Airborne equipment fault detection method based on electromagnetic radiation signal

Abstract

The invention provides an airborne equipment fault detection method based on an electromagnetic radiation signal. The airborne equipment fault detection method comprises the following steps: detecting an electromagnetic radiation frequency band range of equipment in a non-interference environment; n sensors are placed around the equipment to collect electromagnetic radiation signals, and meanwhile M sensors are placed around the equipment and an attenuation shelter is arranged to collect reference signals; dividing the signal into K frequency bands, and designing a band-pass filter to process the signal; counting the energy of each frequency band, combining the signals, comparing the energy and taking the maximum value; judging the frequency band, and if the reference signal energy is greater than the signal, regarding that the noise is zero; and inputting the processed signal into a fault judgment model to identify a fault category. According to the method, faults are rapidly positioned by monitoring electromagnetic radiation changes, and the troubleshooting efficiency is improved. According to the invention, a non-contact and high-precision fault detection method can be realized, the accuracy and efficiency of fault detection are improved, and meanwhile, interference on normal operation of equipment is avoided.

Description

Technical Field

[0001] The present invention relates to the technical field of electromagnetic radiation signal processing, and more specifically, to an airborne equipment fault detection method based on electromagnetic radiation signals. Background Art

[0002] Existing equipment fault detection technologies mainly rely on the monitoring of physical quantities such as temperature, vibration, and sound. These methods have certain limitations. For example, the installation of vibration and temperature sensors requires direct contact with the equipment, which may limit the monitoring range; sound detection is easily affected by environmental noise.

[0003] In addition, traditional detection methods may have difficulty capturing potential minor faults inside the device, resulting in false alarms or missed alarms. Electromagnetic radiation is a natural byproduct of device operation, and its characteristic signals can reflect the operating status of the electronic components inside the device, and there are significant differences between the normal operation and fault status of the device. By monitoring and analyzing the electromagnetic radiation signals of the device, a contactless, real-time device fault detection solution can be implemented, thereby improving the accuracy and efficiency of fault detection. However, considering the high integration of airborne equipment and the large number of minor faults, it is difficult to directly locate the faulty device, especially software operation failures.

[0004] In the process of implementing the embodiments of the present invention, the inventors found that there are at least the following problems or defects in the prior art: the existing fault detection method cannot effectively distinguish minor faults inside the equipment, especially in highly integrated airborne equipment, it is difficult to achieve fast and accurate fault location and diagnosis. Summary of the invention

[0005] The present invention provides an airborne equipment fault detection method based on electromagnetic radiation signals, comprising:

[0006] Step 1: In the interference-free environment of the electromagnetic compatibility laboratory, detect the electromagnetic radiation frequency band range of the equipment by intercepting the range of significant amplitude changes [F min ,F max ];

[0007] Step 2: Place N sensors around the airborne equipment and use electromagnetic radiation collection equipment to collect multi-point electromagnetic radiation signals to obtain x1, x2, ..., x N ;

[0008] Step 3: Place M sensors around the airborne equipment, place an electromagnetic radiation attenuation shelter between the sensor and the airborne equipment, and use the electromagnetic radiation collection equipment to collect multi-point electromagnetic radiation reference signals to obtain r1, r2, ..., r M ;

[0009] Step 4: Divide the electromagnetic radiation signal and the electromagnetic radiation reference signal into K frequency bands, and design bandpass filters f1, f2, ..., f according to the frequency band cutoff frequency. K , and obtain the electromagnetic radiation signal matrix X and the electromagnetic radiation reference signal matrix R after filter processing;

[0010] Step 5: Count the energy of each frequency band of the electromagnetic radiation signal and the electromagnetic radiation reference signal respectively to obtain the electromagnetic radiation signal energy matrix Xe and the electromagnetic radiation reference signal energy matrix Re;

[0011] Step 6: Combine the electromagnetic radiation signal and the electromagnetic radiation reference signal, compare the energy of each frequency band respectively, take the maximum value, and obtain the maximum energy value xe of the electromagnetic radiation signal and the maximum energy value re of the electromagnetic radiation reference signal;

[0012] Step 7: Make a judgment on each frequency band. When the energy of the electromagnetic radiation reference signal is greater than the electromagnetic radiation signal, the current frequency band is regarded as noise, and the energy of the current frequency band is directly set to 0 to obtain the frequency band energy ye of the clean electromagnetic radiation signal;

[0013] Step 8: Input the final ye into the fault judgment model for identification to obtain the fault category c, where the fault judgment model is trained on the historical electromagnetic signal data of the airborne equipment to identify the electromagnetic radiation characteristics under normal and abnormal operating conditions. The training data comes from the historical operation records of the airborne equipment, including electromagnetic radiation signals under various conditions such as normal operation, minor faults, major faults, etc., and c=F(ye).

[0014] Furthermore, in step 4, the electromagnetic radiation signal and the electromagnetic radiation reference signal are divided into K frequency bands, and bandpass filters f1, f2, ..., f are designed according to the cutoff frequencies of the frequency bands. K , and obtaining the electromagnetic radiation signal matrix X and the electromagnetic radiation reference signal matrix R after filter processing, the specific operations are as follows:

[0015]

[0016] In the formula, x ij represents the signal after the i-th electromagnetic radiation signal passes through the j-th bandpass filter, r ij Represents the signal after the i-th electromagnetic radiation reference signal passes through the j-th bandpass filter.

[0017] Furthermore, in the step of respectively counting the energy of each frequency band of the electromagnetic radiation signal and the electromagnetic radiation reference signal in step 5 to obtain the electromagnetic radiation signal energy matrix Xe and the electromagnetic radiation reference signal energy matrix Re, the specific operation is as follows:

[0018]

[0019] In the formula, xe ij represents the energy of the i-th electromagnetic radiation signal in the j-th frequency band, re ij Represents the energy of the i-th electromagnetic radiation reference signal in the j-th frequency band.

[0020] Furthermore, in the step of combining the electromagnetic radiation signal and the electromagnetic radiation reference signal in step 6, comparing the energy of each frequency band, taking the maximum value, and obtaining the maximum energy value xe of the electromagnetic radiation signal and the maximum energy value re of the electromagnetic radiation reference signal, the specific operation is as follows:

[0021] xe=max{xe i11 ,xe 21 ,...,xe N1},max{xe 12 ,xe 22 ,...,xe N2},...,max{xe 1k ,xe 2k ,...,xe NK}

[0022] re=max{re 11 ,re 21 ,...,re M1},max{re 12 ,re 22 ,...,re M2},...,max{re 1k ,re 2k ,...,re MK}

[0023] Further, in step 7, when each frequency band is judged, when the energy of the electromagnetic radiation reference signal is greater than the electromagnetic radiation signal, the current frequency band is regarded as noise, and the energy of the current frequency band is directly set to 0, and the step of obtaining the frequency band energy ye of the clean electromagnetic radiation signal is performed, and the specific operation is as follows:

[0024]

[0025] Furthermore, in step eight, the finally obtained ye is input into the fault judgment model for identification to obtain the fault category c. The fault judgment model adopts the K-NN method, and the identification process is:

[0026] Find the characteristic similarity between the clean electromagnetic radiation signal frequency band energy ye and each group of feature vectors

[0027]

[0028] The feature similarity ξ(ye,T i ) to sort, find the 10 most similar eigenvectors, and count the number of repetitions in the corresponding categories. The category with the largest number is the category corresponding to the frequency band energy ye of the clean electromagnetic radiation signal, where T i is the i-th feature vector in the model feature combination, ye k is the kth element of the clean electromagnetic radiation signal frequency band energy ye, T ik is the kth element of the i-th eigenvector.

[0029] Furthermore, the electromagnetic radiation attenuation shelter is used to reduce the influence of environmental noise on the electromagnetic radiation reference signal.

[0030] Furthermore, the bandpass filter is a Butterworth bandpass filter.

[0031] Furthermore, the electromagnetic radiation signal x i The energy xe in the jth frequency band ij The calculation formula is

[0032]

[0033] Among them, P is the number of sampling points, x ij,i Represents the value of the i-th electromagnetic radiation signal at the i-th sampling point in the j-th frequency band.

[0034] Furthermore, the training steps of the fault judgment model are as follows:

[0035] Step A: In an electromagnetic compatibility laboratory without interference, prepare a fault category to make the airborne equipment have a specified fault, collect electromagnetic radiation signals according to the steps in claim 1, do not collect electromagnetic radiation reference signals, and then divide the frequency bands according to the steps in claim 1 to obtain energy and merge multiple collected signals to obtain a 256-dimensional feature vector;

[0036] Step B: Collect 10 sets of feature vectors during the fault occurrence process. There are C fault categories, and obtain the model feature combination T (50*C, 256);

[0037] Step C: Use the model feature combination to train the fault judgment model so that the fault judgment model can identify the electromagnetic radiation characteristics under normal and abnormal operating conditions.

[0038] The above-mentioned embodiments of the present invention have at least the following beneficial effects: the technical solution of the present invention can be applied to various types of airborne equipment, and has the advantages of high detection accuracy, rapid response, and non-contact. By collecting the electromagnetic radiation signals generated by the airborne equipment and analyzing and processing them, the working status and fault type of the equipment can be identified. This method can monitor without affecting the normal operation of the equipment, thereby improving the accuracy and efficiency of fault detection.

[0039] In addition, the present invention can quickly locate faulty equipment and improve troubleshooting efficiency by monitoring the changes in electromagnetic radiation of the equipment. This method can avoid interference with the normal operation of the equipment and reduce maintenance costs and time, which is of great significance for improving the reliability and safety of airborne equipment. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] The above and other objects, features and advantages of the exemplary embodiments of the present invention will become readily understood by reading the following detailed description with reference to the accompanying drawings. In the accompanying drawings, several embodiments of the present invention are shown in an exemplary and non-limiting manner, in which:

[0041] Figure 1 A schematic flow chart of an airborne equipment fault detection method based on electromagnetic radiation signals provided in one embodiment of the present invention. DETAILED DESCRIPTION

[0042] The principles and spirit of the present invention will be described below with reference to several exemplary embodiments. It should be understood that these embodiments are provided only to enable those skilled in the art to better understand and implement the present invention, and are not intended to limit the scope of the present invention in any way. On the contrary, these embodiments are provided to make the present invention more thorough and complete, and to fully convey the scope of the present invention to those skilled in the art.

[0043] Those skilled in the art know that the embodiments of the present invention can be implemented as a system, device, apparatus, method or computer program product. Therefore, the present invention can be specifically implemented in the following forms, namely: complete hardware, complete software (including firmware, resident software, microcode, etc.), or a combination of hardware and software.

[0044] It should be noted that any number of elements in the drawings is for illustration rather than limitation, and any naming is only for distinction and does not have any limiting meaning.

[0045] Reference below Figure 1 , Figure 1 The following is a flow chart of a method for detecting airborne equipment faults based on electromagnetic radiation signals according to an embodiment of the present invention. Figure 1 As shown, an airborne equipment fault detection method 100 based on electromagnetic radiation signals includes:

[0046] Step 1: In the interference-free environment of the electromagnetic compatibility laboratory, detect the electromagnetic radiation frequency band range of the equipment by intercepting the range of significant amplitude changes [F min ,F max ];

[0047] Step 2: Place N sensors around the airborne equipment and use electromagnetic radiation collection equipment to collect multi-point electromagnetic radiation signals to obtain x1, x2, ..., x N ;

[0048] Step 3: Place M sensors around the airborne equipment, place an electromagnetic radiation attenuation shelter between the sensor and the airborne equipment, and use the electromagnetic radiation collection equipment to collect multi-point electromagnetic radiation reference signals to obtain r1, r2, ..., r M ;

[0049] Step 4: Divide the electromagnetic radiation signal and the electromagnetic radiation reference signal into K frequency bands, and design bandpass filters f1, f2, ..., f according to the frequency band cutoff frequency. K , and obtain the electromagnetic radiation signal matrix X and the electromagnetic radiation reference signal matrix R after filter processing;

[0050] Step 5: Count the energy of each frequency band of the electromagnetic radiation signal and the electromagnetic radiation reference signal respectively to obtain the electromagnetic radiation signal energy matrix Xe and the electromagnetic radiation reference signal energy matrix Re;

[0051] Step 6: Combine the electromagnetic radiation signal and the electromagnetic radiation reference signal, compare the energy of each frequency band respectively, take the maximum value, and obtain the maximum energy value xe of the electromagnetic radiation signal and the maximum energy value re of the electromagnetic radiation reference signal;

[0052] Step 7: Make a judgment on each frequency band. When the energy of the electromagnetic radiation reference signal is greater than the electromagnetic radiation signal, the current frequency band is regarded as noise, and the energy of the current frequency band is directly set to 0 to obtain the frequency band energy ye of the clean electromagnetic radiation signal;

[0053] Step 8: Input the final ye into the fault judgment model for identification to obtain the fault category c, where the fault judgment model is trained on the historical electromagnetic signal data of the airborne equipment to identify the electromagnetic radiation characteristics under normal and abnormal operating conditions. The training data comes from the historical operation records of the airborne equipment, including electromagnetic radiation signals under various conditions such as normal operation, minor faults, major faults, etc., and c=F(ye).

[0054] It should be noted that this method first involves detecting the electromagnetic radiation frequency band range of the device in an interference-free environment of an electromagnetic compatibility laboratory. The electromagnetic compatibility laboratory here refers to a laboratory environment specially designed for testing electromagnetic compatibility, which can isolate external electromagnetic interference and ensure the accuracy of the test results. The electromagnetic radiation frequency band range refers to the frequency range of the electromagnetic radiation signal generated by the device during operation, which is crucial for subsequent signal acquisition and analysis.

[0055] Specifically, the detection method used in this step can be implemented by intercepting the range of significant amplitude changes, that is, monitoring the electromagnetic radiation signal generated by the equipment in the laboratory environment, and recording the frequency interval where the signal amplitude changes significantly. These intervals represent the possible fault characteristic frequency bands of the equipment, providing a basis for subsequent sensor layout and signal acquisition. Parameter settings include selecting appropriate monitoring equipment, setting appropriate sampling frequency and resolution, etc.

[0056] Preferably, the detection step may also include comparative analysis of the electromagnetic radiation frequency band range of the device under different working conditions to determine the most representative fault characteristic frequency band. For example, the electromagnetic radiation signals of the device under normal working conditions and known fault conditions may be compared to find the difference between the two, thereby more accurately determining the fault characteristic frequency band. In addition, it may also be considered to use automated software to assist in the analysis and improve the efficiency and accuracy of the detection process.

[0057] In some embodiments, in step 4, the electromagnetic radiation signal and the electromagnetic radiation reference signal are divided into K frequency bands, and the bandpass filters f1, f2, ..., f are designed according to the cutoff frequencies of the frequency bands. K , and obtaining the electromagnetic radiation signal matrix X and the electromagnetic radiation reference signal matrix R after filter processing, the specific operations are as follows:

[0058]

[0059]

[0060] In the formula, x ij represents the signal after the i-th electromagnetic radiation signal passes through the j-th bandpass filter, r ij Represents the signal after the i-th electromagnetic radiation reference signal passes through the j-th bandpass filter.

[0061] It should be noted that this step involves dividing the electromagnetic radiation signal and the electromagnetic radiation reference signal into K frequency bands, and designing a bandpass filter based on the cutoff frequency of the frequency band. The bandpass filter here is an electronic filter that allows signals in a specific frequency band to pass through while blocking signals in other frequency bands. The frequency band cutoff frequency refers to the highest and lowest frequency points that the bandpass filter allows the signal to pass through. The purpose of designing a bandpass filter is to extract specific frequency band signals related to equipment failure from a complex electromagnetic environment.

[0062] Specifically, in this step, the division of K frequency bands is based on the preliminary detection results of the frequency band range of the electromagnetic radiation of the equipment. The cutoff frequency of each frequency band can be set according to the specific characteristics of the equipment and the fault characteristics. For example, if the change of electromagnetic radiation of the equipment in a specific frequency range is strongly correlated with the fault, then this frequency range can be set as a frequency band.

[0063] More specifically, parameter setting may include determining the center frequency, bandwidth, and filter order of each frequency band, etc. The setting of these parameters will directly affect the effect of signal processing, and therefore needs to be accurately adjusted based on experimental data and theoretical analysis.

[0064] Preferably, this step may also include using computer-aided design software to assist in the design of a bandpass filter to ensure that the performance of the filter meets specific technical requirements. In addition, a tunable filter may be considered, so that the parameters of the filter can be dynamically adjusted according to the characteristics of the electromagnetic radiation signal actually monitored to adapt to different monitoring environments and equipment states. This flexibility can improve the adaptability and accuracy of the fault detection method.

[0065] In some embodiments, in the step of respectively counting the energy of each frequency band of the electromagnetic radiation signal and the electromagnetic radiation reference signal in step 5 to obtain the electromagnetic radiation signal energy matrix Xe and the electromagnetic radiation reference signal energy matrix Re, the specific operation is as follows:

[0066]

[0067]

[0068] In the formula, xe ij represents the energy of the i-th electromagnetic radiation signal in the j-th frequency band, re ij Represents the energy of the i-th electromagnetic radiation reference signal in the j-th frequency band.

[0069] It should be noted that this step involves counting the energy of each frequency band of the electromagnetic radiation signal and the electromagnetic radiation reference signal. The energy here refers to the energy distribution of the signal in a specific frequency band, which is usually obtained by calculating the power spectrum density of the signal. The electromagnetic radiation signal energy matrix and the electromagnetic radiation reference signal energy matrix are data structures that record the energy values ​​of each frequency band and are used for subsequent signal comparison and analysis.

[0070] Specifically, in this step, the energy of each frequency band can be calculated by accumulating the squares of the filtered signals. For example, the energy of the electromagnetic radiation signal in the i-th frequency band can be calculated as the sum of the squares of the signal amplitudes of all sampling points in the frequency band. Parameter settings include determining the number of sampling points, the sampling frequency of the signal, and the specific method of accumulation. The settings of these parameters will directly affect the accuracy and efficiency of energy calculation.

[0071] Preferably, this step may also include using a frequency domain analysis method such as Fast Fourier Transform (FFT) to calculate the energy of the signal. This method can process a large amount of data more efficiently and can provide more accurate energy distribution information.

[0072] Furthermore, it is possible to consider introducing a window function to reduce spectrum leakage and improve the accuracy of energy calculation. The choice of the window function can be determined according to the characteristics of the signal and the specific requirements of the monitoring environment, such as a Hamming window, a Hanning window, or a Blackman window.

[0073] In some embodiments, in the step of combining the electromagnetic radiation signal and the electromagnetic radiation reference signal in step 6, comparing the energy of each frequency band respectively, taking the maximum value, and obtaining the maximum energy value xe of the electromagnetic radiation signal and the maximum energy value re of the electromagnetic radiation reference signal, the specific operation is as follows:

[0074] xe=max{xe 11 ,xe 21 ,...,xe N1},max{xe 12 ,xe 22 ,...,xe N2},...,max{xe 1k ,xe 2k ,...,xe NK}

[0075] re=max{re 11 ,re 21 ,...,re M1},max{re 12 ,re 22 ,...,re M2},...,max{re 1k ,re2k ,...,re MK}

[0076] It should be noted that this step involves the process of combining the electromagnetic radiation signal and the electromagnetic radiation reference signal, and comparing the energy of each frequency band and taking the maximum value. The combination here refers to integrating the signals collected from different sensors for subsequent energy comparison. The maximum value means that in the comparison process, for each pair of corresponding frequency band energies of the electromagnetic radiation signal and the reference signal, the higher one of the two is selected as the result.

[0077] Specifically, in this step, the merging operation can be achieved by summarizing the signal energy values ​​collected by each sensor. For example, for electromagnetic radiation signals, we can compare the energy values ​​of N sensors in the kth frequency band and select the maximum value as the maximum energy value of the frequency band.

[0078] More specifically, parameter settings include determining the frequency band range for comparison, the algorithm for selecting the maximum value (such as direct comparison or using statistical methods), etc. The settings of these parameters will directly affect the results of energy comparison and the accuracy of subsequent fault judgment.

[0079] Preferably, this step may also include using a weighted average method to determine the maximum energy value of each frequency band. For example, different weights may be assigned according to the location of the sensor or the signal-to-noise ratio of the signal, so that when comparing the energy, the data of certain sensors will be assigned a higher weight.

[0080] Furthermore, it is possible to consider introducing threshold judgment and further analyzing the close values ​​to avoid misjudgment caused by small fluctuations. This method can improve the robustness of energy comparison and make fault detection more accurate.

[0081] In some embodiments, in step 7, when the energy of the electromagnetic radiation reference signal is greater than the electromagnetic radiation signal, the current frequency band is regarded as noise, and the energy of the current frequency band is directly set to 0, and the step of obtaining the frequency band energy ye of the clean electromagnetic radiation signal is performed as follows:

[0082]

[0083] It should be noted that this step involves judging each frequency band. When the energy of the electromagnetic radiation reference signal is greater than the electromagnetic radiation signal, the current frequency band is regarded as noise, and the energy of the current frequency band is directly set to 0 to obtain the frequency band energy of the clean electromagnetic radiation signal. The noise here refers to those interference signals in the electromagnetic radiation signal that are irrelevant to the normal operation of the device. Setting the energy to 0 is a processing method that aims to eliminate these interferences so as to more accurately identify the true status of the device.

[0084] Specifically, in this step, the judgment operation can be implemented by comparing the energy value of the electromagnetic radiation signal and the reference signal in each frequency band. If the energy value of the reference signal is higher than the energy value of the signal, it can be considered that the signal in the frequency band is affected by noise, and the energy value of the signal needs to be set to 0.

[0085] More specifically, parameter settings include determining the threshold of energy comparison, selecting which algorithm to use for energy comparison, etc. The settings of these parameters will directly affect the effect of noise processing and the accuracy of subsequent fault detection.

[0086] Preferably, this step may also include using an adaptive algorithm to dynamically adjust the threshold of the energy comparison. For example, the threshold may be adjusted according to historical data and changes in the current environment to adapt to different monitoring conditions.

[0087] Furthermore, it is possible to consider introducing machine learning algorithms to identify and classify noise, which can not only improve the accuracy of noise processing, but also continuously optimize the performance of noise identification over time. This approach can make the fault detection system more intelligent and improve its robustness in complex environments.

[0088] In some embodiments, in the step of inputting the finally obtained ye into the fault judgment model for identification to obtain the fault category c in step eight, the fault judgment model adopts the K-NN method, and the identification process is:

[0089] Find the characteristic similarity between the clean electromagnetic radiation signal frequency band energy ye and each group of feature vectors

[0090]

[0091] The feature similarity ξ(ye,T i ) to sort, find the 10 most similar eigenvectors, and count the number of repetitions in the corresponding categories. The category with the largest number is the category corresponding to the frequency band energy ye of the clean electromagnetic radiation signal, where T i is the i-th feature vector in the model feature combination, ye k is the kth element of the clean electromagnetic radiation signal frequency band energy ye, T ik is the kth element of the i-th eigenvector.

[0092] It should be noted that this step involves inputting the energy of the clean electromagnetic radiation signal into the fault judgment model for identification to determine the fault category. The fault judgment model here refers to an algorithm or system that can identify the type of equipment fault based on the input feature data. Fault category refers to the different types of faults that may occur in the equipment, such as normal operation, minor fault or major fault.

[0093] Specifically, in this step, the recognition process of the fault judgment model can be implemented by the K-NN (K nearest neighbor) method. K-NN is a distance-based classification algorithm that calculates the distance between the test sample and the known sample, finds the nearest K samples, and then predicts the category of the test sample based on the categories of these samples. Parameter settings include selecting the value of K, determining the distance measurement method (such as Euclidean distance or Manhattan distance), etc. The setting of these parameters will directly affect the classification accuracy and efficiency of the model.

[0094] Preferably, this step may also include optimizing the fault judgment model to improve its recognition capability. For example, a cross-validation method may be used to determine the optimal K value, or a feature selection technique may be used to reduce unnecessary features and improve the generalization capability of the model.

[0095] Furthermore, it is possible to consider combining other machine learning algorithms, such as support vector machines (SVM) or neural networks, to improve the accuracy of fault detection. These alternatives can be selected based on the specific application scenario and data characteristics to achieve the best fault detection effect.

[0096] In some embodiments, the electromagnetic radiation attenuation shelter is used to reduce the impact of environmental noise on the electromagnetic radiation reference signal.

[0097] It should be noted that the electromagnetic radiation attenuation shelter mentioned in this step is used to reduce the impact of environmental noise on the electromagnetic radiation reference signal. The electromagnetic radiation attenuation shelter here refers to a physical structure designed to absorb or reduce the propagation of electromagnetic waves, thereby reducing the contamination of the reference signal by external electromagnetic interference. This shelter is crucial to improving the accuracy and reliability of signal acquisition.

[0098] Specifically, electromagnetic radiation attenuation shelters can be made of conductive materials, such as copper or aluminum, which can effectively absorb or shield electromagnetic waves. The design of the shelter includes an enclosed space or container in which the equipment is placed, or a shielding cover around the sensor. The parameter setting involves the size of the shelter, the material selection, and the shielding effectiveness level. The setting of these parameters will directly affect the performance of the shelter, that is, its ability to reduce noise.

[0099] Preferably, the design of the electromagnetic radiation attenuation shelter can also take into account its absorption effect on electromagnetic waves in a specific frequency range. For example, a specific frequency selective shield can be designed to optimize the electromagnetic radiation in a specific frequency band generated when the equipment is running.

[0100] Furthermore, it is possible to consider using multiple layers of shielding materials to further improve the shielding effect. Such a multi-layer structure can include a combination of layers of different thicknesses and materials to suit different electromagnetic environments and application requirements. These alternatives can be customized according to the actual electromagnetic environment and device characteristics to achieve the best noise suppression effect.

[0101] In some embodiments, the bandpass filter is a Butterworth bandpass filter.

[0102] It should be noted that the bandpass filter mentioned in this step is a Butterworth bandpass filter. Butterworth filter is an analog or digital filter known for its flat frequency response within the passband, which makes it very useful in signal processing. A bandpass filter is a filter that only allows signals within a specific frequency range to pass through, while blocking other frequency signals.

[0103] Specifically, the design of a Butterworth bandpass filter involves determining the filter's cutoff frequency, center frequency, and bandwidth. These parameters define the frequency range that the filter allows to pass. For example, if we know that the characteristic signal of a device failure is mainly concentrated in a certain frequency interval, we can set this interval to the center frequency and set the bandwidth to cover the range of this interval. Parameter settings include the order of the filter, which affects the roll-off rate of the filter at the cutoff frequency, and the specific implementation method of the filter, such as analog circuit design or digital signal processing algorithm.

[0104] Preferably, the design of the Butterworth bandpass filter may also include optimizing the order of the filter to achieve the best performance in a specific application. For example, increasing the order of the filter can improve the roll-off rate of the filter, but it will also increase the complexity of the filter and possible phase distortion. Therefore, it is necessary to find a balance between filtering performance and system complexity during design.

[0105] Further, it is possible to consider using digital signal processing techniques to implement the Butterworth filter, which provides more flexibility and adjustment possibilities, especially when rapid adaptation to different monitoring conditions is required. This alternative allows the filter parameters to be adjusted through software updates without changing the hardware design.

[0106] In some embodiments, the electromagnetic radiation signal x i The energy xe in the jth frequency band ij The calculation formula is

[0107]

[0108] Among them, P is the number of sampling points, x ij,i Represents the value of the i-th electromagnetic radiation signal at the i-th sampling point in the j-th frequency band.

[0109] It should be noted that this step describes the energy calculation method of the electromagnetic radiation signal in the kth frequency band. The energy calculation here refers to the quantification of the power distribution of the signal in a specific frequency band by a specific mathematical method. The number of sampling points refers to the number of points into which the signal is discretized during the signal processing process, and these points contain the amplitude information of the signal.

[0110] Specifically, the energy calculation of the electromagnetic radiation signal in the kth frequency band can be achieved by accumulating the squares of the amplitudes of all sampling points of the signal in the frequency band. For example, if a signal has P sampling points in the kth frequency band, the energy of the frequency band can be expressed as the sum of the squares of the amplitudes of these sampling points. Parameter settings include determining the number of sampling points P, the sampling frequency of the signal, and the specific method of accumulation. The settings of these parameters will directly affect the accuracy and efficiency of energy calculation.

[0111] Preferably, the energy calculation in this step may also include using a window function to reduce the influence of spectrum leakage. The window function is a common technique in frequency domain analysis, which reduces the discontinuity at both ends of the sampled signal by multiplying a specific mathematical function, thereby reducing spectrum leakage.

[0112] Furthermore, it is possible to consider using the Fast Fourier Transform (FFT) to efficiently calculate the frequency domain representation of the signal and then calculate the energy. This approach can improve computational efficiency, especially when processing large amounts of data. These alternatives can be selected based on the characteristics of the signal and the processing requirements to achieve the best signal energy calculation effect.

[0113] In some embodiments, the training steps of the fault judgment model are as follows:

[0114] Step A: In an electromagnetic compatibility laboratory without interference, prepare a fault category to make the airborne equipment have a specified fault, collect electromagnetic radiation signals according to the steps in claim 1, do not collect electromagnetic radiation reference signals, and then divide the frequency bands according to the steps in claim 1 to obtain energy and merge multiple collected signals to obtain a 256-dimensional feature vector;

[0115] Step B: Collect 10 sets of feature vectors during the fault occurrence process. There are C fault categories, and obtain the model feature combination T (50*C, 256);

[0116] Step C: Use the model feature combination to train the fault judgment model so that the fault judgment model can identify the electromagnetic radiation characteristics under normal and abnormal operating conditions.

[0117] It should be noted that this step involves the training process of the fault judgment model. The fault judgment model refers to an algorithm or system used to identify the electromagnetic radiation characteristics of airborne equipment under normal and abnormal operating conditions. The training mentioned here refers to the use of historical data to adjust the model parameters so that it can accurately classify new data. A feature vector refers to a set of values ​​extracted from an electromagnetic radiation signal that can represent the key characteristics of the signal.

[0118] Specifically, the training steps of the fault judgment model include preparing fault categories in an interference-free environment of an electromagnetic compatibility laboratory to cause specified faults in the airborne equipment, and collecting electromagnetic radiation signals according to the method described above. In this process, the electromagnetic radiation reference signal is not collected, but the signal is directly obtained from the operation of the equipment, and the frequency band is divided according to the method described above to obtain energy, and multiple collected signals are merged to finally obtain a 256-dimensional feature vector. There are C fault categories, and a model feature combination is formed by collecting 10 groups of feature vectors. Parameter settings include selecting the dimension of the feature vector, determining the number of samples, and selecting a suitable machine learning algorithm. The setting of these parameters will directly affect the performance and accuracy of the model.

[0119] Preferably, the training process of this step may also include the use of cross-validation to evaluate the performance of the model to ensure the generalization ability of the model on unseen data. In addition, different feature selection techniques may be considered to reduce the dimension of the feature vector and improve the training efficiency and prediction speed of the model. For example, principal component analysis (PCA) may be used to reduce the dimension of the feature space while retaining the most important information. These alternatives can be customized according to specific application requirements and data characteristics to achieve the best model training effect.

[0120] The above-mentioned embodiments of the present invention have the following beneficial effects: The airborne equipment fault detection method based on electromagnetic radiation signals described in the present invention can improve the accuracy and efficiency of fault detection. By detecting the electromagnetic radiation frequency band range of the equipment in an interference-free environment, and utilizing multi-point electromagnetic radiation signal acquisition and attenuation shelter technology, the method can monitor the changes in electromagnetic radiation signals in real time when the equipment is running, thereby effectively identifying the normal and abnormal operating states of the equipment. The application of this method can reduce direct contact with the equipment and avoid the installation complexity and interference with equipment operation that may be caused by traditional contact monitoring methods.

[0121] In addition, by designing a bandpass filter to process the signal, counting the energy of each frequency band, comparing the signal energy and identifying the noise frequency band, this technology can effectively extract useful fault characteristic signals from a complex electromagnetic environment. This method can not only improve the sensitivity and specificity of fault detection, but also quickly locate the faulty equipment through the fault judgment model, thereby speeding up the troubleshooting process, reducing the downtime of airborne equipment, and improving the operational reliability and safety of the equipment.

[0122] Furthermore, the storage medium of the embodiment of the present application stores program instructions that can implement all the above methods, wherein the program instructions can be stored in the above storage medium in the form of a software product, including several instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) or a processor to execute all or part of the steps of the methods described in each embodiment of the present application. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, or terminal devices such as a computer, a server, a mobile phone, and a tablet.

[0123] The above descriptions are only some preferred embodiments of the present invention and an explanation of the technical principles used. Those skilled in the art should understand that the scope of the invention involved in the embodiments of the present invention is not limited to the technical solutions formed by a specific combination of the above technical features, but should also cover other technical solutions formed by any combination of the above technical features or their equivalent features without departing from the above inventive concept. For example, the above features are replaced with (but not limited to) technical features with similar functions disclosed in the embodiments of the present invention to form a technical solution.

Claims

1. A method for detecting airborne equipment faults based on electromagnetic radiation signals, characterized in that: The following steps are involved: Step 1: In the interference-free environment of the electromagnetic compatibility laboratory, detect the electromagnetic radiation frequency band range of the equipment by intercepting the range of significant amplitude changes [F min ,F max ]; Step 2: Place N sensors around the airborne equipment and use electromagnetic radiation collection equipment to collect multi-point electromagnetic radiation signals to obtain x1, x2, ..., x N ; Step 3: Place M sensors around the airborne equipment, place an electromagnetic radiation attenuation shelter between the sensor and the airborne equipment, and use the electromagnetic radiation collection equipment to collect multi-point electromagnetic radiation reference signals to obtain r1, r2, ..., r M ; Step 4: Divide the electromagnetic radiation signal and the electromagnetic radiation reference signal into K frequency bands, and design bandpass filters f1, f2, ..., f according to the frequency band cutoff frequency. K , and obtain the electromagnetic radiation signal matrix X and the electromagnetic radiation reference signal matrix R after filter processing; Step 5: Count the energy of each frequency band of the electromagnetic radiation signal and the electromagnetic radiation reference signal respectively to obtain the electromagnetic radiation signal energy matrix Xe and the electromagnetic radiation reference signal energy matrix Re; Step 6: Combine the electromagnetic radiation signal and the electromagnetic radiation reference signal, compare the energy of each frequency band respectively, take the maximum value, and obtain the maximum energy value xe of the electromagnetic radiation signal and the maximum energy value re of the electromagnetic radiation reference signal; Step 7: Make a judgment on each frequency band. When the energy of the electromagnetic radiation reference signal is greater than the electromagnetic radiation signal, the current frequency band is regarded as noise, and the energy of the current frequency band is directly set to 0 to obtain the frequency band energy ye of the clean electromagnetic radiation signal; Step 8: Input the final ye into the fault judgment model for identification to obtain the fault category c, where the fault judgment model is trained on the historical electromagnetic signal data of the airborne equipment to identify the electromagnetic radiation characteristics under normal and abnormal operating conditions. The training data comes from the historical operation records of the airborne equipment, including electromagnetic radiation signals under various conditions such as normal operation, minor faults, major faults, etc., and c=F(ye).

2. The method for detecting airborne equipment faults based on electromagnetic radiation signals according to claim 1, characterized in that: In step 4, the electromagnetic radiation signal and the electromagnetic radiation reference signal are divided into K frequency bands, and bandpass filters f1, f2, ..., f are designed according to the cutoff frequencies of the frequency bands. K , and obtaining the electromagnetic radiation signal matrix X and the electromagnetic radiation reference signal matrix R after filter processing, the specific operations are as follows: In the formula, x ij represents the signal after the i-th electromagnetic radiation signal passes through the j-th bandpass filter, r ij Represents the signal after the i-th electromagnetic radiation reference signal passes through the j-th bandpass filter.

3. The method for detecting airborne equipment faults based on electromagnetic radiation signals according to claim 1, characterized in that: In the step of respectively counting the energy of each frequency band of the electromagnetic radiation signal and the electromagnetic radiation reference signal in step 5 to obtain the electromagnetic radiation signal energy matrix Xe and the electromagnetic radiation reference signal energy matrix Re, the specific operation is as follows: In the formula, xe ij represents the energy of the i-th electromagnetic radiation signal in the j-th frequency band, re ij Represents the energy of the i-th electromagnetic radiation reference signal in the j-th frequency band.

4. The method for detecting airborne equipment faults based on electromagnetic radiation signals according to claim 1, characterized in that: In the step of combining the electromagnetic radiation signal and the electromagnetic radiation reference signal in step 6, comparing the energy of each frequency band respectively, taking the maximum value, and obtaining the maximum energy value xe of the electromagnetic radiation signal and the maximum energy value re of the electromagnetic radiation reference signal, the specific operation is as follows: car=max{car 11 ,car 21 ,...,car N1 },max{car 12 ,car 22 ,...,car N2 },...,max{car 1k ,car 2k ,...,car NK } re＝ max{re 11 ,re 21 ,...,re M1 },max{re 12 ,re 22 ,...,re M2 },...,max{re 1k ,re 2k ,...,re MK }。 5. The method for detecting airborne equipment faults based on electromagnetic radiation signals according to claim 1, characterized in that: In step 7, when judging each frequency band, when the energy of the electromagnetic radiation reference signal is greater than the electromagnetic radiation signal, the current frequency band is regarded as noise, and the energy of the current frequency band is directly set to 0, and the step of obtaining the frequency band energy ye of the clean electromagnetic radiation signal is performed. The specific operation is as follows:

6. The method for detecting airborne equipment faults based on electromagnetic radiation signals according to claim 1, characterized in that: In step eight, the finally obtained ye is input into the fault judgment model for identification to obtain the fault category c. The fault judgment model adopts the K-NN method, and the identification process is: Find the characteristic similarity between the clean electromagnetic radiation signal frequency band energy ye and each group of feature vectors The feature similarity ξ(ye,T i ) to sort, find the 10 most similar eigenvectors, and count the number of repetitions in the corresponding categories. The category with the largest number is the category corresponding to the frequency band energy ye of the clean electromagnetic radiation signal, where T i is the i-th feature vector in the model feature combination, ye k is the kth element of the clean electromagnetic radiation signal frequency band energy ye, T ik is the kth element of the i-th eigenvector.

7. The method for detecting airborne equipment faults based on electromagnetic radiation signals according to claim 1, characterized in that: The electromagnetic radiation attenuation shelter is used to reduce the influence of environmental noise on the electromagnetic radiation reference signal.

8. The method for detecting airborne equipment faults based on electromagnetic radiation signals according to claim 1, characterized in that: The bandpass filter is a Butterworth bandpass filter.

9. The method for detecting airborne equipment faults based on electromagnetic radiation signals according to claim 1, characterized in that: Electromagnetic radiation signal x i The energy xe in the jth frequency band ij The calculation formula is Among them, P is the number of sampling points, x ij,i Represents the value of the i-th electromagnetic radiation signal at the i-th sampling point in the j-th frequency band.

10. The method for detecting airborne equipment faults based on electromagnetic radiation signals according to claim 1, characterized in that: The training steps of the fault judgment model are as follows: Step A: In an electromagnetic compatibility laboratory without interference, prepare a fault category to make the airborne equipment have a specified fault, collect electromagnetic radiation signals according to the steps in claim 1, do not collect electromagnetic radiation reference signals, and then divide the frequency bands according to the steps in claim 1 to obtain energy and merge multiple collected signals to obtain a 256-dimensional feature vector; Step B: Collect 10 sets of feature vectors during the fault occurrence process. There are C fault categories, and obtain the model feature combination T (50*C, 256); Step C: Use the model feature combination to train the fault judgment model so that the fault judgment model can identify the electromagnetic radiation characteristics under normal and abnormal operating conditions.

Images