Aviation equipment failure analysis method and device, electronic equipment and storage medium

Abstract

The aviation equipment failure analysis method and device, electronic device and storage medium provided in the present application relate to the technical field of aviation equipment failure analysis. In the present application, first, sample aviation equipment data is obtained; second, the equipment timing data included in the sample aviation equipment data is subjected to component decoupling processing to obtain at least one sample aviation equipment component data corresponding to the sample aviation equipment data; then, based on the sample aviation equipment data, at least one sample aviation equipment component data and the aviation equipment failure label corresponding to the sample aviation equipment data, the candidate equipment failure analysis model is updated and adjusted to form a target equipment failure analysis model; finally, the target aviation equipment data is analyzed using the target equipment failure analysis model to output the aviation equipment failure analysis data. Based on the above content, the problem of low reliability of aviation equipment failure analysis existing in the prior art can be improved.

Description

Technical Field

[0001] The present application relates to the technical field of aviation equipment failure analysis, and in particular, to an aviation equipment failure analysis method and device, electronic equipment and storage medium. Background Art

[0002] With the development of modern science and technology, the requirements for the safety and reliability of aviation equipment are increasing day by day. The structure, function, load-bearing and service environment of mechanical components are becoming more and more complex, and the subsequent mechanical parts failure is difficult to avoid. Since the failure of mechanical components is often the result of the coupling of multiple factors, the corresponding analysis work is becoming increasingly complex. In order to quickly and accurately diagnose and analyze faults, relying solely on failure analysis experts is far from meeting engineering needs. In the current failure decision-making process, the data is mainly analyzed based on the work experience and professional knowledge of the staff to determine the type and specific cause of the fault, and then formulate corresponding treatment measures. This method relies too much on the experience of personnel, requires a lot of manual participation, is inefficient, and the uncertainty of personnel will affect the quality of the diagnosis results. Therefore, there is a problem of low reliability of aviation equipment failure analysis. Summary of the invention

[0003] In view of this, the purpose of the present application is to provide an aviation equipment failure analysis method and device, electronic equipment and storage medium, so as to improve the problem of low reliability of aviation equipment failure analysis in the prior art.

[0004] To achieve the above objectives, this application adopts the following technical solutions:

[0005] An aviation equipment failure analysis method, comprising:

[0006] Acquire sample aviation equipment data, wherein the sample aviation equipment data includes equipment timing data of the corresponding first aviation equipment, and the equipment timing data is obtained by monitoring the operation process of the first aviation equipment at multiple time points;

[0007] Performing component decoupling processing on the equipment time series data included in the sample aviation equipment data to obtain at least one sample aviation equipment component data corresponding to the sample aviation equipment data;

[0008] Based on the sample aviation equipment data, the at least one sample aviation equipment component data and the aviation equipment failure label corresponding to the sample aviation equipment data, the candidate equipment failure analysis model is updated and adjusted to form a corresponding target equipment failure analysis model, wherein the aviation equipment failure label is used to reflect the failure time and / or failure type of the first aviation equipment, and the candidate equipment failure analysis model belongs to a pre-trained language model;

[0009] The target equipment failure analysis model is used to analyze the target aviation equipment data of the second aviation equipment, and the corresponding aviation equipment failure analysis data is output, wherein the aviation equipment failure analysis data is used to reflect the failure time and / or failure type of the second aviation equipment.

[0010] In a preferred embodiment of the present application, in the above-mentioned aviation equipment failure analysis method, the step of updating and adjusting the candidate equipment failure analysis model based on the sample aviation equipment data, the at least one sample aviation equipment component data and the aviation equipment failure label corresponding to the sample aviation equipment data to form a corresponding target equipment failure analysis model includes:

[0011] Using the candidate equipment failure analysis model, feature mining is performed on the sample aviation equipment data and the at least one sample aviation equipment component data, respectively, and a first equipment feature vector corresponding to the sample aviation equipment data and a second equipment feature vector corresponding to the at least one sample aviation equipment component data are output, wherein the first equipment feature vector and the second equipment feature vector are respectively used to represent semantic information of corresponding data;

[0012] Using the candidate equipment failure analysis model, performing failure analysis based on the first equipment feature vector and the second equipment feature vector, and outputting aviation equipment failure analysis data corresponding to the first aviation equipment, wherein the aviation equipment failure analysis data is used to reflect the failure time and / or failure type of the first aviation equipment;

[0013] Calculating a model training error parameter of the candidate equipment failure analysis model based on the aviation equipment failure analysis data corresponding to the first aviation equipment and the aviation equipment failure label corresponding to the sample aviation equipment data;

[0014] Based on the model training error parameters, the model parameters of the candidate equipment failure analysis model are updated and adjusted to form a corresponding target equipment failure analysis model.

[0015] In a preferred embodiment of the present application, in the above-mentioned aviation equipment failure analysis method, the step of using the candidate equipment failure analysis model to perform feature mining processing on the sample aviation equipment data and the at least one sample aviation equipment component data, and outputting a first equipment feature vector corresponding to the sample aviation equipment data and a second equipment feature vector corresponding to the at least one sample aviation equipment component data, includes:

[0016] Using the candidate equipment failure analysis model, the sample aviation equipment data is subjected to vector space embedding processing and time series information fusion processing to obtain a first equipment feature vector corresponding to the sample aviation equipment data;

[0017] The candidate equipment failure analysis model is used to perform vector space embedding processing on the at least one sample aviation equipment component data to obtain a second equipment feature vector corresponding to the at least one sample aviation equipment component data.

[0018] In a preferred embodiment of the present application, in the above-mentioned aviation equipment failure analysis method, the step of using the candidate equipment failure analysis model to perform vector space embedding processing and time series information fusion processing on the sample aviation equipment data to obtain the first equipment feature vector corresponding to the sample aviation equipment data includes:

[0019] Using the candidate equipment failure analysis model, embedding the equipment time series data included in the sample aviation equipment data in a vector space to obtain a first embedding vector corresponding to the sample aviation equipment data;

[0020] Using the candidate equipment failure analysis model, embedding the position data corresponding to the equipment time series data included in the sample aviation equipment data in a vector space to obtain a second embedding vector corresponding to the sample aviation equipment data, wherein the position data is used to reflect the time series position of the monitoring data corresponding to each time point in the equipment time series data;

[0021] Using the candidate equipment failure analysis model, embedding the time series causal relationship data corresponding to the equipment time series data included in the sample aviation equipment data in a vector space to obtain a third embedding vector corresponding to the sample aviation equipment data, wherein the time series causal relationship data is used to reflect the causal relationship between the monitoring data corresponding to each time point in the equipment time series data;

[0022] The first embedding vector, the second embedding vector and the third embedding vector corresponding to the sample aviation equipment data are superimposed to obtain a first equipment feature vector corresponding to the sample aviation equipment data.

[0023] In a preferred embodiment of the present application, in the above-mentioned aviation equipment failure analysis method, the step of using the candidate equipment failure analysis model to perform failure analysis based on the first equipment feature vector and the second equipment feature vector, and outputting aviation equipment failure analysis data corresponding to the first aviation equipment includes:

[0024] Using the candidate equipment failure analysis model, the first equipment feature vector and the second equipment feature vector are fused to form a corresponding fused equipment feature vector, wherein the semantic information of the sample aviation equipment component data represented by the second equipment feature vector is used as a prefix of the semantic information of the equipment time series data, so as to assist the candidate equipment failure analysis model in understanding the time series semantics of the equipment time series data represented by the first equipment feature vector;

[0025] The candidate equipment failure analysis model is used to perform failure analysis based on the fused equipment feature vector, and aviation equipment failure analysis data corresponding to the first aviation equipment is output.

[0026] In a preferred embodiment of the present application, in the above-mentioned aviation equipment failure analysis method, the step of updating and adjusting the model parameters of the candidate equipment failure analysis model based on the model training error parameter to form a corresponding target equipment failure analysis model includes:

[0027] Based on the model training error parameter, the model parameters of the candidate equipment failure analysis model are updated and adjusted to form a corresponding pending equipment failure analysis model;

[0028] Based on the sample aviation equipment data and / or other sample aviation equipment data, and in combination with corresponding aviation equipment failure labels, the model parameters of the pending equipment failure analysis model are updated and adjusted to form a corresponding target equipment failure analysis model.

[0029] In a preferred embodiment of the present application, in the above-mentioned aviation equipment failure analysis method, the step of updating and adjusting the model parameters of the pending equipment failure analysis model based on the sample aviation equipment data and / or other sample aviation equipment data and in combination with the corresponding aviation equipment failure label to form the corresponding target equipment failure analysis model includes:

[0030] Performing low-rank adaptation and fine-tuning processing on the weight matrix of the pending equipment failure analysis model to obtain an adjusted pending equipment failure analysis model;

[0031] Based on the sample aviation equipment data and / or other sample aviation equipment data, and in combination with corresponding aviation equipment failure labels, the model parameters of the adjusted pending equipment failure analysis model are updated and adjusted to form a corresponding target equipment failure analysis model.

[0032] In a preferred embodiment of the present application, in the above-mentioned aviation equipment failure analysis method, the step of performing component decoupling processing on the equipment timing data included in the sample aviation equipment data to obtain at least one sample aviation equipment component data corresponding to the sample aviation equipment data includes:

[0033] Decoupling processing of failure attribute principal components is performed on equipment time series data included in the sample aviation equipment data to obtain at least one sample aviation equipment component data corresponding to the sample aviation equipment data, wherein the at least one sample aviation equipment component data includes sample aviation equipment trend data, sample aviation equipment seasonal data and / or sample aviation equipment noise data.

[0034] The present application also provides an aviation equipment failure analysis device, comprising:

[0035] A sample data acquisition module, used to acquire sample aviation equipment data, wherein the sample aviation equipment data includes equipment timing data of the corresponding first aviation equipment, and the equipment timing data is obtained by monitoring the operation process of the first aviation equipment at multiple time points;

[0036] A time series component decoupling module, used for performing component decoupling processing on the equipment time series data included in the sample aviation equipment data to obtain at least one sample aviation equipment component data corresponding to the sample aviation equipment data;

[0037] a model updating and adjusting module, configured to update and adjust a candidate equipment failure analysis model based on the sample aviation equipment data, the at least one sample aviation equipment component data, and an aviation equipment failure label corresponding to the sample aviation equipment data, so as to form a corresponding target equipment failure analysis model, wherein the aviation equipment failure label is used to reflect the failure time and / or failure type of the first aviation equipment, and the candidate equipment failure analysis model is a pre-trained language model;

[0038] The equipment failure analysis module is used to analyze the target aviation equipment data of the second aviation equipment using the target equipment failure analysis model, and output corresponding aviation equipment failure analysis data, wherein the aviation equipment failure analysis data is used to reflect the failure time and / or failure type of the second aviation equipment.

[0039] Based on the above, the present application also provides an electronic device, including:

[0040] Memory for storing computer programs;

[0041] The processor connected to the memory is used to execute the computer program stored in the memory to implement the above-mentioned aviation equipment failure analysis method.

[0042] On the basis of the above, the present application also provides a computer-readable storage medium, in which a computer program is stored, and when the computer program is run, each step of the above-mentioned aviation equipment failure analysis method is executed.

[0043] The aviation equipment failure analysis method and device, electronic device and storage medium provided by the present application first obtain sample aviation equipment data; secondly, perform component decoupling processing on the equipment timing data included in the sample aviation equipment data to obtain at least one sample aviation equipment component data corresponding to the sample aviation equipment data; then, based on the sample aviation equipment data, at least one sample aviation equipment component data and the aviation equipment failure label corresponding to the sample aviation equipment data, update and adjust the candidate equipment failure analysis model to form a target equipment failure analysis model; finally, use the target equipment failure analysis model to analyze the target aviation equipment data and output the aviation equipment failure analysis data. Based on the above content, since the equipment timing data will be component decoupled, when updating and adjusting, it can be based not only on the sample aviation equipment data and the corresponding aviation equipment failure label, but also on the sample aviation equipment component data obtained by decoupling, so that the basis can be more sufficient, so that the model can better understand the sample aviation equipment data with the assistance of the sample aviation equipment component data, thereby improving the reliability of failure analysis, and therefore, it can improve the problem of low reliability of aviation equipment failure analysis in the prior art. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] In order to make the above-mentioned objects, features and advantages of the present application more obvious and easy to understand, preferred embodiments are specifically cited below and described in detail with reference to the attached drawings.

[0045] Figure 1 A structural block diagram of an electronic device provided in an embodiment of the present application.

[0046] Figure 2 A schematic flow chart of an aviation equipment failure analysis method provided in an embodiment of the present application.

[0047] Figure 3 A schematic diagram of the embedding process provided in an embodiment of the present application.

[0048] Figure 4 A schematic diagram of the fusion processing provided in an embodiment of the present application.

[0049] Figure 5 A block diagram of an aviation equipment failure analysis device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0050] In order to make the purpose, technical solution and advantages of the embodiments of the present application clearer, 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. The components of the embodiments of the present application described and shown in the drawings here can be arranged and designed in various different configurations.

[0051] Therefore, the following detailed description of the embodiments of the present application provided in the accompanying drawings is not intended to limit the scope of the present application for which protection is sought, but merely represents selected embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in the field without creative work are within the scope of protection of the present application.

[0052] like Figure 1 As shown, an embodiment of the present application provides an electronic device, wherein the electronic device may include a memory, a processor, and an aviation equipment failure analysis device.

[0053] In detail, the memory and the processor are electrically connected directly or indirectly to achieve data transmission or interaction. For example, the memory and the processor can be electrically connected through one or more communication buses or signal lines. The aviation equipment failure analysis device includes at least one software function module stored in the memory in the form of software or firmware. The processor is used to execute an executable computer program stored in the memory, for example, the software function modules and computer programs included in the aviation equipment failure analysis device, so as to implement the aviation equipment failure analysis method provided in the embodiment of the present application.

[0054] Optionally, the memory may be, but is not limited to, a random access memory (RAM), a read only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable read-only memory (EEPROM), etc.

[0055] Furthermore, the processor may be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), a system on chip (SoC), etc.; it may also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components.

[0056] Understandably, Figure 1The structure shown is for illustration only, and the electronic device may also include Figure 1 More or fewer components as shown, or with Figure 1 The different configurations shown, for example, may also include a communication unit for exchanging information with other devices (such as sensors, etc.).

[0057] Combination Figure 2 The present application also provides an aviation equipment failure analysis method applicable to the above electronic device, wherein the method steps defined in the process related to the aviation equipment failure analysis method can be implemented by the electronic device.

[0058] The following will Figure 2 The specific process shown is explained in detail.

[0059] Step S110, obtaining sample aviation equipment data.

[0060] In an embodiment of the present application, the electronic device can obtain sample aviation equipment data. Exemplarily, the specific number of sample aviation equipment data obtained is not limited. Among them, the sample aviation equipment data includes equipment timing data of the corresponding first aviation equipment, and the equipment timing data is obtained by monitoring the operation process of the first aviation equipment at multiple time points. Exemplarily, the equipment timing data can include data of multiple dimensions, such as equipment timing data of the first dimension, equipment timing data of the second dimension, equipment timing data of the third dimension, etc. The specific dimensions can include but are not limited to temperature, vibration, stress, pressure, etc. In addition, in some other implementations, the sample aviation equipment data can also include basic information of the first aviation equipment, such as model information, service time information, and material information.

[0061] Step S120 , performing component decoupling processing on the equipment time series data included in the sample aviation equipment data to obtain at least one sample aviation equipment component data corresponding to the sample aviation equipment data.

[0062] In an embodiment of the present application, after acquiring the sample aviation equipment data, the electronic device may perform component decoupling processing on the equipment time series data included in the sample aviation equipment data to obtain at least one sample aviation equipment component data corresponding to the sample aviation equipment data, that is, extract at least one component from the equipment time series data to obtain at least one corresponding sample aviation equipment component data. In addition, it should be noted that when the equipment time series data includes data of multiple dimensions, component decoupling processing may be performed on the data of each dimension respectively to obtain at least one sample aviation equipment component data corresponding to the data of each dimension.

[0063] Step S130 , based on the sample aviation equipment data, the at least one sample aviation equipment component data and the aviation equipment failure label corresponding to the sample aviation equipment data, the candidate equipment failure analysis model is updated and adjusted to form a corresponding target equipment failure analysis model.

[0064] In an embodiment of the present application, after obtaining at least one sample aviation equipment component data, the electronic device can update and adjust the candidate equipment failure analysis model based on the sample aviation equipment data, the at least one sample aviation equipment component data and the aviation equipment failure label corresponding to the sample aviation equipment data to form a corresponding target equipment failure analysis model. Wherein, the aviation equipment failure label is used to reflect the failure time and / or failure type (i.e., the actual failure situation, such as engine blades, bearings, turbine blades and other components are broken or cracked), and the candidate equipment failure analysis model belongs to a pre-trained language model, such as a large language model, which can process large-scale text data and has strong language understanding and generation capabilities. Exemplarily, the candidate equipment failure analysis model includes but is not limited to the GPT-3 model (Generative Pre-trained Transformer 3, a generative pre-trained model based on the Transformer architecture), the BERT model (Bidirectional Encoder Representations from Transformers, a bidirectional Transformer encoder model), and the Megatron model (a large-scale Transformer language model, mainly used to process very large text corpora).

[0065] Step S140 , using the target equipment failure analysis model, analyzing the target aviation equipment data of the second aviation equipment, and outputting corresponding aviation equipment failure analysis data.

[0066] In the embodiment of the present application, after forming the target equipment failure analysis model, the electronic device can use the target equipment failure analysis model to analyze the target aviation equipment data of the second aviation equipment and output the corresponding aviation equipment failure analysis data. The aviation equipment failure analysis data is used to reflect the failure time and / or failure type of the second aviation equipment.

[0067] Based on the above content, since the equipment timing data will be decoupled into components, when updating and adjusting, it is possible not only to rely on the sample aviation equipment data and the corresponding aviation equipment failure labels, but also to rely on the decoupled sample aviation equipment component data, so that the basis can be more sufficient, and the model can be assisted by the sample aviation equipment component data. Better understanding of the sample aviation equipment data, thereby improving the reliability of failure analysis, therefore, can improve the problem of low reliability of aviation equipment failure analysis in the prior art.

[0068] It should be noted that for step S120, the specific method of performing component decoupling processing on the equipment time series data included in the sample aviation equipment data is not limited and can be selected according to actual needs.

[0069] For example, in an alternative implementation, in order to obtain sufficient and reliable sample aviation equipment component data, the above step S120 may include the following specific contents:

[0070] Decoupling processing of failure attribute principal components is performed on equipment time series data included in the sample aviation equipment data to obtain at least one sample aviation equipment component data corresponding to the sample aviation equipment data, wherein the at least one sample aviation equipment component data includes sample aviation equipment trend data, sample aviation equipment seasonal data and / or sample aviation equipment noise data.

[0071] That is to say, characteristic information such as trend (Trend: the part describing the long-term trend of data changes), season (Seasonality: the part describing the periodic change of data, usually a repetitive pattern related to time), and noise (Noise: random component, reflecting random fluctuations or anomalies in the data) can be extracted from the equipment time series data. For example, for aircraft monitoring data, parameters such as temperature, pressure, and vibration can be decomposed according to time series to identify different patterns that show trend changes, seasonal fluctuations, and random noise over time. This decomposition helps to more accurately capture signals related to aircraft system failures.

[0072] It should be noted that for step S130 , the specific method of updating and adjusting the candidate equipment failure analysis model is not limited and can be selected according to actual needs.

[0073] For example, in an alternative implementation, in order to improve the analysis accuracy of the target timing analysis model formed by updating and adjusting, the above-mentioned step S130 may further include step S131, step S132, step S133 and step S134, and the specific content of each step is described as follows.

[0074] Step S131, using the candidate equipment failure analysis model, respectively perform feature mining processing on the sample aviation equipment data and the at least one sample aviation equipment component data, and output a first equipment feature vector corresponding to the sample aviation equipment data and a second equipment feature vector corresponding to the at least one sample aviation equipment component data.

[0075] In an embodiment of the present application, the candidate equipment failure analysis model can be used to perform feature mining processing on the sample aviation equipment data and the at least one sample aviation equipment component data, respectively, and output a first equipment feature vector corresponding to the sample aviation equipment data and a second equipment feature vector corresponding to the at least one sample aviation equipment component data. The first equipment feature vector and the second equipment feature vector are respectively used to characterize the semantic information of the corresponding data, for example, the first equipment feature vector is used to characterize the semantic information of the sample aviation equipment data, and the second equipment feature vector is used to characterize the semantic information of the sample aviation equipment component data.

[0076] Step S132: using the candidate equipment failure analysis model, performing failure analysis based on the first equipment feature vector and the second equipment feature vector, and outputting aviation equipment failure analysis data corresponding to the first aviation equipment.

[0077] In the embodiment of the present application, after obtaining the first equipment feature vector and the second equipment feature vector, the candidate equipment failure analysis model can be used to perform failure analysis based on the first equipment feature vector and the second equipment feature vector, and output the aviation equipment failure analysis data corresponding to the first aviation equipment, that is, the semantic information of the first equipment feature vector and the second equipment feature vector can be combined to analyze and predict the failure of the first aviation equipment. Wherein, the aviation equipment failure analysis data is used to reflect the failure time and / or failure type of the first aviation equipment. In addition, in some other embodiments, since the sample aviation equipment data can also include basic information of the first aviation equipment, such as model information, service time information and material information, the candidate equipment failure analysis model can also be used to perform feature mining on the model information, service time information and material information, so as to obtain the corresponding third equipment feature vector. In this way, when performing failure analysis, the third equipment feature vector can also be combined, that is, the first equipment feature vector, the second equipment feature vector and the third equipment feature vector are fused, and then, based on the fused output vector, the failure of the first aviation equipment is analyzed and predicted. In this way, the characterization ability of the final fused output vector can be improved to a certain extent, so that the accuracy of the analysis and prediction is higher.

[0078] Step S133: Calculate the model training error parameter of the candidate equipment failure analysis model based on the aviation equipment failure analysis data corresponding to the first aviation equipment and the aviation equipment failure label corresponding to the sample aviation equipment data.

[0079] In an embodiment of the present application, after obtaining the aviation equipment failure analysis data corresponding to the first aviation equipment, the model training error parameter of the candidate equipment failure analysis model can be calculated based on the aviation equipment failure analysis data corresponding to the first aviation equipment and the aviation equipment failure label (i.e., the actual failure situation) corresponding to the sample aviation equipment data. It can be understood that the specific method of calculating the error is not limited. For example, for the error corresponding to the failure time, the mean square error (MSE) can be used for calculation, and for the error corresponding to the failure type, the cross-entropy loss (Cross-Entropy Loss) can be used for calculation.

[0080] Step S134: based on the model training error parameter, the model parameters of the candidate equipment failure analysis model are updated and adjusted to form a corresponding target equipment failure analysis model.

[0081] In an embodiment of the present application, after obtaining the model training error parameter, the model parameters of the candidate equipment failure analysis model can be updated and adjusted based on the model training error parameter to form a corresponding target equipment failure analysis model. For example, the model parameters of the candidate equipment failure analysis model can be updated and adjusted in the direction of reducing the model training error parameter until the model training error parameter is reduced to a preset value or the reduction is less than the preset value. In this way, the corresponding update and adjustment can be completed to obtain the target equipment failure analysis model.

[0082] It can be understood that, in the above step S131, the specific manner of performing feature mining processing on the sample aviation equipment data and the at least one sample aviation equipment component data is not limited and can be selected according to actual needs.

[0083] For example, in an alternative implementation, in order to improve the reliability of feature mining, the above-mentioned step S131 may further include step S131a and step S131b, and the specific content of each step is described as follows.

[0084] Step S131a, using the candidate equipment failure analysis model, performs vector space embedding processing and time series information fusion processing on the sample aviation equipment data to obtain a first equipment feature vector corresponding to the sample aviation equipment data.

[0085] In an embodiment of the present application, the candidate equipment failure analysis model can be used to embed the sample aviation equipment data in the vector space and fuse the timing information to obtain the first equipment feature vector corresponding to the sample aviation equipment data. That is to say, not only is it necessary to embed the equipment timing data included in the sample aviation equipment data into the corresponding vector space, but it is also necessary to fuse the timing semantics in the equipment timing data so that the representation capability of the first equipment feature vector is better.

[0086] Step S131b: using the candidate equipment failure analysis model, embedding the at least one sample aviation equipment component data in a vector space to obtain a second equipment feature vector corresponding to the at least one sample aviation equipment component data.

[0087] In an embodiment of the present application, the candidate equipment failure analysis model can also be used to perform vector space embedding processing on the at least one sample aviation equipment component data to obtain a second equipment feature vector corresponding to the at least one sample aviation equipment component data. In other words, each sample aviation equipment component data can be embedded in the corresponding vector space to obtain the second equipment feature vector corresponding to the at least one sample aviation equipment component data.

[0088] It can be understood that, in the above step S131a, the specific method of performing the embedding process of the vector space and the fusion process of the time series information is not limited and can be selected according to actual needs.

[0089] For example, in an alternative implementation, in order to improve the reliability of the embedding process of the vector space and the fusion process of the time series information, Figure 3 As shown, the above step S131a may further include the following specific implementation contents:

[0090] First, the candidate equipment failure analysis model can be used to perform embedding processing in a vector space on the equipment time series data included in the sample aviation equipment data (such as first performing word segmentation processing and then performing word embedding processing) to obtain a first embedding vector corresponding to the sample aviation equipment data;

[0091] Secondly, the candidate equipment failure analysis model can be used to perform vector space embedding processing on the position data corresponding to the equipment time series data included in the sample aviation equipment data to obtain a second embedding vector corresponding to the sample aviation equipment data, wherein the position data is used to reflect the time series position of the monitoring data corresponding to each time point in the equipment time series data;

[0092] Then, the candidate equipment failure analysis model can be used to perform vector space embedding processing on the time series causal relationship data corresponding to the equipment time series data included in the sample aviation equipment data to obtain a third embedding vector corresponding to the sample aviation equipment data, wherein the time series causal relationship data is used to reflect the causal relationship between the monitoring data corresponding to each time point in the equipment time series data, for example, the state at the previous time point will affect the state at the next time point, but the state at the next time point will not affect the state at the previous time point. Thus, in an alternative implementation, the time series causal relationship can be an influence transfer relationship from front to back between the monitoring data corresponding to each time point in the equipment time series data;

[0093] Finally, the first embedding vector, the second embedding vector and the third embedding vector corresponding to the sample aviation equipment data can be superimposed to obtain the first equipment feature vector corresponding to the sample aviation equipment data; it should be noted that the equipment timing data included in the sample aviation equipment data can be of multiple dimensions. In this way, after the embedding processing of the vector space is performed separately, the local embedding vectors corresponding to each dimension can be aligned (for example, the local embedding vectors corresponding to each dimension can be linearly mapped or nonlinearly mapped so that they can be mapped to the same space, thereby achieving alignment between vectors), thereby forming an aligned first embedding vector, and then superimposed with the second embedding vector and the third embedding vector.

[0094] It can be understood that, in the above step S132, the specific method of performing failure analysis based on the first device characteristic vector and the second device characteristic vector is not limited and can be selected according to actual needs.

[0095] For example, in an alternative implementation, in order to further improve the reliability of failure analysis and thereby obtain aviation equipment failure analysis data with higher accuracy, Figure 4 , the above step S132 may further include the following specific implementation contents:

[0096] First, the candidate equipment failure analysis model can be used to fuse the first equipment feature vector and the second equipment feature vector to form a corresponding fused equipment feature vector, wherein the semantic information of the sample aviation equipment component data represented by the second equipment feature vector is used as a prefix (i.e., a prompt header) of the semantic information of the equipment time series data to assist the candidate equipment failure analysis model in understanding the time series semantics of the equipment time series data represented by the first equipment feature vector. That is to say, the first equipment feature vector and the second equipment feature vector can be spliced ​​to obtain a corresponding fused equipment feature vector;

[0097] Secondly, the candidate equipment failure analysis model can be used to perform failure analysis based on the fused equipment feature vector, and output the aviation equipment failure analysis data corresponding to the first aviation equipment; illustratively, the fused equipment feature vector can be convolved, pooled, activated, and fully connected, and then predicted and output based on the corresponding prediction function (such as softmax function, etc.), so as to obtain the aviation equipment failure analysis data corresponding to the first aviation equipment. In addition, after obtaining the fused equipment feature vector, the third equipment feature vector mentioned above can also be fused.

[0098] It can be understood that, in the above step S134, the specific method of updating and adjusting the model parameters of the candidate equipment failure analysis model is not limited and can be selected according to actual needs.

[0099] For example, in an alternative embodiment, in order to further improve the reliability of model update adjustment so that the obtained target equipment failure analysis model can have better analysis capabilities, the above-mentioned step S134 can further include step S134a and step S134b, and the specific content of each step is described as follows.

[0100] Step S134a, based on the model training error parameter, the model parameters of the candidate equipment failure analysis model are updated and adjusted to form a corresponding pending equipment failure analysis model.

[0101] In an embodiment of the present application, the model parameters of the candidate equipment failure analysis model can be updated and adjusted based on the model training error parameters to form a corresponding pending equipment failure analysis model, such as updating and adjusting the model parameters of the candidate equipment failure analysis model in the direction of reducing the model training error parameters until the model training error parameters converge.

[0102] Step S134b, based on the sample aviation equipment data and / or other sample aviation equipment data, and in combination with the corresponding aviation equipment failure labels, the model parameters of the pending equipment failure analysis model are updated and adjusted to form a corresponding target equipment failure analysis model.

[0103] In an embodiment of the present application, after obtaining the pending equipment failure analysis model, the model parameters of the pending equipment failure analysis model can be updated and adjusted again based on the sample aviation equipment data and / or other sample aviation equipment data and in combination with the corresponding aviation equipment failure labels to form a corresponding target equipment failure analysis model. For example, the sample data for the update and adjustment can be different, or the adjustment method for the update and adjustment can be different, so that the target equipment failure analysis model can have better generalization ability.

[0104] It can be understood that, in the above step S134b, the specific method of updating and adjusting the model parameters of the pending equipment failure analysis model is not limited and can be selected according to actual needs.

[0105] For example, in an alternative implementation, in order to further improve the reliability of the update adjustment, the above step S134b may further include the following specific implementation contents:

[0106] First, the weight matrix of the pending equipment failure analysis model can be subjected to low-rank adaptive fine-tuning processing to obtain an adjusted pending equipment failure analysis model; that is, the weight matrix included in the adjusted pending equipment failure analysis model is a weight matrix subjected to low-rank adaptive fine-tuning processing;

[0107] Secondly, based on the sample aviation equipment data and / or other sample aviation equipment data, and in combination with the corresponding aviation equipment failure labels, the model parameters of the adjusted pending equipment failure analysis model can be updated and adjusted to form a corresponding target equipment failure analysis model. The specific update and adjustment method can refer to the aforementioned related steps.

[0108] For example, in the GPT-3 model, the complexity and computational cost of the model can be reduced by low-rank adaptation of the attention heads in the attention mechanism. For example, assuming that the attention heads of the GPT-3 model need to be low-rank adapted, the specific steps are as follows:

[0109] Singular value decomposition (SVD) application: Singular value decomposition is performed on the weight matrix of the attention head to decompose it into smaller sub-matrices;

[0110] Select the main singular values: retain the main singular values ​​to reduce the rank of the original weight matrix;

[0111] Reconstruct the attention head: Use the new low-rank adapted weight matrix to reconstruct the attention head to replace the original high-rank weight matrix;

[0112] Fine-tuning the model: Fine-tune the entire GPT-3 model to minimize the loss function under the new low-rank adaptation setting to ensure optimal model performance.

[0113] For example, in the BERT model, the self-attention mechanism and feedforward neural network in its multi-layer Transformer encoder can be low-rank adapted. For example, assuming that the Transformer encoder of the BERT model is to be low-rank adapted, the specific steps are as follows:

[0114] Singular Value Decomposition (SVD) application: Singular value decomposition of the self-attention mechanism in each Transformer encoder and the weight matrix of the feedforward neural network;

[0115] Keep important singular values: Select and keep the main singular values ​​in each level to reduce the rank of the weight matrix;

[0116] Reconstruct encoder layer: Reconstruct each Transformer encoder layer with the new low-rank adapted weight matrix;

[0117] Global fine-tuning: Globally fine-tune the entire BERT model to adapt to the new low-rank adaptation parameters while maintaining the model’s language understanding and representation capabilities.

[0118] For example, in the Megatron model, the efficiency and performance of the model can be optimized by performing low-rank adaptation on its multi-layer Transformer structure. For example, assuming that the Transformer structure in the Megatron model needs to be low-rank adapted, the specific steps are as follows:

[0119] Singular Value Decomposition (SVD) application: Singular value decomposition of the attention mechanism in each Transformer layer and the weight matrix of the feedforward neural network;

[0120] Select key singular values: Pick and retain the main singular values ​​to reduce the rank of the weight matrix at each level;

[0121] Update Transformer layer: Update the parameters of each Transformer layer using the new weight matrix after low-rank adaptation;

[0122] Overall model optimization: The entire Megatron model is adjusted through a global fine-tuning process to ensure that efficient language representation and reasoning capabilities are maintained while reducing computational costs.

[0123] Through the above method, low-rank adaptation fine-tuning technology can be effectively used to optimize large models such as GPT-3, BERT and Megatron to reduce the model's computing and storage resource consumption without sacrificing performance.

[0124] In addition, the weight matrix in the model is subjected to singular value decomposition, which can be decomposed into the product form of three matrices: [W = U \Sigma V^T];

[0125] Here, (U) and (V) are orthogonal matrices, (\Sigma) is a diagonal matrix containing the singular values ​​of the original matrix, and by retaining the first few main singular values, the weight matrix (W) can be approximated by a lower rank matrix, such as the new weight matrix (U'\Sigma' V'^T).

[0126] After updating and adjusting the model based on the above steps, the following steps may be included when analyzing and predicting failures of actual aviation equipment data:

[0127] Step 1: For the constructed large model, the basic information that needs to be inferred (time series information, as well as non-decision information such as aircraft model, service time, and materials) can be used as input, and the input sequence is processed by time series embedding, and absolute position embedding (such as the embedding process of the vector space of the position data mentioned above), time series causal embedding (such as the embedding process of the vector space of the time series causal relationship data mentioned above), and multi-parameter alignment embedding (such as the embedding process of the vector space of the equipment time series data mentioned above and the alignment of the local embedding vectors corresponding to each dimension) are constructed respectively; then, information fusion is performed by accumulation to ensure that the time series information is accurately encoded and represented;

[0128] Step 2: Decouple the original sequence (i.e., equipment time series data) into main components to highlight the failure attribute information, such as decomposing it into main components such as trend, season, and noise. The main component is used as a prompt and added to the beginning of the sequence obtained in step 1 to form a specific mark.

[0129] Step 3: Use the result of step 2 as the input of the large model to complete the reasoning (i.e., analytical prediction) of the failure type or failure time.

[0130] Combination Figure 5 The present application also provides an aviation equipment failure analysis device applicable to the above electronic equipment. The aviation equipment failure analysis device may include a sample data acquisition module, a time series component decoupling module, a model update adjustment module, and an equipment failure analysis module.

[0131] In detail, the sample data acquisition module can be used to acquire sample aviation equipment data, wherein the sample aviation equipment data includes equipment timing data of the corresponding first aviation equipment, and the equipment timing data is obtained by monitoring the operation process of the first aviation equipment at multiple time points. Figure 2 As shown in step S110, the relevant contents of the sample data acquisition module can refer to the above description of step S110.

[0132] In detail, the timing component decoupling module can be used to perform component decoupling processing on the equipment timing data included in the sample aviation equipment data to obtain at least one sample aviation equipment component data corresponding to the sample aviation equipment data. Figure 2 As shown in step S120, the relevant contents of the timing component decoupling module can refer to the above description of step S120.

[0133] In detail, the model update adjustment module can be used to update and adjust the candidate equipment failure analysis model based on the sample aviation equipment data, the at least one sample aviation equipment component data and the aviation equipment failure label corresponding to the sample aviation equipment data to form a corresponding target equipment failure analysis model, wherein the aviation equipment failure label is used to reflect the failure time and / or failure type of the first aviation equipment, and the candidate equipment failure analysis model belongs to a pre-trained language model. In the embodiment of the present application, the model update adjustment module can be used to perform Figure 2 As shown in step S130, the relevant contents of the model updating and adjusting module can refer to the above description of step S130.

[0134] In detail, the equipment failure analysis module is used to analyze the target aviation equipment data of the second aviation equipment using the target equipment failure analysis model, and output corresponding aviation equipment failure analysis data, wherein the aviation equipment failure analysis data is used to reflect the failure time and / or failure type of the second aviation equipment. In the embodiment of the present application, the equipment failure analysis module can be used to perform Figure 2 As shown in step S140, the relevant contents of the equipment failure analysis module can refer to the above description of step S140.

[0135] In the embodiment of the present application, corresponding to the above-mentioned aviation equipment failure analysis method applied to the electronic device, a computer-readable storage medium is also provided, in which a computer program is stored, and when the computer program is run, each step of the aviation equipment failure analysis method is executed. Among them, each step executed when the aforementioned computer program is run will not be described one by one here, and reference can be made to the above explanation of the aviation equipment failure analysis method.

[0136] In summary, the aviation equipment failure analysis method and device, electronic device and storage medium provided by the present application first obtain sample aviation equipment data; secondly, perform component decoupling processing on the equipment timing data included in the sample aviation equipment data to obtain at least one sample aviation equipment component data corresponding to the sample aviation equipment data; then, based on the sample aviation equipment data, at least one sample aviation equipment component data and the aviation equipment failure label corresponding to the sample aviation equipment data, the candidate equipment failure analysis model is updated and adjusted to form a target equipment failure analysis model; finally, the target aviation equipment data is analyzed using the target equipment failure analysis model to output the aviation equipment failure analysis data. Based on the above content, since the equipment timing data is component decoupled, when updating and adjusting, it can be based not only on the sample aviation equipment data and the corresponding aviation equipment failure label, but also on the sample aviation equipment component data obtained by decoupling, so that the basis can be more sufficient, so that the model can better understand the sample aviation equipment data with the assistance of the sample aviation equipment component data, thereby improving the reliability of failure analysis, and therefore, it can improve the problem of low reliability of aviation equipment failure analysis in the prior art.

[0137] In several embodiments provided in the embodiments of the present application, it should be understood that the disclosed device and method can also be implemented in other ways. The device and method embodiments described above are merely schematic, for example, the flowcharts and block diagrams in the accompanying drawings show the possible architecture, functions and operations of the device, method and computer program product according to multiple embodiments of the present application. In this regard, each box in the flowchart or block diagram can represent a module, a program segment or a part of a code, and the module, a program segment or a part of a code contains one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a different order from the order marked in the accompanying drawings. For example, two consecutive boxes can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram and / or flowchart, and the combination of boxes in the block diagram and / or flowchart can be implemented with a dedicated hardware-based system that performs a specified function or action, or can be implemented with a combination of dedicated hardware and computer instructions.

[0138] In addition, the functional modules in the various embodiments of the present application may be integrated together to form an independent part, or each module may exist separately, or two or more modules may be integrated to form an independent part.

[0139] If the function is implemented in the form of a software function module and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application is essentially or the part that contributes to the prior art or the part of the technical solution can be embodied in the form of a software product, which is stored in a storage medium, including several instructions to enable a computer device (which can be a personal computer, electronic device, or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk and other media that can store program code. It should be noted that in this article, the term "include", "include" or any other variant thereof is intended to cover non-exclusive inclusion, so that the process, method, article or device including a series of elements includes not only those elements, but also includes other elements that are not explicitly listed, or also includes elements inherent to such process, method, article or device. Without more constraints, an element defined by the phrase "comprising a..." does not exclude the existence of other identical elements in the process, method, article or apparatus comprising the element.

[0140] 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 analyzing failure of aviation equipment, characterized in that: include: Acquire sample aviation equipment data, wherein the sample aviation equipment data includes equipment timing data of the corresponding first aviation equipment, and the equipment timing data is obtained by monitoring the operation process of the first aviation equipment at multiple time points; Performing component decoupling processing on the equipment time series data included in the sample aviation equipment data to obtain at least one sample aviation equipment component data corresponding to the sample aviation equipment data; Based on the sample aviation equipment data, the at least one sample aviation equipment component data and the aviation equipment failure label corresponding to the sample aviation equipment data, the candidate equipment failure analysis model is updated and adjusted to form a corresponding target equipment failure analysis model, wherein the aviation equipment failure label is used to reflect the failure time and / or failure type of the first aviation equipment, and the candidate equipment failure analysis model belongs to a pre-trained language model; Analyzing the target aviation equipment data of the second aviation equipment by using the target equipment failure analysis model, and outputting corresponding aviation equipment failure analysis data, wherein the aviation equipment failure analysis data is used to reflect the failure time and / or failure type of the second aviation equipment; The step of updating and adjusting the candidate equipment failure analysis model based on the sample aviation equipment data, the at least one sample aviation equipment component data and the aviation equipment failure label corresponding to the sample aviation equipment data to form a corresponding target equipment failure analysis model includes: Using the candidate equipment failure analysis model, feature mining is performed on the sample aviation equipment data and the at least one sample aviation equipment component data, respectively, and a first equipment feature vector corresponding to the sample aviation equipment data and a second equipment feature vector corresponding to the at least one sample aviation equipment component data are output, wherein the first equipment feature vector and the second equipment feature vector are respectively used to represent semantic information of corresponding data; Using the candidate equipment failure analysis model, performing failure analysis based on the first equipment feature vector and the second equipment feature vector, and outputting aviation equipment failure analysis data corresponding to the first aviation equipment, wherein the aviation equipment failure analysis data is used to reflect the failure time and / or failure type of the first aviation equipment; Calculating a model training error parameter of the candidate equipment failure analysis model based on the aviation equipment failure analysis data corresponding to the first aviation equipment and the aviation equipment failure label corresponding to the sample aviation equipment data; Based on the model training error parameters, the model parameters of the candidate equipment failure analysis model are updated and adjusted to form a corresponding target equipment failure analysis model.

2. The aviation equipment failure analysis method according to claim 1, characterized in that: The step of using the candidate equipment failure analysis model to perform feature mining processing on the sample aviation equipment data and the at least one sample aviation equipment component data, and outputting a first equipment feature vector corresponding to the sample aviation equipment data and a second equipment feature vector corresponding to the at least one sample aviation equipment component data, comprises: Using the candidate equipment failure analysis model, the sample aviation equipment data is subjected to vector space embedding processing and time series information fusion processing to obtain a first equipment feature vector corresponding to the sample aviation equipment data; The candidate equipment failure analysis model is used to perform vector space embedding processing on the at least one sample aviation equipment component data to obtain a second equipment feature vector corresponding to the at least one sample aviation equipment component data.

3. The aviation equipment failure analysis method according to claim 2, characterized in that: The step of using the candidate equipment failure analysis model to perform vector space embedding processing and time series information fusion processing on the sample aviation equipment data to obtain a first equipment feature vector corresponding to the sample aviation equipment data includes: Using the candidate equipment failure analysis model, embedding the equipment time series data included in the sample aviation equipment data in a vector space to obtain a first embedding vector corresponding to the sample aviation equipment data; Using the candidate equipment failure analysis model, embedding the position data corresponding to the equipment time series data included in the sample aviation equipment data in a vector space to obtain a second embedding vector corresponding to the sample aviation equipment data, wherein the position data is used to reflect the time series position of the monitoring data corresponding to each time point in the equipment time series data; Using the candidate equipment failure analysis model, embedding the time series causal relationship data corresponding to the equipment time series data included in the sample aviation equipment data in a vector space to obtain a third embedding vector corresponding to the sample aviation equipment data, wherein the time series causal relationship data is used to reflect the causal relationship between the monitoring data corresponding to each time point in the equipment time series data; The first embedding vector, the second embedding vector and the third embedding vector corresponding to the sample aviation equipment data are superimposed to obtain a first equipment feature vector corresponding to the sample aviation equipment data.

4. The aviation equipment failure analysis method according to claim 1, characterized in that: The step of using the candidate equipment failure analysis model to perform failure analysis based on the first equipment feature vector and the second equipment feature vector, and outputting aviation equipment failure analysis data corresponding to the first aviation equipment, comprises: Using the candidate equipment failure analysis model, the first equipment feature vector and the second equipment feature vector are fused to form a corresponding fused equipment feature vector, wherein the semantic information of the sample aviation equipment component data represented by the second equipment feature vector is used as a prefix of the semantic information of the equipment time series data, so as to assist the candidate equipment failure analysis model in understanding the time series semantics of the equipment time series data represented by the first equipment feature vector; The candidate equipment failure analysis model is used to perform failure analysis based on the fused equipment feature vector, and aviation equipment failure analysis data corresponding to the first aviation equipment is output.

5. The aviation equipment failure analysis method according to claim 1, characterized in that: The step of updating and adjusting the model parameters of the candidate equipment failure analysis model based on the model training error parameter to form a corresponding target equipment failure analysis model includes: Based on the model training error parameter, the model parameters of the candidate equipment failure analysis model are updated and adjusted to form a corresponding pending equipment failure analysis model; Based on the sample aviation equipment data and / or other sample aviation equipment data, and in combination with corresponding aviation equipment failure labels, the model parameters of the pending equipment failure analysis model are updated and adjusted to form a corresponding target equipment failure analysis model.

6. The aviation equipment failure analysis method according to claim 5, characterized in that: The step of updating and adjusting the model parameters of the pending equipment failure analysis model based on the sample aviation equipment data and / or other sample aviation equipment data and in combination with the corresponding aviation equipment failure label to form the corresponding target equipment failure analysis model includes: Performing low-rank adaptation and fine-tuning processing on the weight matrix of the pending equipment failure analysis model to obtain an adjusted pending equipment failure analysis model; Based on the sample aviation equipment data and / or other sample aviation equipment data, and in combination with corresponding aviation equipment failure labels, the model parameters of the adjusted pending equipment failure analysis model are updated and adjusted to form a corresponding target equipment failure analysis model.

7. The aviation equipment failure analysis method according to any one of claims 1 to 6, characterized in that: The step of performing component decoupling processing on the equipment time series data included in the sample aviation equipment data to obtain at least one sample aviation equipment component data corresponding to the sample aviation equipment data includes: Decoupling processing of failure attribute principal components is performed on equipment time series data included in the sample aviation equipment data to obtain at least one sample aviation equipment component data corresponding to the sample aviation equipment data, wherein the at least one sample aviation equipment component data includes sample aviation equipment trend data, sample aviation equipment seasonal data and / or sample aviation equipment noise data.

8. An aviation equipment failure analysis device, characterized in that: include: A sample data acquisition module, used to acquire sample aviation equipment data, wherein the sample aviation equipment data includes equipment timing data of the corresponding first aviation equipment, and the equipment timing data is obtained by monitoring the operation process of the first aviation equipment at multiple time points; A time series component decoupling module, used for performing component decoupling processing on the equipment time series data included in the sample aviation equipment data to obtain at least one sample aviation equipment component data corresponding to the sample aviation equipment data; a model updating and adjusting module, configured to update and adjust a candidate equipment failure analysis model based on the sample aviation equipment data, the at least one sample aviation equipment component data, and an aviation equipment failure label corresponding to the sample aviation equipment data, so as to form a corresponding target equipment failure analysis model, wherein the aviation equipment failure label is used to reflect the failure time and / or failure type of the first aviation equipment, and the candidate equipment failure analysis model is a pre-trained language model; an equipment failure analysis module, used to analyze the target aviation equipment data of the second aviation equipment by using the target equipment failure analysis model, and output corresponding aviation equipment failure analysis data, wherein the aviation equipment failure analysis data is used to reflect the failure time and / or failure type of the second aviation equipment; Wherein, the updating and adjusting of the candidate equipment failure analysis model based on the sample aviation equipment data, the at least one sample aviation equipment component data and the aviation equipment failure label corresponding to the sample aviation equipment data to form the corresponding target equipment failure analysis model includes: Using the candidate equipment failure analysis model, feature mining is performed on the sample aviation equipment data and the at least one sample aviation equipment component data, respectively, and a first equipment feature vector corresponding to the sample aviation equipment data and a second equipment feature vector corresponding to the at least one sample aviation equipment component data are output, wherein the first equipment feature vector and the second equipment feature vector are respectively used to represent semantic information of corresponding data; Using the candidate equipment failure analysis model, performing failure analysis based on the first equipment feature vector and the second equipment feature vector, and outputting aviation equipment failure analysis data corresponding to the first aviation equipment, wherein the aviation equipment failure analysis data is used to reflect the failure time and / or failure type of the first aviation equipment; Calculating a model training error parameter of the candidate equipment failure analysis model based on the aviation equipment failure analysis data corresponding to the first aviation equipment and the aviation equipment failure label corresponding to the sample aviation equipment data; Based on the model training error parameters, the model parameters of the candidate equipment failure analysis model are updated and adjusted to form a corresponding target equipment failure analysis model.

9. An electronic device, characterized in that: include: Memory for storing computer programs; A processor connected to the memory is used to execute the computer program stored in the memory to implement the aviation equipment failure analysis method described in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, which, when executed, executes the aviation equipment failure analysis method according to any one of claims 1 to 7.

Images