A method, equipment, and medium for fault detection and isolation of complex spacecraft equipment.

Abstract

This application discloses a method, device, and medium for fault detection and isolation of complex spacecraft equipment, relating to the field of fault identification. The method includes: inputting the current operating status data of the complex spacecraft equipment into an operating status prediction model for prediction to obtain the future operating status data distribution; the operating status prediction model is obtained by training a variational Koopman anomaly detector using a time-series training sample set of the complex spacecraft equipment; and calculating the prediction error matrix and the corresponding T based on the current operating status data and the future operating status data distribution. 2 The system calculates the values ​​and SPE statistics, and then determines the operating state of complex spacecraft equipment based on preset limits. When the operating state indicates a fault, it calculates the deviation score of each operating state variable and isolates the fault to determine the fault variables of the operating state. This application realizes effective fault detection and fault variable screening for complex spacecraft equipment.

Description

Technical Field

[0001] This application relates to the field of fault identification technology, and in particular to a method, device and medium for fault detection and isolation of complex spacecraft equipment. Background Technology

[0002] Effective fault detection and isolation (FDI) is crucial for monitoring time-series data of complex spacecraft equipment, ensuring not only the safety of personnel and equipment but also maintaining stable product quality. The primary goal of fault detection is to identify anomalous behavior, while fault isolation involves identifying the variables most likely to cause failure. However, due to the inherent complexity of spacecraft equipment time-series data, accurately modeling the system using expert knowledge to directly achieve fault detection and isolation remains a significant challenge. Therefore, developing advanced automated FDI systems to support operators in managing anomalies is essential. In recent years, data mining techniques have made significant progress, enabling the extraction of information from collected data and facilitating fault detection and isolation.

[0003] Typical data-driven multivariate statistical methods include principal component analysis, partial least squares regression, and independent component analysis, often used for fault detection and isolation in time series data of complex spacecraft equipment. However, these methods assume that process variables are linear and normal, which limits their applicability to nonlinear real-world process data. To address the limitations of traditional methods, deep learning methods have been introduced for fault detection and isolation. These models learn data representations through nonlinear activation functions and complex network architectures, effectively handling complex nonlinear relationships. However, these deep learning models do not model dynamic systems. Furthermore, due to their complexity, deep learning models are often considered "black boxes," making it difficult to explain their internal processes and decision-making mechanisms. Summary of the Invention

[0004] The purpose of this application is to provide a method, device and medium for fault detection and isolation of complex spacecraft equipment, which realizes effective fault detection and fault variable screening of complex spacecraft equipment.

[0005] To achieve the above objectives, this application provides the following solution:

[0006] Firstly, this application provides a method for fault detection and isolation of complex spacecraft equipment, including:

[0007] Acquire current operational status data of complex equipment on a spacecraft; the current operational status data includes multiple operational status variables;

[0008] The current operating status data is input into the operating status prediction model for prediction to obtain the future operating status data distribution; wherein, the operating status prediction model is obtained by training the variational Koopman anomaly detector using a time series training sample set of complex spacecraft equipment; the time series training sample set includes multiple operating status data within a preset time period; the variational Koopman anomaly detector is determined based on a variational autoencoder and a Koopman matrix layer;

[0009] Based on the current operating status data and the distribution of the future operating status data, calculate the prediction error matrix and the corresponding T. 2 Value, SPE statistic;

[0010] Based on preset limits, and according to T corresponding to the current operating status data 2 The operating status of the complex equipment on the spacecraft is determined by the value and SPE statistic; the operating status is either a fault occurred or no fault occurred.

[0011] When the working state is in the event of a fault, the deviation score of each operating state variable is calculated and the fault is isolated based on the current operating state data and the distribution of future operating state data, so as to determine the operating state fault variable.

[0012] Secondly, this application provides a computer device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement a method for fault detection and isolation of complex equipment in a spacecraft.

[0013] Thirdly, this application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements a method for fault detection and isolation of complex equipment in a spacecraft.

[0014] According to the specific embodiments provided in this application, this application achieves the following technical effects: This application provides a method, device, and medium for fault detection and isolation of complex spacecraft equipment. It combines variational autoencoders and Koopman operator theory to obtain a variational Koopman anomaly detector, which is then trained to predict the distribution of future operational state data. Because it combines variational autoencoders and Koopman operator theory, it relies entirely on data-driven approaches and does not require expert knowledge in related fields. Furthermore, the variational Koopman anomaly detector overcomes the dependence of traditional methods on the assumptions of linearity and normality of process variables, as well as the difficulties in interpreting internal processes and decision-making mechanisms in deep learning methods. The uncertainty estimation it generates can be used for fault detection in time-series data of complex spacecraft equipment. Furthermore, based on the deviation score set by the uncertainty output of the variational Koopman anomaly detector, the degree of deviation of each variable after a fault can be defined, thereby intuitively and effectively screening fault-related variables and enabling fault isolation. This application achieves effective fault identification and fault isolation variable screening in time-series data of complex spacecraft equipment. Attached Figure Description

[0015] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0016] Figure 1 This is a flowchart illustrating a method for fault detection and isolation of complex spacecraft equipment, provided as an embodiment of this application.

[0017] Figure 2 This is a schematic diagram illustrating the calculation of the training loss function provided in one embodiment of this application.

[0018] Figure 3 This is a flowchart illustrating a method for fault detection and isolation of complex spacecraft equipment, provided as another embodiment of this application.

[0019] Figure 4 This is a schematic diagram of the structure of a computer device provided in an embodiment of this application. Detailed Implementation

[0020] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0021] This application utilizes time-series data of complex spacecraft equipment for fault detection and fault variable screening, representing a data-driven method that does not require expertise in related fields. In other words, this application achieves fault detection and isolation, addressing the problem that most current fault identification methods focus solely on fault detection. Using this method, the variables causing the fault can be identified promptly, effectively preventing the risk of fault propagation.

[0022] To make the above-mentioned objectives, features and advantages of this application more apparent and understandable, the application will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0023] In one exemplary embodiment, such as Figure 1 As shown, a fault detection and isolation method for complex equipment in a spacecraft is provided. This method is executed by computer equipment, specifically by a terminal or server alone, or by both a terminal and a server. In this embodiment, it includes the following steps 101 to 105.

[0024] Step 101: Obtain the current operating status data of complex equipment on the spacecraft; the current operating status data includes multiple operating status variables.

[0025] Step 102: Input the current operating status data into the operating status prediction model for prediction to obtain the future operating status data distribution; wherein, the operating status prediction model is obtained by training the Variational Koopman Anomaly Detector (VKAD) using a time series training sample set of complex spacecraft equipment; the time series training sample set includes multiple operating status data within a preset time period; the Variational Koopman Anomaly Detector is determined based on a variational autoencoder and a Koopman matrix layer.

[0026] The variational Koopman anomaly detector comprises, in sequence, an input layer, an encoding layer of a variational autoencoder, a Koopman matrix layer, a sampling layer, a decoding layer of the variational autoencoder, and an output layer. The target of the variational Koopman anomaly detector is the signal passing through the encoding layer of the variational autoencoder (corresponding to...). Figure 2 The q-net network in (a) encodes the observable distribution of the current running state data, then propagates the observable distribution forward using an approximate linear operator, and subsequently passes it through the decoding layer of the variational autoencoder (corresponding to...). Figure 2 The p-net network in (a) maps observables back to the state space in the form of a probability distribution, thereby enabling uncertainty prediction.

[0027] In the time-series training sample set, the operating state corresponding to each operational state sample data is a fault-free state. That is, this application uses time-series data under normal operation (fault-free) to train the variational Koopman anomaly detector, enabling it to accurately predict the distribution of future data, learn the mode corresponding to normal operation, and thus achieve fault detection of time-series data of complex spacecraft equipment. Obviously, the training of the variational Koopman anomaly detector in this application is unsupervised training.

[0028] In another embodiment, after training is complete, the running state prediction model can be tested using a time series test sample set. If a large deviation occurs after the test sample data is input into the running state prediction model (because it deviates from the normal mode), it can be considered as a failure.

[0029] The future operating state data predicted by the operating state prediction model follows a multivariate Gaussian distribution and is represented as a predicted probability distribution. For example, the input to the operating state prediction model is a certain time step x(t), which corresponds to the probability prediction (u(t+1), σ) for the next time step x(t+1). 2 (t+1)).

[0030] During the training of the variational Koopman anomaly detector, the training loss function is calculated using the following formula:

[0031]

[0032] Where α1, α2, and α3 are hyperparameters, representing the proportion of each corresponding individual loss function to the total loss function; φ is the parameter of the coding layer of the variational autoencoder, θ is the parameter of the decoding layer of the variational autoencoder, and K is the Kuman operator; For mapping loss, such as Figure 2 The formula for calculating (b) is as follows:

[0033]

[0034] in, For expectations; Let be the approximate posterior distribution of the coding layer output of the variational autoencoder at time t. Let x be the running status data at time t. t The corresponding observable; The output of the decoding layer of the variational autoencoder is x t The corresponding probability distribution, D KL Let KL divergence be the KL divergence. To and The corresponding prior distribution, where L is the number of running state data in the time series training sample set. To obtain from the approximate posterior distribution The l-th observable drawn from the sample. It is an approximate conditional probability distribution; H is the dimension of the function space, which is specified by the designer; and The standard deviation and mean of the h-th dimension of the observable distribution at time t are respectively ∈ (l) To obtain from the standard Gaussian distribution The l-th noise vector extracted from the sample is represented by the element-wise multiplication of ⊙.

[0035] For dynamic loss, this is an additional constraint that ensures the dynamics of the observables are represented linearly in the function space and controlled by matrix K. This effectively defines a linear Kumann operator K to advance the distribution of the observables over time, such as... Figure 2 The formula for calculating (c) dynamic loss is shown below:

[0036]

[0037] in, Let be the approximate posterior distribution of the coding layer output of the variational autoencoder at time t+m. x is the running status data at time t+m. t+m The corresponding observable, K m To recursively propagate the Koopman operator m times, To and The corresponding k-th sample, and are the standard deviation and mean of the h-th dimension of the observable distribution at time t+m, respectively, where m is the time step; ‖·‖ represents the mean square error, averaged over the dimension.

[0038] To predict loss, it is necessary to predict future states, specifically representing the predicted future states as a probability distribution, such as... Figure 2 As shown in (a) below, the formula for calculating the prediction loss is as follows:

[0039]

[0040] in, For x t+m The corresponding approximate posterior probability distribution, This is an approximation of the sampling.

[0041] Step 103: Based on the current operating status data and the distribution of the future operating status data, calculate the prediction error matrix and the corresponding T. 2 Values, SPE statistics, including:

[0042] The prediction error matrix is ​​calculated using the following formula: Among them, e t Let x be the prediction error matrix at time t. t Let μ be the running status data at time t. t Let t be the mean of the observable distribution at time t, where the observable is the predicted distribution of future operating state data.

[0043] Calculate T using the following formula. 2 value: in,() T P is the transpose of A; P is the load matrix, and A is a diagonal matrix consisting of the first k eigenvalues.

[0044] The SPE statistic is calculated using the following formula: Where I is the identity matrix.

[0045] Step 104: Based on the preset limit, according to T corresponding to the current operating status data. 2 The operating status of the complex equipment of the spacecraft is determined by the value and SPE statistic; the operating status is either a fault or no fault.

[0046] In one application example, the preset limit includes a first limit and a second limit; the first limit and the second limit are obtained by training the principal component analysis model based on the prediction error matrix corresponding to each running state data in the time series training sample set after the variational Koopman anomaly detector has been trained.

[0047] The formula for calculating the first limit is:

[0048]

[0049] Among them, T α The first limit is k, the number of principal components retained in the principal component analysis model is k, 1-α is the confidence level, and N is the sample size; F α (k,Nk) follows an F distribution with degrees of freedom k and Nk.

[0050] The formula for calculating the second limit is:

[0051]

[0052] Among them, SPE α The second limit is given, where m and v are the mean and variance of SPE, respectively. Obeying a degree of freedom of 2m 2 / v is a chi-square distribution.

[0053] Based on preset limits, and according to T corresponding to the current operating status data 2The value and SPE statistic determine the operating status of the complex equipment of the spacecraft, including: when the current operating status data corresponds to T 2 When the value is greater than the first limit or the SPE statistic is greater than the second limit, the working status of the complex equipment of the spacecraft is marked as a malfunction.

[0054] Step 105: When the working state is a fault, calculate the deviation score of each operating state variable based on the current operating state data and the distribution of the future operating state data, and perform fault isolation to determine the operating state fault variable.

[0055] The formula for calculating the deviation score of each operating state variable is as follows:

[0056]

[0057] in, Let be the deviation score of the s-th variable in the running status data at time t. Let be the s-th variable in the running state data at time t. and Let be the mean and standard deviation of the predicted probability distribution of the s-th variable, respectively, and |·| be the absolute value. It is a positive value. The magnitude of the value determines the degree to which the variable is affected.

[0058] In one application instance, fault isolation is performed to identify operational status fault variables, including: after obtaining the deviation scores of each operational status variable, drawing an offset degree map based on the deviation scores of each operational status variable to achieve fault isolation; and marking operational status variables with an offset degree greater than a preset degree value as operational status fault variables.

[0059] In another exemplary embodiment of this application, such as Figure 3As shown, the process is divided into two parts: offline modeling and online monitoring. In the offline modeling process, normal process data is first collected and regularized to obtain a time-series training sample set. Regularization is unrelated to the observable distribution; it's simply a data preprocessing step for the input model, primarily performing a minimization-maximization operation, which is beneficial for model training and testing. During this process, the collected time-series training sample set is stored in an M x N table format. The first row and first column of the table are empty. The time-series training sample set includes N-1 variables, stored in the first row of the table excluding empty spaces. The time-series training sample set includes M-1 time points, stored in the first column of the table excluding empty spaces. The time-series training sample set includes (M-1)*(N-1) data points, stored in the table excluding the first row and first column. The first row of the label column stores the label variables, and the other rows store the data points collected at different time points for the label variables. The label column can be any column other than the first column, and the label variables are arbitrary variables. Then, the variational Koopman anomaly detector is trained to predict the data sample distribution at future time points, and the prediction error is calculated. After obtaining the prediction error matrix containing all normal samples, the PCA model is trained to determine T. 2 And SPE control limits.

[0060] In the online monitoring process, the data samples to be monitored are first collected and regularized. Then, the variational Koopman anomaly detector, trained during the offline modeling process, is used to predict the distribution of data samples at future time points based on the data samples at the current time point. Secondly, the prediction error is calculated using the predicted mean and the actual sample value, and then T is used. 2 The SPE statistic is used to monitor whether the prediction error exceeds the control limit, thereby determining whether a fault has occurred. If a fault occurs, the predicted mean and standard deviation of the model output are used to set a deviation score to define the degree of deviation of each variable after the fault, and a deviation graph of each variable is plotted to achieve fault isolation and thus determine the fault-related variables.

[0061] In summary, this application addresses the issues of dependence on the linearity and normality assumptions of process variables and the difficulties in interpreting internal processes and decision-making mechanisms using deep learning methods. It studies a fault detection and isolation model based on Koopman operator theory, which is applicable to complex nonlinear time-series data of spacecraft equipment and provides good interpretability. Secondly, to address the problem that traditional Koopman operator theory does not consider uncertainty, a variational Koopman anomaly detector (VKAD) is studied for fault detection and isolation of uncertain time-series data of complex spacecraft equipment. Finally, the mean and standard deviation output of the VKAD model help establish a bias score, and based on the bias score, a graph reflecting the degree of deviation of each scalar is plotted, thereby achieving fault isolation.

[0062] This application's method utilizes the proposed Variational Koopman Anomaly Detector (VKAD) to detect and isolate faults in time-series data of complex spacecraft equipment. VKAD can accurately capture the dynamic characteristics of complex time series and predict the data distribution at future time points based on data samples at the current time point. After obtaining the predicted distribution from VKAD, the data is then processed via T... 2 The SPE monitoring statistic is used to assess the deviation between the predicted mean and the true sample. If the deviation exceeds T... 2 If the control limits of SPE are not met, the sample is identified as a fault. After a fault is detected, the predicted mean and standard deviation of the model output are used to set a deviation score to define the degree of deviation of each variable after the fault, and a deviation graph of each variable is plotted to achieve fault isolation, thereby determining the candidate set of fault-related variables.

[0063] In one exemplary embodiment, a computer device is provided, which may be a server or a terminal, and its internal structure diagram may be as follows. Figure 4 As shown, this computer device includes a processor, memory, input / output (I / O) interfaces, and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is also connected to the system bus via the I / O interfaces. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and databases. The internal memory provides the environment for the operating system and computer programs stored in the non-volatile storage media to run. The I / O interfaces are used for exchanging information between the processor and external devices. The communication interface is used for communicating with external terminals via a network connection. When the computer program is executed by the processor, it implements a fault detection and isolation method for complex spacecraft equipment.

[0064] Those skilled in the art will understand that Figure 4 The structures shown are merely block diagrams of some structures related to the present application and do not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than shown in the figures, or combine certain components, or have different component arrangements. In an exemplary embodiment, a computer device is provided, including a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the steps in the above-described method embodiments.

[0065] In one exemplary embodiment, a computer-readable storage medium is provided storing a computer program that, when executed by a processor, implements the steps in the above-described method embodiments.

[0066] In one exemplary embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps in the above-described method embodiments.

[0067] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of the relevant data must comply with relevant regulations.

[0068] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM).

[0069] The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0070] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0071] This document uses specific examples to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the methods and core ideas of this application. Furthermore, those skilled in the art will recognize that, based on the ideas of this application, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A method for fault detection and isolation of complex spacecraft equipment, characterized in that, The fault detection and isolation methods for complex spacecraft equipment include: Acquire current operational status data of complex equipment on a spacecraft; the current operational status data includes multiple operational status variables; The current operating status data is input into the operating status prediction model for prediction to obtain the future operating status data distribution; wherein, the operating status prediction model is obtained by training the variational Koopman anomaly detector using a time series training sample set of complex spacecraft equipment; the time series training sample set includes multiple operating status data within a preset time period; the variational Koopman anomaly detector is determined based on a variational autoencoder and a Koopman matrix layer; The variational Koopman anomaly detector comprises an input layer, an encoding layer of a variational autoencoder, a Koopman matrix layer, a sampling layer, a decoding layer of a variational autoencoder, and an output layer arranged sequentially. In the time series training sample set, the working state corresponding to each running state sample data is no fault. The distribution of future running state data predicted by the running state prediction model follows a multivariate Gaussian distribution and is represented in the form of a predicted probability distribution. During the training of the variational Koopman anomaly detector, the training loss function is calculated using the following formula: ； in, , and For hyperparameters; For the parameters of the coding layer of the variational autoencoder, These are the parameters of the decoding layer of the variational autoencoder. K For the Kuman operator; The mapping loss is calculated using the following formula: ； in, For expectations; Let be the approximate posterior distribution of the coding layer output of the variational autoencoder at time t. The running status data at time t The corresponding observable; The output of the decoding layer of the variational autoencoder and The corresponding probability distribution D KL Let KL divergence be the KL divergence. p ( ) for and The corresponding prior distribution, L The number of running state data in the time series training sample set. To obtain from the approximate posterior distribution The first one drawn from An observable measurement It is an approximate conditional probability distribution; H For the dimension of the function space, and The observable distribution at time t, respectively The standard deviation and mean of the dimension, To obtain from the standard Gaussian distribution The first one drawn from l A noise vector, Multiply the elements one by one; The dynamic loss is calculated using the following formula: ； in, Let be the approximate posterior distribution of the coding layer output of the variational autoencoder at time t+m. The running status data at time t+m The corresponding observable, K m To recursively propagate the Koopman operator m times, To and The corresponding number Second sampling, and The first of the observable distributions at time t+m is the [missing information]. The standard deviation and mean of the dimension, For time steps; This represents the mean squared error, averaged over the dimension; To predict the loss, the calculation formula is as follows: ； in, To and The corresponding approximate posterior probability distribution, This is an approximation of the sampling. Based on the current operating status data and the distribution of the future operating status data, calculate the prediction error matrix and its corresponding... Value, SPE statistic; Based on preset limits, and according to the current operating status data... The operating status of the complex equipment on the spacecraft is determined by the value and SPE statistic; the operating status is either a fault occurred or no fault occurred. When the working state is in the event of a fault, the deviation score of each operating state variable is calculated and the fault is isolated based on the current operating state data and the distribution of future operating state data, so as to determine the operating state fault variable.

2. The fault detection and isolation method for complex spacecraft equipment according to claim 1, characterized in that, The distribution of future operating state data predicted by the operating state prediction model follows a multivariate Gaussian distribution and is represented in the form of a prediction probability distribution; based on the current operating state data and the distribution of future operating state data, a prediction error matrix and its corresponding... Values, SPE statistics, including: The prediction error matrix is ​​calculated using the following formula: ;in, Let be the prediction error matrix at time t. The data represents the running status at time t. Let t be the mean of the observable distribution at time t, where the observable is the predicted distribution of future operating state data; Calculate according to the following formula value: ;in,( ) T For transpose; For the load matrix, For the reason before A diagonal matrix composed of eigenvalues; The SPE statistic is calculated using the following formula: .

3. The fault detection and isolation method for complex spacecraft equipment according to claim 1, characterized in that, The preset limit includes a first limit and a second limit; the first limit and the second limit are obtained by training the principal component analysis model based on the prediction error matrix corresponding to each running state data in the time series training sample set after the variational Koopman anomaly detector has been trained; Based on preset limits, and according to the current operating status data... The value and SPE statistic determine the operating status of the complex equipment of the spacecraft, including: when the current operating status data corresponds to When the value is greater than the first limit or the SPE statistic is greater than the second limit, the working status of the complex equipment of the spacecraft is marked as a malfunction.

4. The fault detection and isolation method for complex spacecraft equipment according to claim 3, characterized in that, The formula for calculating the first limit is: ； in, As the first limit, The number of principal components retained in the principal component analysis model, 1- The confidence level is given by N, and the sample size is given by N. It follows an F-distribution with degrees of freedom. and ; The formula for calculating the second limit is: ； in, This is the second limit. and These are the mean and variance of SPE, respectively. Obeying the degree of freedom The chi-square distribution.

5. The fault detection and isolation method for complex spacecraft equipment according to claim 1, characterized in that, The formula for calculating the deviation score of each operating state variable is as follows: ； in, Let be the deviation score of the s-th variable in the running status data at time t. Let be the s-th variable in the running state data at time t. and The first The mean and standard deviation of the predicted probability distribution of each variable. To find the absolute value.

6. The fault detection and isolation method for complex spacecraft equipment according to claim 1, characterized in that, Fault isolation is performed to identify operational status fault variables, including: after obtaining the deviation scores of each operational status variable, drawing an offset degree map based on the deviation scores of each operational status variable to achieve fault isolation; and marking operational status variables with an offset degree greater than a preset degree value as operational status fault variables.

7. A computer device, comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that the processor executes the computer program to implement the fault detection and isolation method for complex spacecraft equipment according to any one of claims 1-6.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the computer program implements the fault detection and isolation method for complex spacecraft equipment as described in any one of claims 1-6.

Images