Equipment hidden fault capturing method and device, equipment, storage medium and program product

Abstract

The invention discloses an equipment hidden fault capturing method and device, equipment, a storage medium and a program product. The method comprises the following steps: acquiring sensor acquisition data acquired by sensors of different function types under at least one same timestamp in a preset time period; generating at least one to-be-evaluated data combination according to the sensor collection data under the same timestamps; the to-be-evaluated data combination comprises sensor acquisition data of sensors of different function types; inputting data collected by each sensor in each to-be-evaluated data combination into a hidden fault prediction model obtained by pre-training to obtain hidden fault probability values corresponding to each to-be-evaluated data combination output by the model; and according to the hidden fault probability value of each to-be-evaluated data combination, carrying out hidden fault capture on the equipment. According to the method, different types of sensor data are acquired and different to-be-evaluated data combinations are generated, so that the efficiency and accuracy of capturing hidden faults are improved.

Description

Technical Field

[0001] The present invention relates to the field of fault detection technology, and in particular to a method, device, equipment, storage medium and program product for capturing hidden faults of equipment. Background Art

[0002] At present, the judgment of faults or abnormalities in equipment is mainly achieved through sensors, and it relies on comparing the data collected by the sensor with the corresponding preset thresholds to determine whether a fault or abnormality has occurred. When the physical quantity measured by the sensor (such as temperature, pressure, vibration, etc.) exceeds the normal operating range, that is, exceeds the set threshold, the sensor will trigger an alarm to indicate a possible fault. However, this method is more effective for explicit faults, that is, faults in which parameters obviously deviate from the normal range, but it has certain limitations and disadvantages when detecting implicit faults.

[0003] Therefore, how to quickly and accurately detect and capture hidden faults in equipment based on the collected sensor data has become a technical problem that technicians urgently need to solve. Summary of the invention

[0004] The present invention provides a method, device, equipment, storage medium and program product for capturing hidden faults of equipment, so as to solve the technical problem of detecting and capturing hidden faults in equipment.

[0005] According to one aspect of the present invention, a method for capturing hidden faults of a device is provided, comprising:

[0006] Acquire sensor data collected by sensors of different functional types at at least one same timestamp within a preset time period; the sensors of different functional types are used to collect data of different data types for at least one device;

[0007] Generate at least one data combination to be evaluated based on the sensor collected data at the same timestamp; the data combination to be evaluated includes sensor collected data of sensors of different functional types;

[0008] Inputting the sensor data collected in each of the data combinations to be evaluated into a pre-trained hidden fault prediction model to obtain hidden fault probability values ​​corresponding to each of the data combinations to be evaluated output by the model;

[0009] Hidden faults of the equipment are captured according to the hidden fault probability values ​​of the data combinations to be evaluated.

[0010] According to another aspect of the present invention, a device for capturing hidden faults of equipment is provided, comprising:

[0011] A data acquisition module, used to acquire sensor data collected by sensors of different functional types at at least one same timestamp within a preset time period; the sensors of different functional types are used to collect data of different data types for at least one device;

[0012] A generating module, configured to generate at least one data combination to be evaluated according to the sensor collected data at the same timestamp; the data combination to be evaluated includes sensor collected data of sensors of different functional types;

[0013] A probability value acquisition module is used to input the sensor collected data in each of the data combinations to be evaluated into a pre-trained hidden fault prediction model to obtain the hidden fault probability value corresponding to each of the data combinations to be evaluated output by the model;

[0014] The fault capture module is used to capture hidden faults of the device according to the hidden fault probability value of each combination of data to be evaluated.

[0015] According to another aspect of the present invention, an electronic device is provided, the electronic device comprising:

[0016] at least one processor; and

[0017] a memory communicatively connected to the at least one processor; wherein,

[0018] The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the device hidden fault capture method described in any embodiment of the present invention.

[0019] According to another aspect of the present invention, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the device hidden fault capture method described in any embodiment of the present invention when executed.

[0020] According to another aspect of the present invention, a computer program product is provided. The computer program product includes a computer program. When the computer program is executed by a processor, the method for capturing hidden faults of a device according to any embodiment of the present invention is implemented.

[0021] The technical solution of the embodiment of the present invention is to obtain sensor data collected by sensors of different functional types at at least one same timestamp within a preset time period; generate at least one data combination to be evaluated based on the sensor data collected at each same timestamp; input each sensor data in each data combination to be evaluated into a hidden fault prediction model obtained by pre-training, and obtain the hidden fault probability values ​​corresponding to each data combination to be evaluated output by the model; and capture hidden faults of the device based on the hidden fault probability values ​​of each data combination to be evaluated. This technical solution can obtain different types of sensor data collected at the same timestamp and generate different data combinations to be evaluated, thereby solving the technical problem that a single sensor data is difficult to determine the state of the device and capturing hidden faults. Furthermore, by using the hidden fault prediction model obtained by pre-training, the hidden fault probability of each data combination to be evaluated is calculated, thereby improving the efficiency and accuracy of capturing hidden faults.

[0022] It should be understood that the contents described in this section are not intended to identify the key or important features of the embodiments of the present invention, nor are they intended to limit the scope of the present invention. Other features of the present invention will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0024] Figure 1 is a flow chart of a method for capturing hidden faults of a device provided according to the first embodiment of the present invention;

[0025] Figure 2 is a flow chart of a method for capturing hidden faults of a device provided according to Embodiment 2 of the present invention;

[0026] Figure 3 2 is a schematic diagram of the structure of a device for capturing hidden faults of equipment provided according to Embodiment 3 of the present invention;

[0027] Figure 4 It is a structural schematic diagram of an electronic device for implementing the device hidden fault capturing method according to an embodiment of the present invention. DETAILED DESCRIPTION

[0028] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0029] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0030] Embodiment 1

[0031] Figure 1 A flowchart of a method for capturing hidden faults of a device is provided for the first embodiment of the present invention. This embodiment is applicable to the case of capturing hidden faults in a device. The method can be executed by a device for capturing hidden faults. The device for capturing hidden faults can be implemented in the form of hardware and / or software. The device for capturing hidden faults can be configured in an electronic device. Figure 1 As shown, the method includes:

[0032] S110, acquiring sensor data collected by sensors of different functional types at at least one same timestamp within a preset time period; sensors of different functional types are used to collect data of different data types for at least one device.

[0033] Among them, the preset time period can be manually set by relevant technical personnel based on actual experience, or it can be automatically generated. For example, the preset time period can be one minute or five minutes, specifically collecting data once every minute, or collecting data once every five minutes. This embodiment does not make specific restrictions.

[0034] Among them, sensors of different functional types may refer to sensors that collect different data types, and may specifically include temperature sensors, humidity sensors, gas sensors, and imaging sensors, etc.

[0035] The same timestamp may be the same time point under a time reference determined based on a global clock synchronization protocol, and the global clock synchronization protocol may include a network time protocol or a precision time protocol. Specifically, the same time point information may be propagated to all sensors of different types to ensure that all sensors use the same time reference, so as to ensure that all sensors can collect data under at least one same timestamp.

[0036] Preferably, a sensor network can be established, in which all sensors belong to the same sensor network, and the sensor data acquisition module is connected to the sensor network to obtain the data collected by different types of sensors at the same timestamp, wherein the data collected may include temperature, humidity and pressure, and then the corresponding sensor type is selected, and then the layout position and number of sensors are determined according to the size and shape of the area to be monitored. Specifically, the sensor data acquisition module can obtain the data collected by different types of sensors at the same timestamp, wherein the sensor network can adopt a star topology structure, wherein a central node is used as a data collector, and different types of sensors are connected to the central node as peripheral nodes, and the central node is connected to the sensor data acquisition module, thereby extracting the data collected by different types of sensors in real time; it can also be a mesh topology structure, wherein sensor nodes can communicate directly with each other, and the mesh topology structure can provide better robustness and reliability, which is not specifically limited in this embodiment.

[0037] S120, generating at least one data combination to be evaluated according to the sensor collected data at the same timestamp;

[0038] The data combination to be evaluated may include sensor collected data at the same timestamp, and specifically may include sensor collected data of sensors of different functional types.

[0039] Optionally, at least one data combination to be evaluated is generated based on the sensor collected data at the same timestamp, including: based on the sensor collected data at the same timestamp and based on the abnormal data judgment conditions of the sensors of each functional type, at least one reference timestamp is determined from the same timestamps; for any reference timestamp, the sensor collected data of the sensors of each functional type at the reference timestamp are combined to obtain the collected data combination at the reference timestamp; and the collected data combination at each reference timestamp is determined as the data combination to be evaluated.

[0040] The reference timestamp may be the time point at which the collected data obtained by one or more sensors first becomes abnormal, or may be any time point after the time point at which the collected data obtained by the sensor first becomes abnormal.

[0041] Among them, the abnormal data judgment condition can be a condition for determining whether the collected data obtained by each type of sensor is abnormal. For example, if the temperature to ensure that the device can operate normally is 90-120 degrees Celsius, it can be determined whether the device operating temperature obtained by the temperature sensor is within this range. If the temperature data collected by the temperature sensor is 100 degrees Celsius, it can be determined that the temperature data is normal; if the collected temperature data is 145 degrees Celsius, it can be determined that the data collected by the temperature sensor is abnormal.

[0042] Specifically, obtaining the combination of collected data at the reference timestamp may be to extract the collected data of all different types of sensors at the time point when the abnormality first occurs in the collected data and at any time point thereafter, and perform standardization or normalization processing so that the extracted collected data are within the same numerical range. For the collected data of different types of sensors within the same numerical range, corresponding feature vectors can be generated. Furthermore, the feature vectors of different types of sensors are combined in a manner of timestamp alignment to obtain the combination of collected data at the reference timestamp, and the combination of collected data at the reference timestamp is used as the data combination to be evaluated.

[0043] S130, inputting the data collected by each sensor in each data combination to be evaluated into a pre-trained hidden fault prediction model, and obtaining hidden fault probability values ​​corresponding to each data combination to be evaluated output by the model.

[0044] Among them, the hidden fault prediction model can be a data model used to calculate the probability value of hidden faults in the equipment based on the collected data of different types of sensors in the data combination to be evaluated. Specifically, it can be pre-trained by relevant technical personnel based on actual experience.

[0045] Specifically, the collection data combination can be to combine the explicit features extracted from the collection data with abnormal results and the implicit features extracted from the collection data with non-abnormal results in the sensor collection data at the reference timestamp. Specifically, the collection data combination can be,

[0046] X n =(a1,a2,…,a m ,b1,b2,…,b d );

[0047] Among them, X n is the dominant feature a in the nth combination m and recessive trait b d is a set of, m is the total number of dominant features in the combination, and d is the total number of recessive features in the combination.

[0048] Furthermore, the explicit features and implicit features in the collected data combination can be input into the pre-trained implicit fault prediction model to calculate the probability value of the equipment having an implicit fault. n As the input of the hidden fault prediction model, the probability vector output by the hidden fault prediction model is It can be expressed as the probability of k combinations triggering hidden faults, k = n; where p i It can be the probability that the hidden fault belongs to the i-th combination. Specifically, the hidden fault probability value can be calculated as follows:

[0049]

[0050] Among them, z i can be the value corresponding to the ith combination in the output of the hidden fault prediction model, and e can be the base of the natural logarithm. It can be used to ensure that all hidden fault probability values ​​add up to 1.

[0051] S140: Capture hidden faults of the equipment according to the hidden fault probability values ​​of each combination of data to be evaluated.

[0052] Among them, according to the calculated hidden fault probability value of each data combination to be evaluated, the sensor acquisition data corresponding to each data combination to be evaluated can be used as a trigger condition for capturing the hidden fault of the equipment. Specifically, the sensor acquisition data corresponding to each evaluation data combination can be input into the hidden fault capture module to capture the hidden fault.

[0053] Optionally, hidden faults of the device are captured according to the hidden fault probability values ​​of each data combination to be evaluated, including: determining a target acquisition data combination according to the hidden fault probability values ​​of each data combination to be evaluated; determining a target hidden fault capture condition according to the sensor acquisition data in the target acquisition data combination; and capturing hidden faults of the device based on the target hidden fault capture condition.

[0054] The target data collection combination may be a combination of sensor data collection of various types at the same timestamp for triggering the capture of hidden faults of the equipment, and specifically may be a sensor data collection combination with the largest hidden fault probability value among the data combinations to be evaluated.

[0055] Specifically, data collected by sensors of various types with the largest hidden fault probability values ​​in the data combinations to be evaluated may be used as trigger conditions for capturing hidden faults of equipment, and hidden faults may be captured by hidden fault capture modules.

[0056] Optionally, after performing hidden fault capture on the device based on the hidden fault capture condition, the method further includes: if the hidden fault capture of the device fails, determining at least one candidate acquisition data combination and the priority of each candidate acquisition data combination according to the hidden fault probability value of each data combination to be evaluated; determining candidate hidden fault capture conditions corresponding to each candidate acquisition data combination according to the sensor acquisition data in each candidate acquisition data combination; and performing hidden fault capture on the device according to each candidate hidden fault capture condition and based on the priority corresponding to each candidate acquisition data combination.

[0057] Among them, the candidate acquisition data combination can be a collection data combination of various types of sensors at the same timestamp whose calculated hidden fault probability value exceeds the preset probability threshold. Specifically, the preset probability threshold can be manually set by relevant technical personnel based on experience, and can also be automatically generated according to the hidden fault probability value of the sensor collection data in the historical time period and the captured hidden fault situation. Exemplarily, the preset probability threshold can be 20%, wherein the calculated hidden fault probability value of the data combination a to be evaluated is 35%; the hidden fault probability value of the data combination b to be evaluated is 23%; the hidden fault probability value of the data combination c to be evaluated is 27%; the hidden fault probability value of the data combination d to be evaluated is 15%, then the candidate acquisition data combination can be determined to be the data combinations a, b and c to be evaluated.

[0058] Among them, determining the priority of each candidate data collection combination may be determining the hidden fault probability value of each candidate data collection combination, and prioritizing them according to the hidden fault probability value. Continuing with the above example, the priority of each candidate data collection combination may be that the priority of the data combination to be evaluated a is the highest, the priority of the data combination to be evaluated c is the second highest, and the priority of the data combination to be evaluated b is the lowest.

[0059] Specifically, the acquired data corresponding to each type of transmitter may be sequentially input into the hidden fault capture module according to the priority of each candidate data acquisition combination as a trigger condition for capturing the hidden fault. For example, the sensor acquisition data corresponding to the data acquisition combination a with the largest hidden fault probability value may be used as a trigger condition for capturing the hidden fault for hidden fault capture; if the capture fails, the sensor acquisition data corresponding to the data acquisition combination c may be used as a trigger condition for capturing the hidden fault, and so on, the sensor acquisition data corresponding to the data acquisition combination b may be used as a trigger condition for capturing the hidden fault, until the hidden fault is completely captured.

[0060] For example, the analysis of the hidden faults that may exist in the device through the above embodiment can obtain the final and only analysis result: the temperature of module A reaches 200°C, and module B stops running, causing the hidden fault of the device. Then the trigger condition for the hidden fault of the device can be: the temperature of device A reaches 200°C, and device B stops running.

[0061] The technical solution of the embodiment of the present invention is to obtain sensor data collected by sensors of different functional types at at least one same timestamp within a preset time period; generate at least one data combination to be evaluated based on the sensor data collected at each same timestamp; input each sensor data in each data combination to be evaluated into a hidden fault prediction model obtained by pre-training, and obtain the hidden fault probability values ​​corresponding to each data combination to be evaluated output by the model; and capture hidden faults of the device based on the hidden fault probability values ​​of each data combination to be evaluated. This technical solution can obtain different types of sensor data collected at the same timestamp and generate different data combinations to be evaluated, thereby solving the technical problem that it is difficult to determine the state of the device with a single sensor data and to capture hidden faults of the device. Furthermore, the hidden fault prediction model obtained by pre-training can be used to calculate the hidden fault probability of each data combination to be evaluated, thereby improving the efficiency and accuracy of capturing hidden faults.

[0062] Embodiment 2

[0063] Figure 2 This is a flow chart of a method for capturing hidden faults of a device provided in the second embodiment of the present invention. This embodiment further optimizes the method for capturing hidden faults of a device based on the above technical solutions.

[0064] Furthermore, before the step of "obtaining sensor data collected by sensors of different functional types at at least one same timestamp within a preset time period; sensors of different functional types are used to collect data of different data types for at least one device", add the step of "obtaining historical data corresponding to each type of functional sensor at at least one abnormal time point within a historical time period; the historical data is abnormal explicit data or normal implicit data; for each historical data at any abnormal time point, according to the data type of the historical data, generating label data of the corresponding historical data, and combining each historical data to obtain a sample data combination at the abnormal time point; inputting the sample data combination at each abnormal time point into a pre-constructed semi-supervised network model for model training until the model training end condition is met to obtain a hidden fault prediction model that has completed the training." to improve the model training method for the hidden fault prediction model.

[0065] like Figure 2 As shown, the method includes:

[0066] S210: Acquire historical collection data corresponding to each type of functional sensor at at least one abnormal time point in a historical time period.

[0067] The abnormal time point may be a time point at which the collected data obtained by the sensor becomes abnormal.

[0068] The historical collected data may be abnormal explicit data or normal implicit data. The abnormal explicit data may be sensor collected data with abnormal results; the normal implicit data may be sensor collected data with normal results. Specifically, when abnormal data appears in the collected data obtained by the sensor during the historical time period, the collected data of each type of sensor at the same time point is obtained.

[0069] Optionally, historical collected data corresponding to each type of functional sensor at at least one abnormal time point in a historical time period is obtained, including: obtaining multi-source heterogeneous data in the historical time period; determining multiple abnormal time points based on the multi-source heterogeneous data; and obtaining historical collected data corresponding to each type of functional sensor at each abnormal time point.

[0070] Among them, multi-source heterogeneous data can be a data set that does not include sensor collection data and has different data formats, data structures, data contents and data sources. Specifically, it can be equipment operation data acquired in a historical time period, which may include equipment maintenance data, alarm data and repair data.

[0071] The abnormal time point may be a time point at which abnormal sensor data is collected in multi-source heterogeneous data, and specifically may be a time point at which abnormal data is collected from various types of sensors in a historical time period. Specifically, the historical data collected from various types of sensors may be acquired based on the time point at which abnormal data is collected from various types of sensors in a historical time period.

[0072] S220. For each historical collection data at any abnormal time point, generate label data of the corresponding historical collection data according to the data type of the historical collection data, and combine each historical collection data to obtain a sample data combination at the abnormal time point.

[0073] The data type may include explicit data with abnormal results in the historical data collected by various types of sensors, and may also include implicit data with normal results. The label data may be data generated by marking the explicit features of the explicit data in the historical data.

[0074] Specifically, the explicit features of the marked explicit collected data and the implicit features of the unmarked implicit collected data in the historical collected data of each type of sensor, as well as the multi-source heterogeneous features, can be combined to obtain the sensor collected data combination at the abnormal time point. Among them, the multi-source heterogeneous features can be obtained by extracting features from the abnormal sensor collected data appearing in the multi-source heterogeneous data.

[0075] S230, inputting the sample data combination at each abnormal time point into a pre-built semi-supervised network model for model training until the model training end condition is met, thereby obtaining a hidden fault prediction model that has completed training.

[0076] Among them, model training can be supervised training through explicit features of explicit data in marked historical collected data, and unsupervised training through implicit features and multi-source heterogeneous features of unmarked implicit collected data, and iterative adjustment of model parameters. Specifically, in the process of iterative adjustment of model parameters, the gradient of model parameters can be determined by calculating the loss function, and the model parameters can be updated along the direction of the negative gradient. Among them, the adjustment of model parameters can be achieved through gradient descent methods, such as stochastic gradient descent and other algorithms. When the change of the loss function is less than the preset threshold, it is determined that the model has converged, that is, the end condition of model training is met, and a trained implicit fault prediction model can be obtained.

[0077] S240, acquiring sensor data collected by sensors of different functional types at at least one same timestamp within a preset time period; sensors of different functional types are used to collect data of different data types for at least one device.

[0078] S250: Generate at least one data combination to be evaluated based on the sensor collected data at the same timestamp; the data combination to be evaluated includes sensor collected data of sensors of different functional types.

[0079] 260. Input the data collected by each sensor in each data combination to be evaluated into a pre-trained hidden fault prediction model to obtain hidden fault probability values ​​corresponding to each data combination to be evaluated output by the model.

[0080] S270: Capture hidden faults of the equipment according to the hidden fault probability values ​​of each combination of data to be evaluated.

[0081] The technical solution of the embodiment of the present invention is to obtain the historical collected data corresponding to each type of functional sensor at at least one abnormal time point in the historical time period; for each historical collected data at any abnormal time point, according to the data type of the historical collected data, obtain the sample data combination at the abnormal time point; input it into the pre-built semi-supervised network model for model training to obtain a hidden fault prediction model that has completed training. Furthermore, by inputting at least one combination of data to be evaluated generated according to the sensor collected data at the same timestamp into the hidden fault prediction model that has completed training, the hidden fault probability value corresponding to each combination of data to be evaluated is calculated, thereby improving the efficiency and accuracy of capturing hidden faults of the equipment. It solves the technical problem that it is difficult to determine the status of the equipment with a single sensor data and to capture hidden faults of the equipment.

[0082] Embodiment 3

[0083] Figure 3 The schematic diagram of the structure of a device hidden fault capture device provided in the third embodiment of the present invention. The device hidden fault capture device provided in the embodiment of the present invention can be applied to the situation where different types of sensor data are combined to capture hidden faults in the device. The device hidden fault capture device can be implemented in the form of hardware and / or software, such as Figure 3 As shown, it specifically includes: a data acquisition module 310, a generation module 320, a probability value acquisition module 330 and a fault capture module 340. Among them,

[0084] The data acquisition module 310 is used to acquire sensor data collected by sensors of different functional types at at least one same timestamp within a preset time period; the sensors of different functional types are used to collect data of different data types for at least one device;

[0085] A generating module 320, configured to generate at least one data combination to be evaluated according to the sensor collected data at the same timestamp; the data combination to be evaluated includes sensor collected data of sensors of different functional types;

[0086] The probability value acquisition module 330 is used to input the sensor collected data in each of the data combinations to be evaluated into the pre-trained hidden fault prediction model, and obtain the hidden fault probability value corresponding to each of the data combinations to be evaluated output by the model;

[0087] The fault capturing module 340 is used to capture hidden faults of the device according to the hidden fault probability value of each combination of data to be evaluated.

[0088] This technical solution can obtain different types of sensor data collected at the same timestamp and generate different combinations of data to be evaluated, solving the technical problem that it is difficult to determine the status of the device with a single sensor data and capture hidden faults. Furthermore, by using the hidden fault prediction model obtained by pre-training, the hidden fault probability of each combination of data to be evaluated is calculated, thereby improving the efficiency and accuracy of capturing hidden faults.

[0089] Optionally, the generation module 320 is specifically used to determine at least one reference timestamp from each of the same timestamps according to the sensor acquisition data at each of the same timestamps and based on the abnormal data judgment conditions of each of the functional type sensors; for any reference timestamp, combine the sensor acquisition data of each functional type sensor at the reference timestamp to obtain the acquisition data combination at the reference timestamp; determine the acquisition data combination at each of the reference timestamps as the data combination to be evaluated.

[0090] Optionally, the fault capture module 340 is specifically used to determine a target data acquisition combination according to the hidden fault probability value of each data combination to be evaluated; determine a target hidden fault capture condition according to the sensor acquisition data in the target data acquisition combination; and perform hidden fault capture on the device based on the target hidden fault capture condition.

[0091] Optionally, the fault capture module 340 is also specifically used to, after performing hidden fault capture on the device based on the hidden fault capture condition, if the hidden fault capture of the device fails, determine at least one candidate acquisition data combination according to the hidden fault probability value of each data combination to be evaluated, and determine the priority of each candidate acquisition data combination; determine the candidate hidden fault capture conditions corresponding to each candidate acquisition data combination according to the sensor acquisition data in each candidate acquisition data combination; and perform hidden fault capture on the device according to each candidate hidden fault capture condition and based on the priority corresponding to each candidate acquisition data combination.

[0092] Optionally, the device also includes a hidden fault prediction model training module 350.

[0093] Optionally, the implicit fault prediction model training module 350 is used to obtain historical collected data corresponding to each type of functional sensor at at least one abnormal time point in a historical time period; the historical collected data is abnormal explicit data or normal implicit data; for each historical collected data at any abnormal time point, according to the data type of the historical collected data, generate label data of the corresponding historical collected data, and combine the historical collected data to obtain the sample data combination at the abnormal time point; input the sample data combination at each abnormal time point into a pre-constructed semi-supervised network model for model training until the model training end condition is met to obtain a hidden fault prediction model that has completed the training.

[0094] Optionally, the hidden fault prediction model training module 350 includes: a multi-source heterogeneous data acquisition unit, an abnormal time point determination unit and a historical collection data acquisition unit.

[0095] Optionally, a multi-source heterogeneous data acquisition unit is used to acquire multi-source heterogeneous data in a historical time period;

[0096] An abnormal time point determination unit, used to determine multiple abnormal time points according to the multi-source heterogeneous data;

[0097] The historical collection data acquisition unit is used to acquire the historical collection data corresponding to each type of functional sensor at each abnormal time point.

[0098] The device for capturing hidden faults of equipment provided by the embodiment of the present invention can execute the method for capturing hidden faults of equipment provided by any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0099] Embodiment 4

[0100] Figure 4 A schematic diagram of the structure of an electronic device 40 that can be used to implement an embodiment of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices (such as helmets, glasses, watches, etc.) and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present invention described and / or required herein.

[0101] like Figure 4As shown, the electronic device 40 includes at least one processor 41, and a memory connected to the at least one processor 41, such as a read-only memory (ROM) 42, a random access memory (RAM) 43, etc., wherein the memory stores a computer program that can be executed by at least one processor, and the processor 41 can perform various appropriate actions and processes according to the computer program stored in the read-only memory (ROM) 42 or the computer program loaded from the storage unit 48 to the random access memory (RAM) 43. In the RAM 43, various programs and data required for the operation of the electronic device 40 can also be stored. The processor 41, the ROM 42, and the RAM 43 are connected to each other through a bus 44. An input / output (I / O) interface 45 is also connected to the bus 44.

[0102] A number of components in the electronic device 40 are connected to the I / O interface 45, including: an input unit 46, such as a keyboard, a mouse, etc.; an output unit 47, such as various types of displays, speakers, etc.; a storage unit 48, such as a disk, an optical disk, etc.; and a communication unit 49, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 49 allows the electronic device 40 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0103] The processor 41 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the processor 41 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any appropriate processor, controller, microcontroller, etc. The processor 41 executes the various methods and processes described above, such as a device hidden fault capture method.

[0104] In some embodiments, the device hidden fault capture method can be implemented as a computer program, which is tangibly contained in a computer-readable storage medium, such as a storage unit 48. In some embodiments, part or all of the computer program can be loaded and / or installed on the electronic device 40 via the ROM 42 and / or the communication unit 49. When the computer program is loaded into the RAM 43 and executed by the processor 41, one or more steps of the device hidden fault capture method described above can be performed. Alternatively, in other embodiments, the processor 41 can be configured to execute the device hidden fault capture method in any other appropriate manner (for example, by means of firmware).

[0105] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0106] Computer programs for implementing the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that when the computer program is executed by the processor, the functions / operations specified in the flow chart and / or block diagram are implemented. The computer program may be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.

[0107] In the context of the present invention, a computer-readable storage medium may be a tangible medium that may contain or store a computer program for use by or in combination with an instruction execution system, device or equipment. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices or equipment, or any suitable combination of the foregoing. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0108] To provide interaction with a user, the systems and techniques described herein may be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and a pointing device (e.g., a mouse or trackball) through which the user can provide input to the electronic device. Other types of devices may also be used to provide interaction with the user; for example, the feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form (including acoustic input, voice input, or tactile input).

[0109] The systems and techniques described herein may be implemented in a computing system that includes backend components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes frontend components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such backend components, middleware components, or frontend components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), a blockchain network, and the Internet.

[0110] A computing system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The client and server relationship is generated by computer programs running on the corresponding computers and having a client-server relationship with each other. The server may be a cloud server, also known as a cloud computing server or cloud host, which is a host product in the cloud computing service system to solve the defects of difficult management and weak business scalability in traditional physical hosts and VPS services.

[0111] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps described in the present invention can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solution of the present invention can be achieved, and this document does not limit this.

[0112] The above specific implementations do not constitute a limitation on the protection scope of the present invention. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for capturing hidden faults of equipment, characterized in that: include: Acquire sensor data collected by sensors of different functional types at at least one same timestamp within a preset time period; The sensors of different functional types are used to collect data of different data types from at least one device; Generate at least one data combination to be evaluated based on the sensor collected data at the same timestamp; the data combination to be evaluated includes sensor collected data of sensors of different functional types; Inputting the data collected by each sensor in each of the data combinations to be evaluated into a pre-trained hidden fault prediction model, respectively, to obtain hidden fault probability values ​​corresponding to each of the data combinations to be evaluated output by the model; Hidden faults of the equipment are captured according to the hidden fault probability values ​​of the data combinations to be evaluated.

2. The method according to claim 1, characterized in that The step of generating at least one combination of data to be evaluated according to the sensor collected data at the same timestamp comprises: According to the sensor collected data at the same timestamps, based on the abnormal data judgment conditions of the functional type sensors, at least one reference timestamp is determined from the same timestamps; For any reference timestamp, the sensor collected data of sensors of each functional type at the reference timestamp are combined to obtain a collection data combination at the reference timestamp; The collected data combinations at the reference timestamps are determined as the data combinations to be evaluated.

3. The method according to claim 1, characterized in that The step of capturing hidden faults of the device according to the hidden fault probability values ​​of the data combinations to be evaluated includes: Determining a target acquisition data combination according to the hidden fault probability value of each of the data combinations to be evaluated; Determine the target hidden fault capture condition according to the sensor collection data in the target collection data combination; Based on the target hidden fault capturing condition, hidden fault capturing is performed on the device.

4. The method according to claim 3, characterized in that After performing hidden fault capture on the device based on the hidden fault capture condition, the method further includes: If the hidden fault of the device fails to be captured, then at least one candidate acquisition data combination is determined according to the hidden fault probability value of each of the data combinations to be evaluated, and the priority of each of the candidate acquisition data combinations is determined; Determining candidate hidden fault capture conditions corresponding to each candidate collection data combination according to the sensor collection data in each candidate collection data combination; According to each of the candidate hidden fault capturing conditions and based on the priorities corresponding to each of the candidate collected data combinations, hidden fault capturing is performed on the device.

5. The method according to claim 1, characterized in that The model training method of the hidden fault prediction model is as follows: Acquire historical collected data corresponding to each type of functional sensor at at least one abnormal time point in a historical time period; the historical collected data is abnormal explicit data or normal implicit data; For each historical collection data at any abnormal time point, generate label data of the corresponding historical collection data according to the data type of the historical collection data, and combine the historical collection data to obtain the sample data combination at the abnormal time point; The sample data combination at each abnormal time point is input into a pre-built semi-supervised network model for model training until the model training end condition is met, thereby obtaining a hidden fault prediction model that has completed training.

6. The method according to claim 5, characterized in that The acquiring of historical collected data corresponding to each type of functional sensor at at least one abnormal time point in a historical time period includes: Acquire multi-source heterogeneous data in historical time periods; Determining multiple abnormal time points according to the multi-source heterogeneous data; The historical collected data corresponding to each type of functional sensor at each abnormal time point are obtained.

7. A device for capturing hidden faults of equipment, characterized in that: include: A data acquisition module, used to acquire sensor data collected by sensors of different functional types at at least one same timestamp within a preset time period; The sensors of different functional types are used to collect data of different data types from at least one device; A generating module, configured to generate at least one data combination to be evaluated according to the sensor collected data at the same timestamp; the data combination to be evaluated includes sensor collected data of sensors of different functional types; A probability value acquisition module is used to input the sensor collected data in each of the data combinations to be evaluated into a pre-trained hidden fault prediction model to obtain the hidden fault probability value corresponding to each of the data combinations to be evaluated output by the model; The fault capture module is used to capture hidden faults of the device according to the hidden fault probability value of each combination of data to be evaluated.

8. An electronic device, characterized in that: The electronic device comprises: at least one processor; and a memory communicatively connected to the at least one processor; wherein, The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the device hidden fault capture method according to any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the device hidden fault capture method according to any one of claims 1 to 6 when executed.

10. A computer program product, characterized in that The computer program product comprises a computer program, and when the computer program is executed by a processor, the method for capturing hidden faults of a device according to any one of claims 1 to 6 is implemented.

Images