Fusion diagnosis method for potential multiple faults of aero-engine gas path in transition state

Abstract

The invention discloses a fusion diagnosis method for potential multiple faults of an aero-engine gas path in a transition state, and belongs to the field of aero-engine fault diagnosis. The fusion diagnosis method for the potential multiple faults of the aero-engine gas path in the transition state comprises the steps that collaborative noise reduction and multi-dimensional representation of aero-engine gas path transition state data are studied to obtain multi-dimensional gas path transition state sample data; constructing an aero-engine gas path transition state potential multi-fault fusion diagnosis model, training the diagnosis model by using the multi-dimensional gas path transition state sample data, and outputting a fault diagnosis result by using the trained aero-engine gas path transition state potential multi-fault fusion diagnosis model; the aero-engine gas path transition state potential multi-fault fusion diagnosis model is explained, and a fault mechanism is revealed; accurate mapping and mechanism revealing from limited gas path monitoring information to potential multiple faults are achieved, and theoretical and application support is provided for interpretable intelligent operation and maintenance of an aero-engine.

Description

Technical Field

[0001] The present invention belongs to the field of aircraft engine fault diagnosis, and in particular relates to a method for fusion diagnosis of potential multiple faults in an aircraft engine gas path under a transition state. Background Art

[0002] During the steady-state operation of an aero-engine, indirect engine fault diagnosis can be achieved by mapping the monitoring parameter deviation to the performance change of the gas path components. However, in the transition state of rapid switching of the steady state, the engine speed, power, and aerodynamic and thermal parameters all change dramatically and nonlinearly with time, and the component working line is closer to the fault boundaries such as surge, stall, and idle stop, making it difficult to accurately characterize the transition state conditions based on limited monitoring information. At the same time, the instantaneous measurement interference in the transition state has a greater impact, and the fault mechanism is more complex, which can easily cause multiple faults and lead to functional failure. Domestic and foreign research results have been accumulated in the areas of aero-engine gas path monitoring data acquisition and preprocessing, data-driven aero-engine gas path fault diagnosis, and improvement of the interpretability of fault diagnosis models. However, most of the current research focuses on gas path fault diagnosis under steady-state conditions. In the transition state where the working conditions change rapidly and the monitoring parameters fluctuate dramatically, the gas path monitoring sample construction, potential multiple fault diagnosis, and fault interpretability cognition need to be further studied. Therefore, a potential multiple fault fusion diagnosis method for aero-engine gas path in the transition state is proposed. Summary of the invention

[0003] In view of the deficiencies in the prior art, the purpose of the present invention is to provide a method for fusion diagnosis of potential multiple faults in the air path of an aircraft engine in a transition state, thereby solving the problems in the prior art.

[0004] The purpose of the present invention can be achieved through the following technical solutions:

[0005] The method for fusion diagnosis of multiple potential faults in the gas path of an aircraft engine under a transition state includes the following steps:

[0006] Study the coordinated noise reduction and multi-dimensional characterization of aircraft engine gas path transition state data to obtain multi-dimensional gas path transition state sample data;

[0007] Construct a potential multiple fault fusion diagnosis model for aircraft engine gas path transition state, and use multi-dimensional gas path transition state sample data to train the diagnosis model, and use the trained potential multiple fault fusion diagnosis model for aircraft engine gas path transition state to output fault diagnosis results;

[0008] The potential multiple fault fusion diagnosis model of aircraft engine gas path transition state is explained and the fault mechanism is revealed.

[0009] Furthermore, the process of obtaining multi-dimensional gas path transition state sample data is as follows:

[0010] 1) Input the multi-parameter time series samples with random noise into the multi-head self-attention Seq2Seq model, use the bidirectional gated recurrent unit to encode the input samples of different sizes, and realize feature focusing through the multi-head self-attention mechanism;

[0011] 2) After the multi-head attention bidirectional gated recurrent unit decoding operation, the reconstructed sample is obtained, and the error between the original noise-free sample and the reconstructed sample is used as the loss function to reversely optimize the multi-head self-attention sequence-to-sequence model to obtain the collaborative denoising result of multi-monitoring parameter variable-scale time series samples;

[0012] 3) The nonlinear false nearest neighbor algorithm and average mutual information method are used to select the appropriate embedding dimension and delay time, obtain the delay vector and reconstruct the high-dimensional phase space to construct multi-dimensional gas path transition state sample data.

[0013] Furthermore, the aircraft engine gas path transition state potential multiple fault fusion diagnosis model includes: a potential fault reasoning unit, a multiple fault decoupling unit and a deep meta-transfer learning diagnosis unit;

[0014] Among them, the fault features obtained by processing the acquired multi-dimensional gas path transition state sample data through the feature extractor are respectively input into the sample weight learner based on the multiple fault decoupling unit, the potential fault reasoning unit and the deep meta-transfer learning diagnosis unit. The comprehensive loss function is the weighted sum of the reasoning results, the diagnosis results and the sample weights. The model parameters and sample weights are iteratively learned through the back propagation of the comprehensive loss function to obtain the optimal fusion diagnosis model.

[0015] Furthermore, the process of the potential fault inference unit inferring the potential fault is as follows:

[0016] A probability graph consisting of known gas path monitoring samples, unknown confounding factors, transition state conditions and potential faults is constructed. The inference network is encoded to obtain a multivariate Gaussian distribution, which is then decoded to obtain a model network to infer potential faults based on the conditional probability distribution.

[0017] Furthermore, the multiple fault decoupling unit inputs the fault features obtained by processing the multi-dimensional gas path transition state sample data through the feature extractor into the smoothed Wigner-Ville distribution to extract non-stationary signal features, and then uses random Fourier features to project it into a high-dimensional space to eliminate the correlation of multiple fault coupling features and learn fault weights.

[0018] Furthermore, the deep meta-transfer learning diagnostic unit selects a deep residual convolutional neural network with a unidirectional moving convolution kernel as a learner, and combines it with a meta-learning method that adopts gradient consistency and loss approximation strategies to tune the meta-learner and the base learner; the meta-learner and the base learner are used as the known working condition source domain model and the new working condition target domain model of the transfer learning model, respectively, and the deviation of the features extracted by the same hidden layer of the source domain model and the target domain model for the same sample is calculated, and the hidden layer feature deviation exceeding the threshold is added to the target domain loss function for optimizing the meta-transfer learning model.

[0019] The potential multiple fault fusion diagnosis system of aircraft engine gas path in transition state includes:

[0020] Sample data acquisition module: Research on the coordinated noise reduction and multi-dimensional characterization of aero-engine gas path transition state data to obtain multi-dimensional gas path transition state sample data;

[0021] Diagnostic model building module: Construct a potential multiple fault fusion diagnostic model for aircraft engine gas path transition state, train the diagnostic model using multi-dimensional gas path transition state sample data, and use the trained potential multiple fault fusion diagnostic model for aircraft engine gas path transition state to output fault diagnosis results;

[0022] And, the diagnostic model interpretation module: interprets the potential multiple fault fusion diagnostic model of the aircraft engine gas path transition state and reveals the fault mechanism.

[0023] A computer storage medium stores a readable program, which, when running, can implement the above-mentioned method for fusion diagnosis of potential multiple faults in the air path of an aero-engine in a transitional state.

[0024] An electronic device, comprising: a processor, a memory, a communication interface and a communication bus, wherein the processor, the memory and the communication interface communicate with each other through the communication bus;

[0025] The memory is used to store at least one executable instruction, and the executable instruction enables the processor to execute operations corresponding to the above-mentioned method for fusion diagnosis of potential multiple faults in the air path of an aircraft engine in a transition state.

[0026] A computer program product includes computer instructions, wherein the computer instructions instruct a computing device to execute operations corresponding to the above-mentioned method for fusion diagnosis of potential multiple faults in an aircraft engine gas path under a transition state.

[0027] Beneficial effects of the present invention:

[0028] 1. The present invention realizes gas path transition state information perception by converting aircraft engine gas path monitoring data under low signal-to-noise ratio and poor information background into high-quality multi-dimensional samples that characterize the complex change process of transition state;

[0029] 2. The present invention constructs an accurate fusion diagnosis model for potential multiple faults in the gas path of an aero-engine under transitional state by using counterfactual reasoning of faults under potential causality and decoupling of multiple faults under strong correlation;

[0030] 3. The present invention improves the credibility of the gas path transition state fusion fault diagnosis results through the explainable traceability attribution and graphical cognition of the fault mechanism of the gas path transition state potential multiple fault fusion diagnosis model. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0032] Figure 1 is a fault diagnosis flow chart of the present invention;

[0033] Figure 2 is a flow chart of collaborative denoising and multi-dimensional characterization of transition state data of the present invention;

[0034] Figure 3 It is a framework diagram of the fault diagnosis model of the present invention. DETAILED DESCRIPTION

[0035] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0036] Example 1

[0037] like Figure 1 As shown, the potential multiple fault fusion diagnosis method of the aircraft engine gas path in the transition state includes the following steps:

[0038] S1, study the coordinated noise reduction and multi-dimensional characterization of aircraft engine gas path transition state data to obtain multi-dimensional gas path transition state sample data;

[0039] In view of the difficulties that it is difficult to retain fault information while reducing the noise of transition state gas path data under low signal-to-noise ratio, and that short time series data of limited gas path monitoring parameters are difficult to characterize the state of the transition state system, the present invention studies the coordinated noise reduction and multi-dimensional characterization of aircraft engine gas path transition state data, so as to obtain high-quality multi-dimensional gas path transition state samples for subsequent fault feature extraction and diagnosis model training.

[0040] like Figure 2 As shown in the figure, the process of collaborative denoising of aircraft engine gas path transition state data is as follows: variable-length time series data of finite gas path monitoring parameters in the transition state are collected, and after adding random disturbances, they are input into the multi-head self-attention Seq2Seq model to obtain reconstructed data. The error between the reconstructed data and the undisturbed original data is calculated and reversely optimized to achieve collaborative denoising of variable time series data of multiple monitoring parameters.

[0041] Specifically, the multi-parameter time series samples with added random noise are first input into the multi-head self-attention Seq2Seq model; then the bidirectional gated recurrent unit is used to encode the input samples of different sizes, and feature focusing is achieved through the multi-head self-attention mechanism; the reconstructed samples are obtained through the multi-head attention bidirectional gated recurrent unit decoding operation, and the error between the original noise-free sample and the reconstructed sample is used as the loss function to reversely optimize the multi-head self-attention sequence to sequence model, and the collaborative denoising result of multi-monitoring parameter variable-scale time series samples is obtained.

[0042] The process of dimensional characterization of aircraft engine gas path transition state data is as follows: using nonlinear false neighbor algorithm and average mutual information method to select appropriate embedding dimension and delay time, obtain delay vector and reconstruct high-dimensional phase space. Thus, multi-dimensional gas path transition state samples are constructed on the basis of understanding the nonlinear dynamic characteristics of the gas path, which are used to characterize the complex time-varying laws of gas path transition state monitoring data.

[0043] Specifically, after determining the embedding dimension and delay time, the monitoring data can be mapped to a high-dimensional space, so that samples of size (number of monitoring parameters × embedding dimension) can be obtained at each moment to characterize the nonlinear change characteristics of the gas path transition state.

[0044] S2, constructing a potential multiple fault fusion diagnosis model for the gas path transition state of an aircraft engine, and using the multi-dimensional gas path transition state sample data obtained in S1 to train the diagnosis model, and using the trained potential multiple fault fusion diagnosis model for the gas path transition state of an aircraft engine to output fault diagnosis results;

[0045] like Figure 3As shown in the figure, the potential multiple fault fusion diagnosis model of the transition state of the aircraft engine gas path includes: a potential fault reasoning unit, a multiple fault decoupling unit, and a deep meta-transfer learning diagnosis unit; a probability map consisting of known gas path monitoring samples, unknown confounding factors, transition state conditions and potential faults is constructed, the inference network is encoded to obtain a multivariate Gaussian distribution, and the model network is obtained after decoding, so as to infer the potential fault according to the conditional probability distribution; the transition state fault characteristics are input into the smoothed Wigner-Ville distribution for non-stationary signal feature extraction, and then the random Fourier features are used to project it into a high-dimensional space to eliminate the correlation of multiple fault coupling features and learn the fault weights; combining the potential fault reasoning, multiple fault decoupling and deep meta-transfer learning diagnosis models, the transition state potential multiple fault fusion diagnosis results are obtained by multi-model parameter optimization.

[0046] 1) Potential Fault Reasoning Unit

[0047] The causal variational autoencoder is used to analyze the impact of different fault types on the transition state gas path state, and the transition state counterfactual potential fault reasoning results considering unknown factors are obtained;

[0048] The process of reasoning about potential failures is:

[0049] A probability graph consisting of known gas path monitoring samples, unknown confounding factors, transition state conditions and potential faults is constructed, the inference network is encoded to obtain a multivariate Gaussian distribution, and the model network is obtained after decoding, so as to infer potential faults according to the conditional probability distribution.

[0050] 2) Multiple fault decoupling unit

[0051] The joint time-frequency analysis method is used to extract the non-stationary signal characteristics of the transition state fault characteristics and eliminate the correlation of multiple fault coupling characteristics, so as to accurately distinguish the multiple fault characteristics in the gas path transition state condition;

[0052] The process of multiple fault decoupling is:

[0053] The multi-dimensional gas path transition state sample data obtained in S1 is processed by the feature extractor to obtain the fault features, which are input into the smoothed Wigner-Ville distribution for non-stationary signal feature extraction, and then projected into the high-dimensional space using random Fourier features to eliminate the correlation of multiple fault coupling features and learn the fault weights.

[0054] 3) Deep meta-learning diagnostic unit: A deep residual convolutional neural network with unidirectional convolution kernel movement is selected as the learner, and a meta-learning method using gradient consistency and loss approximation strategy is used to tune the meta-learner and base learner. The meta-learner and base learner are used as the known working condition source domain model and the new working condition target domain model of the transfer learning model, respectively. The deviation of the features extracted by the same hidden layer of the source domain model and the target domain model for the same sample is calculated, and the hidden layer feature deviation exceeding the threshold is added to the target domain loss function to optimize the meta-transfer learning model.

[0055] Combining potential fault reasoning, multiple fault decoupling and deep meta-transfer learning diagnosis units, the transition state potential multiple fault fusion diagnosis results are obtained through multi-unit parameter optimization.

[0056] The steps for training the diagnostic model using the multi-dimensional gas path transition state sample data obtained in S1 are as follows:

[0057] Step 1: The input timing fault samples are weighted by the sample weight learner, and the inference error and diagnosis error are calculated by combining the output results of the potential fault reasoning unit and the deep meta-transfer learning diagnosis unit. These errors are weighted and combined according to the sample weights to generate a comprehensive training error.

[0058] Step 2: Optimize the model through the back-propagation algorithm, including the parameters of the sample weight learner, the potential fault inference model, and the deep meta-transfer learning diagnosis model. The goal is to minimize the comprehensive training error, thereby improving the model's ability to decouple fault modes and reasoning accuracy.

[0059] Step 3: After completing the model training, the results of the potential fault reasoning model and the deep meta-transfer learning diagnosis model are integrated to obtain the final diagnosis result based on the diagnosis contribution of the sample weight distribution, so as to achieve accurate diagnosis of complex multiple faults and efficient adaptability of the model.

[0060] S3, explain the potential multiple fault fusion diagnosis model of aircraft engine gas path transition state and reveal the fault mechanism;

[0061] In response to the pain point of the "black box nature" of the fault fusion diagnosis model and the challenge of difficulty in understanding the model decision-making process and fault mechanism, the present invention studies the interpretation of the aircraft engine gas path transition state fault diagnosis model and the revelation of the fault mechanism, so as to improve the credibility of the gas path transition state fusion diagnosis results.

[0062] 1) Explanation of the fault fusion diagnosis model

[0063] The multi-order game interaction method is used to solve the contribution of transition state features to the fault fusion diagnosis results, and the transition state feature significance map is obtained by combining the transition state features and unit contribution. A hierarchical class activation mapping is constructed to reversely transfer the transition state feature significance map to obtain the importance of each hidden layer in the feature extractor. Through the fusion of multi-layer feature maps, the reverse reconstruction of the input layer is realized based on the unified modeling of feature visualization interpretation and expression capability.

[0064] 2) Revealing the gas path transition state failure mechanism based on the time series knowledge graph

[0065] After obtaining the gas path transition state fusion fault diagnosis results and the corresponding input end key monitoring information, it is necessary to sort out their logical relationships and form an understanding of the fault mechanism. Based on the consideration of the complex temporal relationship of the transition state, the present invention constructs a temporal knowledge graph to integrate the existing monitoring information and fault diagnosis results into entities and embed relationships, so as to dynamically and intuitively understand the fault mechanism.

[0066] Based on similar inventive concepts, an embodiment of the present invention also provides a computer storage medium storing a readable program, which can execute the above-mentioned method for fusion diagnosis of potential multiple faults in the air path of an aircraft engine in a transition state when the program is running.

[0067] Based on similar inventive concepts, an embodiment of the present invention provides an electronic device, comprising: a processor, a memory, a communication interface and a communication bus, wherein the processor, the memory and the communication interface communicate with each other through the communication bus;

[0068] The memory is used to store at least one executable instruction, and the executable instruction enables the processor to execute operations corresponding to the above-mentioned method for fusion diagnosis of potential multiple faults in the air path of an aircraft engine in a transition state.

[0069] Based on similar inventive concepts, an embodiment of the present invention also provides a computer program product, including computer instructions, which instruct a computing device to execute operations corresponding to the above-mentioned method for fusion diagnosis of potential multiple faults in the air path of an aircraft engine in a transition state.

[0070] Example 2

[0071] In this embodiment, a system for fusion diagnosis of multiple potential faults in the gas path of an aircraft engine in a transition state is disclosed, specifically comprising:

[0072] Sample data acquisition module: Research on the coordinated noise reduction and multi-dimensional characterization of aero-engine gas path transition state data to obtain multi-dimensional gas path transition state sample data;

[0073] Diagnostic model building module: Construct a potential multiple fault fusion diagnostic model for aircraft engine gas path transition state, train the diagnostic model using multi-dimensional gas path transition state sample data, and use the trained potential multiple fault fusion diagnostic model for aircraft engine gas path transition state to output fault diagnosis results;

[0074] And, the diagnostic model interpretation module: interprets the potential multiple fault fusion diagnostic model of the aircraft engine gas path transition state and reveals the fault mechanism.

[0075] The method of the present invention may be implemented in hardware, firmware, or as software or computer code that may be stored in a recording medium (such as a CDROM, RAM, floppy disk, hard disk, or magneto-optical disk), or as computer code that is originally stored in a remote recording medium or a non-temporary machine-readable medium downloaded over a network and will be stored in a local recording medium, so that the method described herein may be stored in such software processing on a recording medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware (such as an ASIC or FPGA). It is understood that a computer, processor, microprocessor controller, or programmable hardware includes a storage component (e.g., RAM, ROM, flash memory, etc.) that can store or receive software or computer code, and when the software or computer code is accessed and executed by a computer, processor, or hardware, the method described herein is implemented. In addition, when a general-purpose computer accesses the code for implementing the method shown herein, the execution of the code converts the general-purpose computer into a dedicated computer for executing the method shown herein.

[0076] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments, and the above embodiments and descriptions are only for explaining the principles of the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, and these changes and improvements all fall within the scope of the present invention to be protected.

Claims

1. A method for fusion diagnosis of potential multiple faults in the gas path of an aircraft engine under transitional state, characterized in that: The following steps are involved: Study the coordinated noise reduction and multi-dimensional characterization of aircraft engine gas path transition state data to obtain multi-dimensional gas path transition state sample data; Construct a potential multiple fault fusion diagnosis model for aircraft engine gas path transition state, and use multi-dimensional gas path transition state sample data to train the diagnosis model, and use the trained potential multiple fault fusion diagnosis model for aircraft engine gas path transition state to output fault diagnosis results; The potential multiple fault fusion diagnosis model of aircraft engine gas path transition state is explained and the fault mechanism is revealed.

2. The method for fusion diagnosis of potential multiple faults in the gas path of an aircraft engine in a transient state according to claim 1 is characterized in that: The process of obtaining multi-dimensional gas path transition state sample data is as follows: 1) Input the multi-parameter time series samples with random noise into the multi-head self-attention Seq2Seq model, use the bidirectional gated recurrent unit to encode the input samples of different sizes, and realize feature focusing through the multi-head self-attention mechanism; 2) After the multi-head attention bidirectional gated recurrent unit decoding operation, the reconstructed sample is obtained, and the error between the original noise-free sample and the reconstructed sample is used as the loss function to reversely optimize the multi-head self-attention sequence-to-sequence model to obtain the collaborative denoising result of multi-monitoring parameter variable-scale time series samples; 3) The nonlinear false nearest neighbor algorithm and average mutual information method are used to select the appropriate embedding dimension and delay time, obtain the delay vector and reconstruct the high-dimensional phase space to construct multi-dimensional gas path transition state sample data.

3. The method for fusion diagnosis of potential multiple faults in the gas path of an aircraft engine in a transient state according to claim 1 is characterized in that: The aircraft engine gas path transition state potential multiple fault fusion diagnosis model comprises: a potential fault reasoning unit, a multiple fault decoupling unit and a deep meta-transfer learning diagnosis unit; Among them, the fault features obtained by processing the acquired multi-dimensional gas path transition state sample data through the feature extractor are respectively input into the sample weight learner based on the multiple fault decoupling unit, the potential fault reasoning unit and the deep meta-transfer learning diagnosis unit. The comprehensive loss function is the weighted sum of the reasoning results, the diagnosis results and the sample weights. The model parameters and sample weights are iteratively learned through the back propagation of the comprehensive loss function to obtain the optimal fusion diagnosis model.

4. The method for fusion diagnosis of potential multiple faults in the gas path of an aircraft engine in a transient state according to claim 3 is characterized in that: The process of the potential fault inference unit inferring the potential fault is as follows: A probability graph consisting of known gas path monitoring samples, unknown confounding factors, transition state conditions and potential faults is constructed. The inference network is encoded to obtain a multivariate Gaussian distribution, which is then decoded to obtain a model network to infer potential faults based on the conditional probability distribution.

5. The method for fusion diagnosis of potential multiple faults in the gas path of an aircraft engine in a transient state according to claim 3 is characterized in that: The multiple fault decoupling unit inputs the fault features obtained by processing the multi-dimensional gas path transition state sample data through a feature extractor into a smoothed Wigner-Ville distribution to extract non-stationary signal features, and then uses random Fourier features to project it into a high-dimensional space to eliminate the correlation of multiple fault coupling features and learn fault weights.

6. The method for fusion diagnosis of potential multiple faults in the gas path of an aircraft engine in a transient state according to claim 3 is characterized in that: The deep meta-transfer learning diagnostic unit selects a deep residual convolutional neural network with a unidirectional moving convolution kernel as a learner, and combines it with a meta-learning method using gradient consistency and loss approximation strategies to tune the meta-learner and the base learner; the meta-learner and the base learner are used as the known working condition source domain model and the new working condition target domain model of the transfer learning model respectively, and the deviation of the features extracted by the same hidden layer of the source domain model and the target domain model for the same sample is calculated, and the hidden layer feature deviation exceeding the threshold is added to the target domain loss function to optimize the meta-transfer learning model.

7. A fusion diagnosis system for potential multiple faults in the air path of an aircraft engine in a transitional state, characterized in that: include: Sample data acquisition module: Research on the coordinated noise reduction and multi-dimensional characterization of aero-engine gas path transition state data to obtain multi-dimensional gas path transition state sample data; Diagnostic model building module: Construct a potential multiple fault fusion diagnostic model for aircraft engine gas path transition state, train the diagnostic model using multi-dimensional gas path transition state sample data, and use the trained potential multiple fault fusion diagnostic model for aircraft engine gas path transition state to output fault diagnosis results; And, the diagnostic model interpretation module: interprets the potential multiple fault fusion diagnostic model of the aircraft engine gas path transition state and reveals the fault mechanism.

8. A computer storage medium storing a readable program, characterized in that: When the program is running, it can execute the method for fusion diagnosis of potential multiple faults in the air path of an aircraft engine in a transitional state as described in any one of claims 1 to 6.

9. An electronic device, characterized in that: include: A processor, a memory, a communication interface and a communication bus, wherein the processor, the memory and the communication interface communicate with each other via the communication bus; The memory is used to store at least one executable instruction, and the executable instruction enables the processor to perform operations corresponding to the method for fusion diagnosis of potential multiple faults in the air path of an aircraft engine in a transition state according to any one of claims 1-6.

10. A computer program product comprising computer instructions, characterized in that: The computer instructions instruct the computing device to execute operations corresponding to the method for fusion diagnosis of potential multiple faults in the air path of an aircraft engine in a transitional state as described in any one of claims 1-6.

Images