greater than or equal to the maximum number of dimensions; If yes, determine to eliminate the multiple feature dimensions; If not, it is determined that the multiple feature dimensions will not be eliminated.

10. The method according to any one of claims 1 to 9, characterized in that: Determining a plurality of target dimension combinations from the plurality of dimension combinations according to the device information of the plurality of target devices includes: Determining, based on the device information of the multiple target devices, a target impact degree of each dimension combination on the device failure; Among the multiple dimension combinations, the dimension combinations whose target impact on the equipment failure is greater than or equal to the preset threshold are determined as the multiple target dimension combinations.

11. The method according to claim 10, characterized in that For any dimension combination; according to the device information of the multiple target devices, determining the target impact of the dimension combination on the device failure includes: Determine that the dimension combination corresponds to a plurality of combination feature information, wherein the combination feature information includes feature information corresponding to each feature dimension in the dimension combination; Determine the impact of each combination of characteristic information on equipment failure; Determine the statistical value of the influence of each combination of characteristic information on the equipment failure as the initial influence of the dimension combination on the equipment failure; According to the initial impacts of the multiple dimensional combinations, a target impact of each dimensional combination on the device failure is determined.

12. The method according to claim 11, characterized in that For any second combination of feature information among the plurality of combination feature information, determining the influence of the second combination of feature information on the device failure includes: determining at least one second target device from the plurality of target devices according to the second combined feature information, wherein the device information of the second target device includes the second combined feature information; determining a fourth number of failed devices in the at least one second target device, determining a fifth number of the at least one second target device; The ratio of the fourth number to the fifth number is determined as the influence of the second combination feature information on the equipment failure.

13. The method according to claim 11, characterized in that For any first dimension combination among the multiple dimension combinations; determining a target impact degree of the first dimension combination on the device failure according to the initial impact degrees of the multiple dimension combinations, including: Determine, among the multiple dimension combinations, multiple second dimension combinations corresponding to the first dimension combination, wherein the multiple feature dimensions included in the second dimension combination include the multiple feature dimensions in the first dimension combination; For any second dimensional combination, a ratio of a sixth number of characteristic dimensions in the first dimensional combination to a seventh number of characteristic dimensions in the second dimensional combination is determined as a characteristic ratio of the first dimensional combination in the second dimensional combination; Determine the product of the initial impact of the second dimensional combination on the equipment failure and the characteristic ratio as the intermediate impact of the first dimensional combination in the second dimensional combination; The statistical value of the intermediate influence of the first dimensional combination in each second dimensional combination is determined as the target influence of the first dimensional combination on the equipment failure.

14. A device for determining equipment failure factors, characterized in that: Applied to a computing device, the apparatus comprises: an acquisition module, a first determination module and a second determination module, wherein: The acquisition module is used to acquire device information of multiple target devices, wherein the device information includes feature information and historical fault information of the target devices in multiple feature dimensions; The first determining module is used to determine a plurality of target feature dimensions from the plurality of feature dimensions according to the computing performance of the computing device and the device information of each target device; The second determination module is used to determine multiple dimension combinations corresponding to the multiple target feature dimensions, and determine multiple target dimension combinations in the multiple dimension combinations according to the device information of the multiple target devices, wherein the dimension combinations include at least one of the target feature dimensions, and the target dimension combinations are combinations of fault factors whose impact on the equipment failure is greater than or equal to a preset threshold.

15. A computing device, characterized in that: include: at least one processor; as well as a memory communicatively coupled to the at least one processor; The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the computing device to perform the method according to any one of claims 1 to 13.

16. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-executable instructions, and when the processor executes the computer-executable instructions, the method according to any one of claims 1 to 13 is implemented.

17. A computer program product comprising a computer program, characterized in that When the computer program is executed by a processor, the method according to any one of claims 1 to 13 is implemented.

Images