Image voiceprint multi-source perception data cross mutual verification method and device and electronic equipment

Abstract

The invention provides an image voiceprint multi-source perception data cross mutual verification method and device and electronic equipment. The method comprises the steps that a data feature set of each data channel in a current observation time window is acquired, fault identification is carried out on the electrical equipment according to the data feature set, a fault identification result is generated, and the fault identification result comprises a fault data sequence of each data channel; obtaining a target fault data sequence of the target data channel from the fault recognition result, and determining other data channels related to the target data channel and other fault data sequences corresponding to the other data channels according to the starting time of the target fault data sequence; and determining a cross mutual verification result of the target data channel at least according to the target fault data sequence and other fault data sequences, and determining an equipment fault identification result of the target data channel according to the cross mutual verification result. The problem that a traditional power distribution network equipment fault identification cross-checking method is low in fault identification accuracy is solved.

Description

Technical Field

[0001] The present application relates to the technical field of cross-verification of multi-source perception data of images and voiceprints, and in particular to a method for cross-verification of multi-source perception data of images and voiceprints, a device for cross-verification of multi-source perception data of images and voiceprints, a computer-readable storage medium and an electronic device. Background Art

[0002] With the continuous advancement of smart grid technology and the accelerated advancement of new distribution network construction, the number of equipment in the distribution network has increased significantly, and the frequency and types of equipment failures have also increased. As a key participant in power transmission and distribution, the stable operation of distribution network equipment is directly related to the overall reliability and safety of the power system. Traditional fault identification methods often rely on manual inspections and empirical judgments, which are not only inefficient, but also difficult to achieve early warning and accurate positioning of faults. In recent years, distribution network equipment fault identification methods based on big data, artificial intelligence and Internet of Things technologies have emerged. These methods can automatically identify abnormal equipment states, predict potential faults, and quickly locate fault points when faults occur by monitoring equipment operation data in real time and using machine learning algorithms to analyze data features. Therefore, researching and developing efficient and intelligent distribution network equipment fault identification methods is of great significance for improving fault response speed and ensuring power supply quality.

[0003] Cross-checking based on different identification results can avoid false alarms and missed alarms, and improve the reliability, accuracy and certainty of distribution network equipment fault identification. However, the existing cross-checking methods still have the following problems. First, the traditional cross-checking method ignores the synergy of empirical performance and correlation, and the reliability of the cross-checking results is low. Second, the traditional distribution network equipment fault identification cross-checking method cannot update parameters according to different identification and detection situations, and the identification accuracy and certainty are low. Summary of the invention

[0004] The main purpose of the present application is to provide an image and voiceprint multi-source perception data cross-verification method, an image and voiceprint multi-source perception data cross-verification device, a computer-readable storage medium and an electronic device, so as to at least solve the problem of low accuracy in fault identification by traditional distribution network equipment fault identification cross-verification methods.

[0005] In order to achieve the above-mentioned purpose, according to one aspect of the present application, a method for cross-verification of image and voiceprint multi-source perception data is provided, comprising: obtaining a data feature set of each data channel in a current observation time window, and performing fault identification on electrical equipment according to the data feature set to generate a fault identification result, wherein the data feature set includes data features of the image data channel and data features of the voiceprint data channel, and the fault identification result includes a fault data sequence of each of the data channels; obtaining a target fault data sequence of a target data channel from the fault identification result, and determining other data channels related to the target data channel and other fault data sequences corresponding to the other data channels according to the start time of the target fault data sequence; determining the cross-verification result of the target data channel at least according to the target fault data sequence and the other fault data sequences, and determining the equipment fault identification result of the target data channel according to the cross-verification result.

[0006] Optionally, determining the cross verification result of the target data channel at least according to the target fault data sequence and the other fault data sequences corresponding to the other data channels includes: when it is determined that the target data channel is the image data channel, according to a first formula:

[0007] Determine the cross-check result of the target data channel, wherein: is the empirical performance of image-based fault detection accuracy h, For other data channels The cross validation weight for fault h, For image data channel The empirical performance of the recognition accuracy of fault h, For image data channel With other data channels The correlation between the data, β 1 and β 2 is the weight coefficient, The fault type identification result of the target data channel.

[0008] Optionally, determining the cross verification result of the target data channel at least according to the target fault data sequence and the other fault data sequences corresponding to the other data channels includes: when it is determined that the target data channel is the voiceprint data channel, according to a second formula:

[0009] Determine the cross-check result of the target data channel, wherein: is the empirical performance based on the accuracy of detecting fault h, For other data channels The cross validation weight for fault h, Voiceprint data channel The empirical performance of the recognition accuracy of fault h, Voiceprint data channel With other data channels The correlation between the data, β 1 and β 2 is the weight coefficient, The fault type identification result of the target data channel.

[0010] Optionally, before determining the cross-verification result of the target data channel according to the first formula, the method further includes: according to a third formula:

[0011] Determine the fault type identification result of the target data channel Among them, Me{·} means taking the majority, C im,exp Represents a collection of other data channels determined by empirical performance.

[0012] Optionally, after determining the device fault identification result of the target data channel according to the cross-verification result, the method further includes: obtaining an actual maintenance result; and determining whether the device fault identification result of the target data channel is accurate according to the maintenance result.

[0013] Optionally, after determining whether the equipment fault identification result of the target data channel is accurate based on the maintenance result, the method further includes: in the case that the equipment fault identification result of the target data channel is inaccurate, determining the identification error type of the target data channel, wherein the identification error type includes a false negative error and a false positive error, the false negative error indicating that the electrical equipment actually fails but the fault is not correctly identified, and the false negative error indicating that the electrical equipment actually does not fail but is identified as having failed; updating the fault identification algorithm of the target data channel according to the identification error type.

[0014] Optionally, the fault identification algorithm of the target data channel is updated according to the identification error type, including: when the target data channel is an image data channel and the identification error type is a false negative error, updating the fault identification algorithm and, based on the identification accuracy and certainty of the target data channel, updating the correlation of the channel time data of the target data channel; when the target data channel is an image data channel and the identification error type is a false positive error, updating the correlation of the channel time data of the target data channel based on the identification accuracy and certainty of the target data channel.

[0015] According to another aspect of the present application, there is provided an image and voiceprint multi-source perception data cross-verification device, comprising: a first acquisition unit, used to acquire a data feature set of each data channel in a current observation time window, and perform fault identification on electrical equipment according to the data feature set to generate a fault identification result, wherein the data feature set includes data features of the image data channel and data features of the voiceprint data channel, and the fault identification result includes a fault data sequence of each of the data channels; a first determination unit, used to acquire a target fault data sequence of a target data channel from the fault identification result, and determine other data channels related to the target data channel and other fault data sequences corresponding to the other data channels according to the start time of the target fault data sequence; a second determination unit, used to determine the cross-verification result of the target data channel according to at least the target fault data sequence and the other fault data sequences, and determine the equipment fault identification result of the target data channel according to the cross-verification result.

[0016] According to another aspect of the present application, a computer-readable storage medium is provided, wherein the computer-readable storage medium includes a stored program, wherein when the program is running, the device where the computer-readable storage medium is located is controlled to execute any one of the image voiceprint multi-source perception data cross-verification methods.

[0017] According to another aspect of the present application, an electronic device is provided, comprising: one or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and are configured to be executed by the one or more processors, and the one or more programs include a method for executing any one of the image voiceprint multi-source perception data cross-verification methods.

[0018] By applying the technical solution of the present application, the data feature set of each data channel in the current observation time window is obtained, and the fault identification of the electrical equipment is performed according to the data feature set to generate a fault identification result, wherein the data feature set includes the data features of the image data channel and the data features of the voiceprint data channel, and the fault identification result includes the fault data sequence of each data channel; the target fault data sequence of the target data channel is obtained from the fault identification result, and other data channels related to the target data channel and other fault data sequences corresponding to other data channels are determined according to the start time of the target fault data sequence; the cross-verification result of the target data channel is determined at least according to the target fault data sequence and other fault data sequences, and the equipment fault identification result of the target data channel is determined according to the cross-verification result. For each data channel of the image and voiceprint, the data features of each data channel in the current observation window are extracted, and fault identification is performed according to the data features. For a certain data channel, the data domain cross-verification sequence of other channels is extracted according to the start time of its fault data sequence; finally, according to the empirical performance of other channels on the fault identification accuracy and the correlation of the data between channels, the data is cross-verified to improve the reliability of the cross-verification result. The problem of low accuracy in fault identification using the traditional cross-check method for fault identification of distribution network equipment is solved. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] The drawings constituting part of the present application are used to provide a further understanding of the present application. The illustrative embodiments and descriptions of the present application are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0020] Figure 1 A hardware structure block diagram of a mobile terminal for executing a method for cross-verification of image and voiceprint multi-source perception data provided in an embodiment of the present application is shown;

[0021] Figure 2 A schematic diagram of a process for cross-verification of multi-source perception data of image and voiceprint provided according to an embodiment of the present application is shown;

[0022] Figure 3 A schematic diagram of a process for cross-verification of multi-source perception data of image and voiceprint provided in accordance with an embodiment of the present application is shown;

[0023] Figure 4 A structural block diagram of a device for cross-verification of image and voiceprint multi-source perception data provided according to an embodiment of the present application is shown.

[0024] The above drawings include the following reference numerals:

[0025] 102, processor; 104, memory; 106, transmission device; 108, input and output devices. DETAILED DESCRIPTION

[0026] It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0027] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. 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 creative work should fall within the scope of protection of the present application.

[0028] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequential order. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present application described here. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0029] As introduced in the background technology, the accuracy of fault identification by the traditional distribution network equipment fault identification cross-verification method is low. In order to solve the problem of low fault identification accuracy by the traditional distribution network equipment fault identification cross-verification method, the embodiments of the present application provide an image voiceprint multi-source perception data cross-verification method, an image voiceprint multi-source perception data cross-verification device, a computer-readable storage medium and an electronic device.

[0030] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the accompanying drawings in the embodiments of the present invention.

[0031] The method embodiments provided in the embodiments of the present application can be executed in a mobile terminal, a computer terminal or a similar computing device. Taking running on a mobile terminal as an example, Figure 1 1 is a hardware structure block diagram of a mobile terminal of a method for cross-verification of multi-source perception data of image and voiceprint according to an embodiment of the present invention. Figure 1 As shown, the mobile terminal may include one or more ( Figure 1Only one is shown in the figure) a processor 102 (the processor 102 may include but is not limited to a processing device such as a microprocessor MCU or a programmable logic device FPGA) and a memory 104 for storing data, wherein the mobile terminal may also include a transmission device 106 and an input / output device 108 for communication functions. It can be understood by those skilled in the art that Figure 1 The structure shown is only for illustration and does not limit the structure of the mobile terminal. Figure 1 More or fewer components as shown, or with Figure 1 Different configurations are shown.

[0032] The memory 104 can be used to store computer programs, for example, software programs and modules of application software, such as the computer program corresponding to the image voiceprint multi-source perception data cross-verification method in the embodiment of the present invention. The processor 102 executes various functional applications and data processing by running the computer program stored in the memory 104, that is, to implement the above method. The memory 104 may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some examples, the memory 104 may further include a memory remotely arranged relative to the processor 102, and these remote memories may be connected to the mobile terminal via a network. Examples of the above networks include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof. The transmission device 106 is used to receive or send data via a network. The specific examples of the above networks may include a wireless network provided by a communication provider of the mobile terminal. In one example, the transmission device 106 includes a network adapter (Network Interface Controller, referred to as NIC), which can be connected to other network devices through a base station so as to communicate with the Internet. In one example, the transmission device 106 may be a radio frequency (RF) module, which is used to communicate with the Internet wirelessly.

[0033] In this embodiment, a method for cross-verification of multi-source perception data of images and voiceprints running on a mobile terminal, a computer terminal or a similar computing device is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0034] Figure 2 1 is a flow chart of a method for cross-verification of multi-source perception data of image and voiceprint according to an embodiment of the present application. Figure 2As shown, the method comprises the following steps:

[0035] Step S201, obtaining a data feature set of each data channel in the current observation time window, and performing fault identification on the electrical equipment according to the data feature set to generate a fault identification result, wherein the data feature set includes data features of the image data channel and data features of the voiceprint data channel, and the fault identification result includes a fault data sequence of each of the data channels;

[0036] Specifically, for each data channel of image and voiceprint, extract the data features of each data channel in the current observation time window. Consider T observation time windows in total, and the set is represented by T = {1,...,t,...,T}. Consider N image data channels, and the set is represented by in represents the nth image data channel. Considering M voiceprint data channels, the set is represented as in Represents the mth voiceprint data channel. Definition I n (t) is the image data channel in the tth time window The image data set V m (t) is the voiceprint data channel in the tth time window The feature extraction function is used to extract the data feature set of each data channel in the current observation time window, which is expressed as: is the image data channel in the tth time window The data feature set; is the voiceprint data channel in the tth time window The data feature set; f ext (·) is the feature extraction function.

[0037] Step S202, obtaining a target fault data sequence of a target data channel from the fault identification result, and determining other data channels related to the target data channel and other fault data sequences corresponding to the other data channels according to the start time of the target fault data sequence;

[0038] Specifically, fault identification can be performed based on the extracted data features, and the fault type, fault probability and corresponding data sequence range, i.e., the fault data sequence, are marked in the data sequence of the current observation window. For a certain data channel, the data domain cross-verification sequence of other channels is extracted based on the starting time of its fault data sequence.

[0039] Step S203, determining a cross-verification result of the target data channel at least according to the target fault data sequence and the other fault data sequences, and determining a device fault identification result of the target data channel according to the cross-verification result.

[0040] Specifically, according to the empirical performance of other channels on fault recognition accuracy and the correlation between channel data, cross-verification of data is performed to verify the fault type and fault probability. When the fault type recognition result of the target data channel is the same as that of most other image data channels, the original fault type recognition result is maintained; otherwise, the fault type recognition result is re-determined according to the mode of the fault type recognition results of the expert channel set, thereby improving the reliability of the cross-verification results.

[0041] Through this embodiment, the data feature set of each data channel in the current observation time window is obtained, and the fault identification of the electrical equipment is performed according to the data feature set to generate a fault identification result, wherein the data feature set includes the data features of the image data channel and the data features of the voiceprint data channel, and the fault identification result includes the fault data sequence of each data channel; the target fault data sequence of the target data channel is obtained from the fault identification result, and other data channels related to the target data channel and other fault data sequences corresponding to other data channels are determined according to the start time of the target fault data sequence; the cross-verification result of the target data channel is determined at least according to the target fault data sequence and other fault data sequences, and the equipment fault identification result of the target data channel is determined according to the cross-verification result. For each data channel of the image and voiceprint, the data features of each data channel in the current observation window are extracted, and fault identification is performed according to the data features. For a certain data channel, the data domain cross-verification sequence of other channels is extracted according to the start time of its fault data sequence; finally, according to the empirical performance of other channels on the fault identification accuracy and the correlation of the data between channels, the data is cross-verified to improve the reliability of the cross-verification result. The problem of low accuracy in fault identification using the traditional cross-check method for fault identification of distribution network equipment is solved.

[0042] In a specific implementation process, the cross verification result of the target data channel is determined at least according to the target fault data sequence and the other fault data sequences corresponding to the other data channels, including: when the target data channel is determined to be the image data channel, according to the first formula:

[0043] Determine the cross-validation results of the above target data channels, where: is the empirical performance of image-based fault detection accuracy h, For other data channels The cross validation weight for fault h, For image data channel The empirical performance of the recognition accuracy of fault h, For image data channel With other data channels The correlation between the data, β1 and β 2 is the weight coefficient, This is the fault type identification result of the target data channel.

[0044] The formula of this method means that the image data channel The detection result of fault h needs to comprehensively consider the accuracy of fault h detection based on the image. The fault probability of detecting fault h, as well as the situation of other image channels. The more other channels that detect the occurrence of fault h, the higher the fault probability, and the higher the cross-validation result of the probability of occurrence of fault h. In addition, other image channels with higher empirical performance in the accuracy of fault h recognition, and the image data channels with higher Other image channels with higher correlation are in the image data channel has a higher weight in the cross-validation results.

[0045] Specifically, determining the cross verification result of the target data channel at least according to the target fault data sequence and the other fault data sequences corresponding to the other data channels includes: when it is determined that the target data channel is the voiceprint data channel, according to the second formula:

[0046] Determine the cross-validation results of the above target data channels, where: is the empirical performance based on the accuracy of detecting fault h, For other data channels The cross validation weight for fault h, Voiceprint data channel The empirical performance of the recognition accuracy of fault h, Voiceprint data channel With other data channels The correlation between the data, β 1 and β 2 is the weight coefficient, This is the fault type identification result of the target data channel.

[0047] The formula of this method means that the voiceprint data channel The detection result of fault h needs to comprehensively consider the accuracy of detecting fault h based on voiceprint. The fault probability of detecting fault h, as well as the situation of other voiceprint channels. The more other channels that detect fault h, the higher the fault probability, and the higher the cross-verification result of the fault probability. In addition, other voiceprint channels with higher empirical performance in the recognition accuracy of fault h, and the voiceprint data channels Other image channels with high correlation are in the voiceprint data channel has a higher weight in the cross-validation results.

[0048] More specifically, before determining the cross-verification result of the target data channel according to the first formula, the method further includes: according to the third formula:

[0049] Determine the above fault type identification result of the above target data channel Among them, Me{·} means taking the majority, C im,exp Represents a collection of other data channels determined by empirical performance.

[0050] The formula of this method means that when the image data channel If the fault type recognition result of the expert channel set is the same as that of most other image data channels, the original fault type recognition result is maintained; otherwise, the fault type recognition result is re-determined according to the mode of the fault type recognition results of the expert channel set.

[0051] Furthermore, after determining the equipment fault identification result of the target data channel according to the cross-verification result, the method further includes: obtaining an actual maintenance result; and determining whether the equipment fault identification result of the target data channel is accurate according to the maintenance result.

[0052] This method calculates the accuracy and certainty of fault identification based on actual maintenance results. The actual maintenance results are defined as This means that the actual inspection result is fault h. In particular, Indicates that there is no fault in the actual maintenance, and defines the image data channel obtained based on the actual maintenance results in the t-th time window The fault identification accuracy is Voiceprint data channel The fault identification accuracy is Respectively expressed as:

[0053] The formula means that the fault identification accuracy is related to the fault type identification result of the image data channel or the voiceprint data channel. When the fault type identification result of the image data channel or the voiceprint data channel is the same as the actual inspection result, the fault identification accuracy is 1; otherwise, the fault identification accuracy is 0.

[0054] Furthermore, after determining whether the equipment fault identification result of the target data channel is accurate based on the maintenance result, the method also includes: in the case that the equipment fault identification result of the target data channel is inaccurate, determining the identification error type of the target data channel, wherein the identification error type includes a false negative error and a false positive error, the false negative error indicating that the electrical equipment actually fails but the fault is not correctly identified, and the false negative error indicating that the electrical equipment actually does not fail but is identified as having failed; updating the fault identification algorithm of the target data channel according to the identification error type.

[0055] For false negative errors, that is, a fault actually occurred but was not correctly identified, this method requires updating the existing fault identification algorithm and updating the correlation of channel time data based on the identification accuracy and certainty of the original channel. It can be expressed as: Where η FN is the correlation update step size in the case of false negative errors; γ∈(0,1) is a constant. Taking the image data channel as an example,

[0056] When the image data channel When a false negative error occurs, if the original channel The fault recognition accuracy is 1, which means the fault is successfully recognized. and The correlation between the data and the original channel The higher the fault identification certainty, the larger the correlation update step size; conversely, if the original channel If the fault recognition accuracy is 0, that is, the fault is not successfully recognized, the fixed step size is reduced. and The correlation between data.

[0057] Specifically, the fault identification algorithm of the target data channel is updated according to the above-mentioned identification error type, including: when the above-mentioned target data channel is an image data channel and the above-mentioned identification error type is a false negative error, the above-mentioned fault identification algorithm is updated, and at the same time, based on the identification accuracy and certainty of the above-mentioned target data channel, the correlation of the channel time data of the above-mentioned target data channel is updated; when the above-mentioned target data channel is an image data channel and the above-mentioned identification error type is a false positive error, based on the identification accuracy and certainty of the above-mentioned target data channel, the correlation of the above-mentioned channel time data of the above-mentioned target data channel is updated.

[0058] For false positive errors, that is, a fault is identified as occurring even though it actually does not occur, this method only needs to update the correlation of the channel time data based on the recognition accuracy and certainty of the original channel.

[0059] It is expressed as:

[0060] Where η FP is the correlation update step size in the case of false positive errors. For example, the formula means that when the voiceprint data channel When a false negative error occurs, if the original channel The fault recognition accuracy is 1, which means the fault is successfully recognized. and The correlation between the data and the original channel The higher the fault identification certainty, the larger the correlation update step; conversely, if the original channel If the fault recognition accuracy is 0, that is, the fault is not successfully recognized, then reduce and The correlation between the data and the original channel The higher the fault identification certainty, the smaller the correlation update step size.

[0061] In order to enable those skilled in the art to more clearly understand the technical solution of the present application, the implementation process of the image voiceprint multi-source perception data cross-verification method of the present application will be described in detail below in combination with specific embodiments.

[0062] This embodiment relates to a specific method for cross-verification of multi-source perception data of image and voiceprint, such as Figure 3 As shown, it includes an image voiceprint multi-source perception data cross-verification method based on experience-correlation collaboration, and a cross-verification parameter updating method considering fault recognition accuracy and affirmation. First, the data features of each data channel are extracted, and fault recognition is performed based on the data features; secondly, according to the starting time of the fault data sequence of a certain data channel, the data domain cross-verification sequence of other channels is extracted; then, according to the empirical performance of other channels on the fault recognition accuracy, and the correlation of the data between channels, the data is cross-verified; then, based on the actual maintenance results, the accuracy and affirmation of fault recognition are calculated; finally, the fault recognition algorithm is updated for false negative errors and the channel data correlation is updated based on the channel recognition accuracy and affirmation, and the channel data correlation is updated for false positive errors.

[0063] The details are as follows:

[0064] 1. Image and voiceprint multi-source perception data cross-validation method based on experience-correlation collaboration:

[0065] According to the data features of each data channel of image and voiceprint, fault identification is performed based on the data features, and the fault type, fault probability and corresponding data sequence range are marked in the data sequence of the current observation window, that is, the fault data sequence. For a certain data channel, according to the starting time of its fault data sequence, the data domain cross-verification sequence of other channels is extracted. According to the empirical performance of other channels on the accuracy of fault identification and the correlation of data between channels, the data is cross-verified.

[0066] Step 1: For each data channel of image and voiceprint, extract the data features of each data channel in the current observation time window. Consider T observation time windows in total, and the set is represented by T = {1,...,t,...,T}. Consider N image data channels, and the set is represented by in represents the nth image data channel. Considering M voiceprint data channels, the set is represented as in Represents the mth voiceprint data channel. Definition I n (t) is the image data channel in the tth time window The image data set V m (t) is the voiceprint data channel in the tth time window The feature extraction function is used to extract the data feature set of each data channel in the current observation time window, which is expressed as:

[0067] In the formula, is the image data channel in the tth time window The data feature set; is the voiceprint data channel in the tth time window The data feature set; f ext (·) is the feature extraction function.

[0068] According to the extracted data features, fault identification can be performed. The fault type, fault probability and corresponding data sequence range are marked in the data sequence of the current observation window, that is, the fault data sequence. A total of H types of faults are considered, and the set is expressed as H = {1,...,h,...,H}. The fault identification result is expressed as:

[0069]

[0070] In the formula, and are the image data channels in the tth time window respectively. and voiceprint data channel Fault identification results; are the image data I in the tth time window respectively.n (t) Failure type, failure probability and failure data sequence; are the voiceprint data V in the tth time window respectively m (t) Failure type, failure probability and failure data sequence; f ide (·) is the fault identification function.

[0071] Step 2: For a certain data channel, extract the data domain cross-verification sequence of other channels according to the starting time of its fault data sequence, and define The fault data series The start time, The fault data series The starting time of the fault data sequence is used to extract the data domain cross-verification sequence of other channels, which is expressed as:

[0072] In the formula, is the fault data sequence in the tth time window Start time Extracted image data channels The data domain cross-validation sequence; is the fault data sequence in the tth time window Start time Extracted voiceprint data channel The data domain cross-validation sequence; f ext,cro (·) is the data domain cross-validation sequence extraction function.

[0073] Based on the empirical performance of other channels on fault identification accuracy and the correlation between channel data, cross-check the data to verify the fault type and fault probability. Define the image data channel of the tth time window The cross-verification result of the fault type is It is expressed as:

[0074]

[0075] In the formula, Me{·} means taking the mode; C im,exp is a set of image data expert channels determined by empirical performance. The formula means that when the image data channel If the fault type recognition result of the expert channel set is the same as that of most other image data channels, the original fault type recognition result is maintained; otherwise, the fault type recognition result is re-determined according to the mode of the fault type recognition results of the expert channel set.

[0076] After confirming the fault type, define the image data channel of the tth time window The cross-validation result of the failure probability is It is expressed as:

[0077] In the formula, is the empirical performance of the accuracy of image-based fault detection h; For image data channel The cross-validation weight for fault h; For image data channel Empirical performance of identification accuracy of fault h; For image data channel and The correlation between the data; β 1 and β 2 is the weight coefficient. The formula means that the image data channel The detection result of fault h needs to comprehensively consider the accuracy of fault h detection based on the image. The fault probability of detecting fault h, as well as the situation of other image channels. The more other channels that detect the occurrence of fault h, the higher the fault probability, and the higher the cross-validation result of the probability of occurrence of fault h. In addition, other image channels with higher empirical performance in the accuracy of fault h recognition, and the image data channels with higher Other image channels with higher correlation are in the image data channel has a higher weight in the cross-validation results.

[0078] Similarly, define the voiceprint data channel of the tth time window The cross-verification result of the fault type is It is expressed as:

[0079]

[0080] In the formula, C vo,exp is the set of expert channels of voiceprint data determined by empirical performance. The formula means that when the voiceprint data channel If the fault type recognition result of the expert channel is the same as that of most other voiceprint data channels, the original fault type recognition result is maintained; otherwise, the fault type recognition result is re-determined according to the mode of the fault type recognition results of the expert channel set;

[0081] Define the voiceprint data channel of the tth time window The cross-validation result of the failure probability is It is expressed as:

[0082]

[0083] In the formula, is the empirical performance of the accuracy of fault h detection based on voiceprint; Voiceprint data channel The cross-validation weight for fault h; Voiceprint data channel Empirical performance of identification accuracy of fault h; Voiceprint data channel and The formula means that the voiceprint data channel The detection result of fault h needs to comprehensively consider the accuracy of detecting fault h based on voiceprint. The fault probability of detecting fault h, as well as the situation of other voiceprint channels. The more other channels that detect fault h, the higher the fault probability, and the higher the cross-verification result of the fault probability. In addition, other voiceprint channels with higher empirical performance in the recognition accuracy of fault h, and the voiceprint data channels Other image channels with high correlation are in the voiceprint data channel has a higher weight in the cross-validation results.

[0084] This embodiment proposes a method for cross-verification of multi-source perception data of images and voiceprints based on experience-correlation collaboration. First, for each data channel of the image and voiceprint, extract the data features of each data channel in the current observation window, and perform fault identification based on the data features; second, mark the fault type, fault probability and corresponding fault data sequence in the data sequence of the current observation window; then, for a certain data channel, extract the data domain cross-verification sequence of other channels based on the starting time of its fault data sequence; finally, perform cross-verification of data based on the empirical performance of other channels on fault identification accuracy and the correlation of data between channels to improve the reliability of the cross-verification results.

[0085] 2. Cross-validation parameter updating method considering fault identification accuracy and certainty:

[0086] Step 1: Based on the actual maintenance results, calculate the accuracy and certainty of fault identification. Define the actual maintenance results as This means that the actual inspection result is fault h. In particular, Indicates that there is no fault in the actual maintenance, and defines the image data channel obtained based on the actual maintenance results in the t-th time window The fault identification accuracy is Voiceprint data channel The fault identification accuracy is Respectively expressed as:

[0087]

[0088] The formula means that the fault identification accuracy is related to the fault type identification result of the image data channel or the voiceprint data channel. When the fault type identification result of the image data channel or the voiceprint data channel is the same as the actual inspection result, the fault identification accuracy is 1; otherwise, the fault identification accuracy is 0.

[0089] Define the image data channel obtained based on the actual maintenance results in the tth time window The fault identification certainty is Voiceprint data channel The fault identification certainty is Respectively expressed as:

[0090]

[0091] The formula means that the fault identification certainty is related to the fault type and fault probability identification results of the image data channel or voiceprint data channel. When the actual inspection result is no fault, the fault identification certainty is the interpolation of the fault probability identification result and 1; when the actual inspection result is the same as the fault type identification result, the fault identification certainty is the same as the fault probability identification result; otherwise, that is, when a fault actually occurs and the actual inspection result is different from the fault type identification result, the fault identification certainty is 0.

[0092] Step 2: For false negative errors, that is, a fault actually occurred but was not correctly identified, the existing fault identification algorithm needs to be updated, and the correlation of the channel time data needs to be updated based on the identification accuracy and certainty of the original channel. It can be expressed as:

[0093]

[0094] Where η FN is the correlation update step size in the case of false negative errors; γ∈(0,1) is a constant. Taking the image data channel as an example, the formula means that when the image data channel When a false negative error occurs, if the original channel The fault recognition accuracy is 1, which means the fault is successfully recognized. and The correlation between the data and the original channel The higher the fault identification certainty, the larger the correlation update step size; conversely, if the original channel If the fault recognition accuracy is 0, that is, the fault is not successfully recognized, the fixed step size is reduced. and The correlation between data.

[0095] Step 3: For false positive errors, that is, a fault does not actually occur but is identified as occurring, it is only necessary to update the correlation of the channel time data based on the recognition accuracy and certainty of the original channel, which is expressed as:

[0096]

[0097] Where η FP is the correlation update step size in the case of false positive errors. For example, the formula means that when the voiceprint data channel When a false negative error occurs, if the original channel The fault recognition accuracy is 1, which means the fault is successfully recognized. and The correlation between the data and the original channel The higher the fault identification certainty, the larger the correlation update step; conversely, if the original channel If the fault recognition accuracy is 0, that is, the fault is not successfully recognized, then reduce and The correlation between the data and the original channel The higher the fault identification certainty, the smaller the correlation update step size.

[0098] This embodiment proposes a cross-validation parameter update method that considers the accuracy and certainty of fault identification. First, based on the actual maintenance results, the accuracy and certainty of fault identification are calculated; second, for false negative errors, the existing fault identification algorithm needs to be updated, and at the same time, based on the recognition accuracy and certainty of the original channel, the correlation of the channel time data is updated; finally, for false positive errors, based on the recognition accuracy and certainty of the original channel, the correlation of the channel time data is updated to improve the accuracy and certainty of distribution network equipment fault identification.

[0099] The embodiment of the present application also provides an image voiceprint multi-source perception data cross-verification device. It should be noted that the image voiceprint multi-source perception data cross-verification device of the embodiment of the present application can be used to execute the image voiceprint multi-source perception data cross-verification method provided by the embodiment of the present application. The device is used to implement the above-mentioned embodiments and preferred implementation modes, and the descriptions that have been made will not be repeated. As used below, the term "module" can be a combination of software and / or hardware that implements a predetermined function. Although the device described in the following embodiments is preferably implemented in software, the implementation of hardware, or a combination of software and hardware, is also possible and conceived.

[0100] The following is an introduction to the image and voiceprint multi-source perception data cross-verification device provided in an embodiment of the present application.

[0101] Figure 4Schematic diagram of a device for cross-verification of multi-source perception data of image and voiceprint according to an embodiment of the present application. Figure 4 As shown, the device comprises:

[0102] The first acquisition unit 41 is used to acquire a data feature set of each data channel in the current observation time window, and perform fault identification on the electrical equipment according to the data feature set to generate a fault identification result, wherein the data feature set includes data features of the image data channel and data features of the voiceprint data channel, and the fault identification result includes a fault data sequence of each of the data channels;

[0103] A first determining unit 42 is used to obtain a target fault data sequence of a target data channel from the fault identification result, and determine other data channels related to the target data channel and other fault data sequences corresponding to the other data channels according to the start time of the target fault data sequence;

[0104] The second determination unit 43 is used to determine the cross verification result of the target data channel at least according to the target fault data sequence and the other fault data sequence, and determine the device fault identification result of the target data channel according to the cross verification result.

[0105] In this embodiment, a first acquisition unit is used to acquire a data feature set of each data channel within a current observation time window, and perform fault identification on the electrical equipment according to the above data feature set to generate a fault identification result, wherein the above data feature set includes data features of the image data channel and data features of the voiceprint data channel, and the above fault identification result includes fault data sequences of each of the above data channels; a first determination unit is used to acquire a target fault data sequence of the target data channel from the above fault identification result, and determine other data channels related to the target data channel and other fault data sequences corresponding to the above other data channels according to the start time of the above target fault data sequence; a second determination unit is used to determine a cross-verification result of the above target data channel according to at least the above target fault data sequence and the above other fault data sequences, and determine the equipment fault identification result of the above target data channel according to the above cross-verification result. For each data channel of image and voiceprint, the data features of each data channel in the current observation window are extracted, and fault identification is performed based on the data features. For a certain data channel, the data domain cross-verification sequence of other channels is extracted according to the starting time of its fault data sequence; finally, based on the empirical performance of other channels on the accuracy of fault identification and the correlation of data between channels, the data is cross-verified to improve the reliability of the cross-verification results. This solves the problem of low accuracy of fault identification in the traditional distribution network equipment cross-verification method.

[0106] As an optional solution, the second determining unit includes a first determining module, which is used to determine, when the target data channel is determined to be the voiceprint data channel, according to a second formula:

[0107] Determine the cross-validation results of the above target data channels, where: is the empirical performance based on the accuracy of detecting fault h, For other data channels The cross validation weight for fault h, Voiceprint data channel The empirical performance of the recognition accuracy of fault h, Voiceprint data channel With other data channels The correlation between the data, β 1 and β 2 is the weight coefficient, This is the fault type identification result of the target data channel.

[0108] In an optional solution, the second determining unit includes a second determining module, which is used to determine that the target data channel is the voiceprint data channel according to a second formula:

[0109] Determine the cross-validation results of the above target data channels, where: is the empirical performance based on the accuracy of detecting fault h, For other data channels The cross validation weight for fault h, Voiceprint data channel The empirical performance of the recognition accuracy of fault h, Voiceprint data channel With other data channels The correlation between the data, β 1 and β 2 is the weight coefficient, This is the fault type identification result of the target data channel.

[0110] In an optional solution, the second determination unit further includes a third determination module, which is used to determine the cross verification result of the target data channel according to the first formula before determining the cross verification result of the target data channel according to the third formula:

[0111] Determine the above fault type identification result of the above target data channel Among them, Me{·} means taking the majority, C im,exp Represents a collection of other data channels determined by empirical performance.

[0112] An optional solution, the device also includes a second acquisition unit and a third determination unit; the second acquisition unit is used to obtain the actual maintenance result after determining the equipment fault identification result of the above-mentioned target data channel according to the above-mentioned cross-verification result; the third determination unit is used to determine whether the above-mentioned equipment fault identification result of the above-mentioned target data channel is accurate according to the above-mentioned maintenance result.

[0113] An optional scheme, the device also includes a fourth determination unit and an update unit; the fourth determination unit is used to determine whether the above-mentioned equipment fault identification result of the above-mentioned target data channel is accurate according to the above-mentioned maintenance result, and in the case that the above-mentioned equipment fault identification result of the above-mentioned target data channel is inaccurate, determine the identification error type of the above-mentioned target data channel, wherein the above-mentioned identification error type includes a false negative error and a false positive error, the above-mentioned false negative error indicates that the above-mentioned electrical equipment actually fails but the fault is not correctly identified, and the above-mentioned false negative error indicates that the above-mentioned electrical equipment actually does not fail but is identified as having a fault; the update unit is used to update the fault identification algorithm of the above-mentioned target data channel according to the above-mentioned identification error type.

[0114] An optional solution, the update unit includes a first update module and a second update module; the first update module is used to update the fault identification algorithm when the target data channel is an image data channel and the identification error type is a false negative error, and at the same time, based on the identification accuracy and certainty of the target data channel, update the correlation of the channel time data of the target data channel; the second update module is used to update the correlation of the channel time data of the target data channel based on the identification accuracy and certainty of the target data channel when the target data channel is an image data channel and the identification error type is a false positive error.

[0115] The above-mentioned image voiceprint multi-source perception data cross-verification device includes a processor and a memory. The above-mentioned first acquisition unit, first determination unit and second determination unit are stored in the memory as program units, and the processor executes the above-mentioned program units stored in the memory to realize the corresponding functions. The above-mentioned modules are all located in the same processor; or, the above-mentioned modules are located in different processors in any combination.

[0116] The processor includes a kernel, which calls the corresponding program unit from the memory. One or more kernels can be set, and the problem of low accuracy of fault identification in the traditional distribution network equipment fault identification cross-check method can be solved by adjusting the kernel parameters.

[0117] The memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM, and the memory includes at least one memory chip.

[0118] An embodiment of the present invention provides a computer-readable storage medium, which includes a stored program. When the program is running, the device where the computer-readable storage medium is located is controlled to execute the image voiceprint multi-source perception data cross-verification method.

[0119] Specifically, the image and voiceprint multi-source perception data cross-verification method includes:

[0120] Step S201, obtaining a data feature set of each data channel in the current observation time window, and performing fault identification on the electrical equipment according to the data feature set to generate a fault identification result, wherein the data feature set includes data features of the image data channel and data features of the voiceprint data channel, and the fault identification result includes a fault data sequence of each of the data channels;

[0121] Step S202, obtaining a target fault data sequence of a target data channel from the fault identification result, and determining other data channels related to the target data channel and other fault data sequences corresponding to the other data channels according to the start time of the target fault data sequence;

[0122] Step S203, determining a cross-verification result of the target data channel at least according to the target fault data sequence and the other fault data sequences, and determining a device fault identification result of the target data channel according to the cross-verification result.

[0123] An embodiment of the present invention provides a processor, and the processor is used to run a program, wherein the image and voiceprint multi-source perception data cross-verification method is executed when the program is running.

[0124] Specifically, the image and voiceprint multi-source perception data cross-verification method includes:

[0125] Step S201, obtaining a data feature set of each data channel in the current observation time window, and performing fault identification on the electrical equipment according to the data feature set to generate a fault identification result, wherein the data feature set includes data features of the image data channel and data features of the voiceprint data channel, and the fault identification result includes a fault data sequence of each of the data channels;

[0126] Step S202, obtaining a target fault data sequence of a target data channel from the fault identification result, and determining other data channels related to the target data channel and other fault data sequences corresponding to the other data channels according to the start time of the target fault data sequence;

[0127] Step S203, determining a cross-verification result of the target data channel at least according to the target fault data sequence and the other fault data sequences, and determining a device fault identification result of the target data channel according to the cross-verification result.

[0128] An embodiment of the present invention provides an electronic device, the device comprising a processor, a memory, and a program stored in the memory and executable on the processor, and when the processor executes the program, at least the following steps are implemented:

[0129] Step S201, obtaining a data feature set of each data channel in the current observation time window, and performing fault identification on the electrical equipment according to the data feature set to generate a fault identification result, wherein the data feature set includes data features of the image data channel and data features of the voiceprint data channel, and the fault identification result includes a fault data sequence of each of the data channels;

[0130] Step S202, obtaining a target fault data sequence of a target data channel from the fault identification result, and determining other data channels related to the target data channel and other fault data sequences corresponding to the other data channels according to the start time of the target fault data sequence;

[0131] Step S203, determining a cross-verification result of the target data channel at least according to the target fault data sequence and the other fault data sequences, and determining a device fault identification result of the target data channel according to the cross-verification result.

[0132] The devices in this article can be servers, PCs, PADs, mobile phones, etc.

[0133] The present application also provides a computer program product, which, when executed on a data processing device, is suitable for executing a program for initializing at least the following method steps:

[0134] Step S201, obtaining a data feature set of each data channel in the current observation time window, and performing fault identification on the electrical equipment according to the data feature set to generate a fault identification result, wherein the data feature set includes data features of the image data channel and data features of the voiceprint data channel, and the fault identification result includes a fault data sequence of each of the data channels;

[0135] Step S202, obtaining a target fault data sequence of a target data channel from the fault identification result, and determining other data channels related to the target data channel and other fault data sequences corresponding to the other data channels according to the start time of the target fault data sequence;

[0136] Step S203, determining a cross-verification result of the target data channel at least according to the target fault data sequence and the other fault data sequences, and determining a device fault identification result of the target data channel according to the cross-verification result.

[0137] Obviously, those skilled in the art should understand that the above modules or steps of the present invention can be implemented by a general computing device, they can be concentrated on a single computing device, or distributed on a network composed of multiple computing devices, they can be implemented by a program code executable by a computing device, so that they can be stored in a storage device and executed by the computing device, and in some cases, the steps shown or described can be executed in a different order than here, or they can be made into individual integrated circuit modules, or multiple modules or steps therein can be made into a single integrated circuit module for implementation. Thus, the present invention is not limited to any specific combination of hardware and software.

[0138] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that include computer-usable program code.

[0139] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes 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 generate 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.

[0140] 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.

[0141] 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.

[0142] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0143] The memory may include non-permanent memory 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.

[0144] 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 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.

[0145] 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.

[0146] From the above description, it can be seen that the above embodiments of the present application achieve the following technical effects:

[0147] 1) A method for cross-verification of multi-source perception data of image and voiceprint of the present application includes: obtaining a data feature set of each data channel in the current observation time window, and performing fault identification on the electrical equipment according to the data feature set to generate a fault identification result, wherein the data feature set includes data features of the image data channel and data features of the voiceprint data channel, and the fault identification result includes the fault data sequence of each data channel; obtaining the target fault data sequence of the target data channel from the fault identification result, and determining other data channels related to the target data channel and other fault data sequences corresponding to other data channels according to the start time of the target fault data sequence; determining the cross-verification result of the target data channel according to at least the target fault data sequence and other fault data sequences, and determining the equipment fault identification result of the target data channel according to the cross-verification result. For each data channel of the image and voiceprint, extract the data features of each data channel in the current observation window, and perform fault identification according to the data features. For a certain data channel, extract the data domain cross-verification sequence of other channels according to the start time of its fault data sequence; finally, perform cross-verification of data according to the empirical performance of other channels on fault identification accuracy and the correlation of data between channels to improve the reliability of cross-verification results. The problem of low accuracy in fault identification using the traditional cross-check method for fault identification of distribution network equipment is solved.

[0148] 2) An image and voiceprint multi-source perception data cross-verification device of the present application includes: a first acquisition unit, used to acquire a data feature set of each data channel in a current observation time window, and identify faults of electrical equipment according to the data feature set to generate a fault identification result, wherein the data feature set includes data features of the image data channel and data features of the voiceprint data channel, and the fault identification result includes a fault data sequence of each data channel; a first determination unit, used to acquire a target fault data sequence of a target data channel from the fault identification result, and determine other data channels related to the target data channel and other fault data sequences corresponding to other data channels according to the start time of the target fault data sequence; a second determination unit, used to determine the cross-verification result of the target data channel according to at least the target fault data sequence and other fault data sequences, and determine the equipment fault identification result of the target data channel according to the cross-verification result. For each data channel of image and voiceprint, the data features of each data channel in the current observation window are extracted, and fault identification is performed based on the data features. For a certain data channel, the data domain cross-verification sequence of other channels is extracted according to the starting time of its fault data sequence; finally, based on the empirical performance of other channels on the accuracy of fault identification and the correlation of data between channels, the data is cross-verified to improve the reliability of the cross-verification results. This solves the problem of low accuracy of fault identification in the traditional distribution network equipment cross-verification method.

[0149] The above description is only the preferred 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 modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A method for cross-verification of multi-source perception data of image and voiceprint, characterized in that: include: Acquire a data feature set of each data channel within the current observation time window, and perform fault identification on the electrical equipment according to the data feature set to generate a fault identification result, wherein the data feature set includes data features of the image data channel and data features of the voiceprint data channel, and the fault identification result includes a fault data sequence of each of the data channels; Acquire a target fault data sequence of a target data channel from the fault identification result, and determine other data channels related to the target data channel and other fault data sequences corresponding to the other data channels according to the start time of the target fault data sequence; A cross-verification result of the target data channel is determined at least according to the target fault data sequence and the other fault data sequences, and a device fault identification result of the target data channel is determined according to the cross-verification result.

2. The method according to claim 1, characterized in that Determining a cross-verification result of the target data channel at least according to the target fault data sequence and the other fault data sequences corresponding to the other data channels, comprises: When it is determined that the target data channel is the image data channel, according to the first formula: Determine the cross-check result of the target data channel, wherein: is the empirical performance of image-based fault detection accuracy h, For other data channels The cross validation weight for fault h, For image data channel The empirical performance of the recognition accuracy of fault h, For image data channel With other data channels The correlation between the data, β 1 and β 2 is the weight coefficient, The fault type identification result of the target data channel.

3. The method according to claim 1, characterized in that Determining a cross-verification result of the target data channel at least according to the target fault data sequence and the other fault data sequences corresponding to the other data channels, comprises: When it is determined that the target data channel is the voiceprint data channel, according to the second formula: Determine the cross-check result of the target data channel, wherein: is the empirical performance based on the accuracy of detecting fault h, For other data channels The cross validation weight for fault h, Voiceprint data channel The empirical performance of the recognition accuracy of fault h, Voiceprint data channel With other data channels The correlation between the data, β 1 and β 2 is the weight coefficient, The fault type identification result of the target data channel.

4. The method according to claim 2, characterized in that: Before determining the cross-check result of the target data channel according to the first formula, the method further includes: According to the third formula: Determine the fault type identification result of the target data channel Among them, Me{·} means taking the mode, C im,exp Represents a collection of other data channels determined by empirical performance.

5. The method according to claim 1, characterized in that After determining the device fault identification result of the target data channel according to the cross verification result, the method further includes: Obtain actual maintenance results; Determine whether the device fault identification result of the target data channel is accurate based on the inspection result.

6. The method according to claim 5, characterized in that After determining whether the device fault identification result of the target data channel is accurate according to the inspection result, the method further includes: In the case where the device fault identification result of the target data channel is inaccurate, determining the identification error type of the target data channel, wherein the identification error type includes a false negative error and a false positive error, wherein the false negative error indicates that the electrical device actually fails but the fault is not correctly identified, and the false negative error indicates that the electrical device actually does not fail but is identified as having failed; The fault identification algorithm of the target data channel is updated according to the identification error type.

7. The method according to claim 6, characterized in that Updating the fault identification algorithm of the target data channel according to the identification error type includes: When the target data channel is an image data channel and the identification error type is a false negative error, updating the fault identification algorithm and updating the correlation of the channel time data of the target data channel based on the identification accuracy and affirmation of the target data channel; In a case where the target data channel is an image data channel and the recognition error type is a false positive error, the relevance of the channel time data of the target data channel is updated based on the recognition accuracy and certainty of the target data channel.

8. A device for cross-verification of multi-source perception data of image and voiceprint, characterized in that: include: a first acquisition unit, configured to acquire a data feature set of each data channel within a current observation time window, and perform fault identification on the electrical equipment according to the data feature set to generate a fault identification result, wherein the data feature set includes data features of an image data channel and data features of a voiceprint data channel, and the fault identification result includes a fault data sequence of each of the data channels; A first determining unit, configured to obtain a target fault data sequence of a target data channel from the fault identification result, and determine other data channels related to the target data channel and other fault data sequences corresponding to the other data channels according to a start time of the target fault data sequence; The second determination unit is used to determine the cross-verification result of the target data channel at least according to the target fault data sequence and the other fault data sequences, and determine the device fault identification result of the target data channel according to the cross-verification result.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium includes a stored program, wherein when the program is executed, the device where the computer-readable storage medium is located is controlled to execute the image voiceprint multi-source perception data cross-verification method according to any one of claims 1 to 7.

10. An electronic device, characterized in that: include: One or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and are configured to be executed by the one or more processors, and the one or more programs include a method for executing the image voiceprint multi-source perception data cross-verification method as described in any one of claims 1 to 7.

Images