Fault monitoring method and device for aero-engine and storage medium

Abstract

The invention relates to the technical field of aero-engine state monitoring and fault diagnosis, in particular to a fault monitoring method and device for an aero-engine and a storage medium. The method comprises the steps that vibration signals of all parts in the working process of the aero-engine are obtained; dividing the vibration signal into a steady-state signal, a periodic impact signal or an unsteady-state transient signal based on the short-time energy and the zero-crossing rate; when the vibration signal is a steady-state signal, determining whether a fault exists or not by adopting a time domain feature extraction and classification model; when the vibration signal is a periodic impact signal, Fourier transform and envelope spectrum analysis are adopted to determine whether a fault exists; and when the vibration signal is an unstable transient signal, determining whether a fault exists or not by adopting time-frequency analysis. Compared with a single analysis algorithm, the method has the advantages that different algorithms are matched for fault monitoring based on the characteristics of the vibration signals, the monitoring accuracy is improved, and the false alarm rate is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of aircraft engine state monitoring and fault diagnosis, and in particular to an aircraft engine fault monitoring method, device and storage medium. Background Art

[0002] As the core power unit of aircraft, the operating status of aircraft engines directly impacts flight safety and equipment lifespan. Vibration signals, a key parameter characterizing engine mechanical performance, can reveal early signs of a variety of typical faults, including bearing wear, blade breakage, rotor imbalance, and surge. Therefore, real-time monitoring and intelligent analysis of vibration signals have become a research priority in aircraft engine health management.

[0003] However, when using vibration signals for fault monitoring, traditional monitoring methods often rely on single analysis algorithms, such as frequency-domain feature extraction based on spectrum analysis or time-domain statistical methods based on metrics such as RMS value and peak value. These methods fail to incorporate the characteristics of the vibration signal for analysis and processing, resulting in high false alarm rates and delayed responses. Furthermore, existing alarm systems typically provide only simple audio and visual prompts, lacking the ability to differentiate and address different types of anomalies. Summary of the Invention

[0004] In view of this, the present invention provides a method, device and storage medium for fault monitoring of an aircraft engine to solve the problem of high false alarm rate in fault monitoring in the prior art.

[0005] In a first aspect, the present invention provides a method for fault monitoring of an aircraft engine, the method comprising: obtaining vibration signals from various parts of the aircraft engine during operation; dividing the vibration signals into steady-state signals, periodic impact signals, or non-steady-state transient signals based on short-time energy and zero-crossing rate; when the vibration signal is a steady-state signal, using time domain feature extraction and classification models to determine whether a fault exists; when the vibration signal is a periodic impact signal, using Fourier transform and envelope spectrum analysis to determine whether a fault exists; when the vibration signal is a non-steady-state transient signal, using time-frequency analysis to determine whether a fault exists.

[0006] In the present invention, the vibration signals are first classified based on the short-time energy and the zero-crossing rate, and then the corresponding data mining algorithm is used to process the vibration signals of different types, thereby realizing dynamic matching of the signal features. Compared with the use of a single analysis algorithm, this method performs fault monitoring based on the feature matching of the vibration signal with different algorithms, thereby improving the monitoring accuracy and reducing the false alarm rate.

[0007] In an optional embodiment, the vibration signal is divided into a steady-state signal, a periodic impact signal or a non-steady-state transient signal based on short-time energy and zero-crossing rate, including: when the short-time energy is less than or equal to the short-time energy baseline and the zero-crossing rate is less than or equal to the first zero-crossing rate threshold, the vibration signal is divided into a steady-state signal; when the short-time energy is greater than the short-time energy baseline and the periodic detection condition is met, the vibration signal is divided into a periodic impact signal; when the short-time energy is greater than the transient event energy threshold and the zero-crossing rate is greater than the second zero-crossing rate threshold, the vibration signal is divided into a non-steady-state transient signal, the transient event energy threshold is greater than the short-time energy baseline, and the second zero-crossing rate threshold is greater than the first zero-crossing rate threshold.

[0008] In this invention, the coordinated determination of short-term energy and zero-crossing rate allows for precise differentiation between steady-state, periodic impulse, and non-steady-state transient signals. Furthermore, classification based on different signal feature thresholds effectively captures differences in engine operating states, providing accurate initial classification for fault diagnosis.

[0009] In an optional embodiment, when the vibration signal is a steady-state signal, a time domain feature extraction and classification model is used to determine whether a fault exists, including: extracting the peak value, kurtosis, pulse factor and root mean square of the vibration signal as time domain features; standardizing the time domain features; inputting the standardized time domain features into a support vector machine classifier to obtain an output result of whether a fault exists.

[0010] In this method, time-domain features such as peak value and kurtosis are first extracted from steady-state vibration signals. These features are then normalized and fed into a support vector machine classifier. This approach leverages time-domain features to directly reflect the signal's essential characteristics, while standardization eliminates dimensionality. Support vector machines excel at small sample sizes and nonlinear classification, enabling precise and efficient identification of steady-state signal faults. This approach provides a reliable means for fault diagnosis, helping to promptly identify potential hazards and ensure stable operation.

[0011] In an optional embodiment, when the vibration signal is a periodic impact signal, Fourier transform and envelope spectrum analysis are used to determine whether a fault exists, including: performing a Fourier transform on the vibration signal to determine the characteristic frequency; extracting the envelope of the vibration signal and performing a Fourier transform to obtain the envelope spectrum; and determining whether a fault exists based on the relationship between the peak value in the envelope spectrum that exceeds three standard deviations of the baseline and the characteristic frequency and its multiples.

[0012] This method combines Fourier transform with envelope spectrum analysis for periodic impulse signals. The Fourier transform precisely locates the characteristic frequency, while envelope spectrum analysis highlights the periodicity of the impulse. Fault identification is performed by correlating the baseline peak value (three times the standard deviation) with the characteristic frequency multiplication factor. This method effectively captures the subtle impulse characteristics of periodic faults, overcomes the limitations of conventional spectrum analysis, improves the accuracy and reliability of fault identification, and provides an efficient and precise technical path for periodic fault diagnosis.

[0013] In an optional embodiment, when the vibration signal is a non-steady-state transient signal, time-frequency analysis is used to determine whether a fault exists, including: using continuous wavelet transform to determine the wavelet coefficients of the vibration signal at different scales and time points; determining a dynamic threshold based on the modulus square, mean and standard deviation of the wavelet coefficients; determining whether a fault exists based on whether multiple scales at the same time point exceed the dynamic threshold; and verifying the determined existing fault using the time-frequency energy distribution obtained by the Hilbert-Huang transform.

[0014] In this paper, a continuous wavelet transform is used to capture wavelet coefficients at different scales and time points for non-steady-state transient vibration signals. Dynamic thresholds are determined using the energy distribution represented by the squared modulus of the coefficients. Faults are then determined using multi-scale over-threshold detection, and verified using the Hilbert-Huang transform. This approach can accurately identify short-term, sudden, non-steady-state faults and adapt to transient anomalies such as surge, overcoming the limitations of traditional methods and improving the timeliness and accuracy of fault diagnosis under complex operating conditions.

[0015] In an optional embodiment, before obtaining the vibration signals of various parts during the operation of the aircraft engine, the method further includes: obtaining the parts of the aircraft engine to be monitored and the types of sensors used for vibration signal acquisition, the parts including the compressor, turbine, rotor bearing, combustion chamber, tail nozzle and engine mounting bracket, and the sensor types including piezoelectric accelerometer, micro-electromechanical system accelerometer, fiber grating vibration sensor and capacitive vibration sensor; determining the type of sensor installed at each part according to the environment in which each part is located and the frequency band of the vibration signal generated by each part.

[0016] In this invention, a variety of sensors are combined according to monitoring requirements and environmental conditions to form a multimodal data acquisition network. Through the collaboration of piezoelectric, MEMS, fiber optic, and capacitive sensors, the full frequency range of vibration signals is covered, ensuring monitoring reliability under complex working conditions.

[0017] In an optional embodiment, after obtaining vibration signals from various parts of the aircraft engine during operation, the method further includes: using a wavelet threshold denoising algorithm to remove high-frequency noise and low-frequency drift in the vibration signal; and performing normalization processing and signal segmentation processing on the vibration signal.

[0018] In this invention, the collected aircraft engine vibration signals are first denoised using a wavelet threshold algorithm to accurately filter out high-frequency noise and low-frequency drift, restoring the signal's true nature. Normalization and segmentation are then performed to unify signal characteristics and adapt to batch processing requirements. This effectively improves signal quality and lays a solid foundation for subsequent signal classification and fault diagnosis, including steady-state, periodic impact, and non-steady-state transients.

[0019] In an optional embodiment, the method further includes: when it is determined that a fault exists, controlling the buzzer to alarm at a preset frequency and controlling the red light to flash; when it is determined that no fault exists, controlling the buzzer to mute and controlling the green light to be always on.

[0020] In this invention, when a fault occurs, the buzzer sounds at a preset frequency and the red light flashes, triggering an immediate warning and facilitating a quick response from maintenance personnel. When there is no fault, the buzzer becomes silent and the green light stays on, indicating the equipment is in a normal state. This improves alarm recognition through the coordinated feedback of sound and light.

[0021] In a second aspect, the present invention provides a fault monitoring device for an aircraft engine, the device comprising: a signal acquisition module for acquiring vibration signals from various parts of the aircraft engine during operation; a signal classification module for dividing the vibration signal into a steady-state signal, a periodic impact signal or a non-steady-state transient signal based on short-time energy and zero-crossing rate; a first fault judgment module for determining whether a fault exists by using time domain feature extraction and a classification model when the vibration signal is a steady-state signal; a second fault judgment module for determining whether a fault exists by using Fourier transform and envelope spectrum analysis when the vibration signal is a periodic impact signal; and a third fault judgment module for determining whether a fault exists by using time-frequency analysis when the vibration signal is a non-steady-state transient signal.

[0022] In a third aspect, the present invention provides a computer device comprising: a memory and a processor, the memory and the processor being communicatively connected to each other, the memory storing computer instructions, and the processor executing the aircraft engine fault monitoring method of the first aspect or any corresponding embodiment thereof by executing the computer instructions.

[0023] In a fourth aspect, the present invention provides a computer-readable storage medium having computer instructions stored thereon, the computer instructions being used to enable a computer to execute the aircraft engine fault monitoring method of the first aspect or any corresponding embodiment thereof.

[0024] In a fifth aspect, the present invention provides a computer program product comprising computer instructions for causing a computer to execute the method for monitoring faults of an aircraft engine according to the first aspect or any corresponding embodiment thereof. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the specific embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0026] Figure 1 is a flow chart of a method for monitoring a fault of an aircraft engine according to an embodiment of the present invention;

[0027] Figure 2 is a structural block diagram of a fault monitoring device for an aircraft engine according to an embodiment of the present invention;

[0028] Figure 3 1 is a schematic diagram of the working process of the fault monitoring device for an aircraft engine according to an embodiment of the present invention;

[0029] Figure 4 is a flowchart of an implementation method of a signal acquisition module according to an embodiment of the present invention;

[0030] Figure 5 is a flowchart of an implementation method of a preprocessing module according to an embodiment of the present invention;

[0031] Figure 6 is a flowchart of an implementation method of a signal classification module according to an embodiment of the present invention;

[0032] Figure 7 is a flowchart of an implementation method of a data mining module according to an embodiment of the present invention;

[0033] Figure 8 is a flowchart of an implementation method of an alarm and display module according to an embodiment of the present invention;

[0034] Figure 9 Schematic diagram of the hardware structure of a computer device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0035] As described in the background technology, the traditional fault monitoring method of aircraft engines uses a single analysis algorithm and has the problem of a high false alarm rate. Specifically, under the actual operating conditions of aircraft engines, the vibration signal has multimodal characteristics, and a single algorithm is difficult to take into account the feature extraction requirements of different signal types. For example, the time domain resolution of spectrum analysis for transient signals is insufficient, and the time domain statistical method has low sensitivity to the fault frequency of periodic shocks, resulting in an increased false alarm rate (especially in complex background noise scenarios). In addition, existing systems generally use a fixed threshold alarm mechanism, which only prompts abnormalities through unified sound and light signals, lacks a hierarchical judgment of the fault type and a targeted response strategy, making it difficult for maintenance personnel to locate the source of the fault in a timely manner, and prolonging the emergency response time.

[0036] Furthermore, with the rise of multimodal data mining technology in industrial equipment monitoring, it integrates signal processing and machine learning algorithms to achieve adaptive feature extraction under complex operating conditions. For example, methods for classifying rotating machinery vibration signals based on wavelet packet decomposition fail to consider the real-time switching of signal modalities under dynamic conditions. Existing methods still face bottlenecks in algorithm switching latency (typically >200ms) and the efficiency of multi-source feature fusion, making them unable to meet the real-time requirements of high-speed aircraft engine applications (>10,000 rpm).

[0037] Therefore, the existing technology has the following core defects:

[0038] 1. Insufficient algorithm adaptability: A single analysis method cannot cover the characteristics of multi-modal vibration signals, resulting in missed detection of key fault information or misjudgment due to noise interference;

[0039] 2. Real-time bottleneck: Traditional signal classification and algorithm matching processes have computational delays and cannot meet the millisecond-level response requirements of high-speed engine operation.

[0040] 3. Rigid alarm mechanism: Lack of abnormality type classification and visual decision-making support functions restricts fault handling efficiency.

[0041] In view of this, the present embodiment provides a method for fault monitoring of an aircraft engine, which integrates dynamic signal classification, adaptive algorithm matching, and intelligent monitoring of a multi-level alarm mechanism to achieve accurate mining and rapid response of aircraft engine vibration signals.

[0042] To make the purpose, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without making creative efforts shall fall within the scope of protection of the present invention.

[0043] According to an embodiment of the present invention, an embodiment of a fault monitoring method for an aircraft engine is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0044] In this embodiment, a method for monitoring faults of an aircraft engine is provided, which can be used in electronic devices such as computers, mobile phones, tablet computers, etc. Figure 1 FIG. 1 is a flow chart of a method for monitoring a fault of an aircraft engine according to an embodiment of the present invention. Figure 1 As shown, the process includes the following steps:

[0045] Step S101, obtaining vibration signals from various parts of the aircraft engine during operation. Specifically, in actual applications, sensors can be installed at key parts of the aircraft engine to collect vibration signals from corresponding parts. In order to facilitate the accuracy of the collected data, this embodiment uses high-precision vibration sensors such as piezoelectric accelerometers and other devices to collect vibration signals at a certain sampling frequency. For example, the sampling frequency can be no less than 20kHz, so as to obtain more data points and improve data quality. The key parts for installing vibration sensors can be determined according to actual conditions. For example, they can be installed at parts such as bearings, blades, and rotors. In addition, in order to facilitate the analysis of transient events, after obtaining the vibration signal, it is first divided into time windows, that is, the vibration signal is segmented and processed, and then the vibration signal in each window is separately judged for faults.

[0046] Step S102: Classify the vibration signal into a steady-state signal, a periodic impact signal, or a non-steady-state transient signal based on short-time energy and zero-crossing rate. Specifically, short-time energy (STE) represents the sum of the squares of the amplitude of the signal within a small interval (a certain time window), which is used to measure the energy intensity of the signal within the time window. The short-time energy is calculated using the following formula:

[0047]

[0048] Where x k (n) represents the nth sampling point of the kth signal window; N represents the window length, (such as N = f s ×T, T = 50ms, corresponding to N = 20kHz × 0.05s = 1000 points, f s represents the sampling frequency, and T represents the sampling period).

[0049] The zero-crossing rate (ZCR) indicates the number of times a signal crosses zero within a window, reflecting the signal's frequency component. The zero-crossing rate (ZCR) is calculated using the following formula:

[0050]

[0051] Where sign(x) represents the sign function (1 if x > 0, -1 otherwise); the denominator 2N is the normalization coefficient, which ensures that the ZCR range is [0, 1].

[0052] Vibration signals typically include multiple modalities, including steady-state signals, periodic impact signals, and non-steady-state transient signals. Steady-state signals include low-frequency vibrations caused by bearing wear, shaft misalignment, and poor lubrication. Periodic impact signals include periodic impact faults caused by sudden increases in harmonic components due to blade breakage, gear tooth breakage, and rolling element spalling. Non-steady-state transient signals include sudden changes in time-frequency energy caused by transient anomalies such as surge and combustion instability. Due to these characteristics, different types of signals have different short-term energy and zero-crossing rates. Therefore, short-term energy and zero-crossing rate can be used to classify vibration signals within each window.

[0053] In step S103, if the vibration signal is steady-state, a time-domain feature extraction and classification model are used to determine whether a fault exists. Specifically, for steady-state signals, time-domain features such as peak values ​​are first extracted. Based on the extracted features, a predetermined classification model is used to determine whether a fault has occurred within the time window.

[0054] Step S104: When the vibration signal is a periodic impact signal, Fourier transform and envelope spectrum analysis are used to determine whether a fault exists. Specifically, for periodic impact signals, it is necessary to extract periodic impact features from them for fault judgment. Specifically, in this embodiment, Fourier transform is used to first locate characteristic frequencies that may be strongly correlated with the fault (that is, it can be understood that the fault will cause energy anomalies at a specific frequency), and then envelope spectrum analysis is used to extract the impact envelope, that is, to focus on the impact periodic features; finally, the presence of a fault is determined based on the relationship between the peak value of the impact envelope and the characteristic frequency.

[0055] In step S105, if the vibration signal is a non-steady-state transient signal, time-frequency analysis is used to determine whether a fault exists. Specifically, non-steady-state transient signals are short-lived, sudden, and have variable frequencies. This embodiment utilizes the time-frequency focusing capability of wavelet transform to accurately locate the time point and frequency range of the abnormality. Therefore, time-frequency analysis can accurately determine whether a fault has occurred.

[0056] The aircraft engine fault monitoring method provided by an embodiment of the present invention first classifies vibration signals based on short-time energy and zero-crossing rate, and then processes the vibration signals of different types using corresponding data mining algorithms, thereby achieving dynamic matching of signal features. In addition, compared with the use of a single analysis algorithm, this method performs fault monitoring based on the matching of different algorithms based on the characteristics of the vibration signal, thereby improving monitoring accuracy and reducing the false alarm rate.

[0057] In this embodiment, a method for monitoring faults of an aircraft engine is provided, and the process includes the following steps:

[0058] Step S201: Obtain the parts of the aircraft engine to be monitored and the types of sensors used for vibration signal acquisition. The parts include the compressor, turbine, rotor bearings, combustion chamber, tail nozzle, and engine mounting bracket. The sensor types include piezoelectric accelerometers, micro-electromechanical system accelerometers, fiber grating vibration sensors, and capacitive vibration sensors. The type of sensor installed in each part is determined based on the environment in which each part is located and the frequency band of the vibration signal generated by each part.

[0059] Specifically, to accurately collect vibration signals from various parts of an aircraft engine, the vibration sensors installed in the aircraft engine must meet the requirements of high temperature, high-frequency response, high sensitivity, and strong anti-interference capabilities. Furthermore, the sensors must support analog (4-20mA) or digital (IEPE) output to facilitate integration with the acquisition system. For this reason, the vibration sensors used in this embodiment include piezoelectric accelerometers, microelectromechanical system (MEMS) accelerometers, fiber Bragg grating (FBG) vibration sensors, and capacitive vibration sensors.

[0060] Among them, the piezoelectric accelerometer uses the piezoelectric effect of piezoelectric materials (such as quartz and ceramics) to convert mechanical vibrations into electrical signals during operation, thereby realizing the collection of vibration signals. This accelerometer has a high-frequency response (0.5Hz to 20kHz), which is suitable for capturing transient impact signals, high sensitivity (such as 100mV / g), excellent signal-to-noise ratio, and a small size that is easy to embed. Therefore, this accelerometer can be installed in high-frequency vibration areas such as engine compressors and turbine blades; it is also suitable for working in high-temperature environments (it needs to be equipped with a high-temperature resistant housing, and the operating temperature range can reach -50℃ to +260℃). For example, this accelerometer can be installed in the engine bearing seat to monitor vibration signals generated by rotor imbalance or abnormal gear meshing.

[0061] The core principle of a micro-electro-mechanical system (MEMS) accelerometer is to output an electrical signal by causing capacitance changes through the deformation of a micromechanical structure (such as a cantilever beam). This accelerometer features a wide bandwidth (DC to 5kHz), making it suitable for steady-state signal monitoring. It also boasts high integration, can be directly embedded in a data acquisition board, and is low-cost and suitable for multi-point deployment. Therefore, this accelerometer can be installed in medium- and low-frequency vibration areas, such as engine casings and accessory gearboxes, or in applications requiring low-power, integrated monitoring. For example, this accelerometer can be used to monitor the overall vibration level of an engine and assist in identifying problems such as loose mounting bases.

[0062] The core operating principle of a fiber Bragg Grating (FBG) vibration sensor is to use the wavelength offset of the fiber Bragg Grating (FBG) to reflect strain changes caused by external vibration. This vibration sensor features high-temperature resistance (up to 600°C), making it suitable for extreme environments; long-distance transmission without signal attenuation; and the ability to multiplex multiple sensors in series, reducing wiring complexity. Therefore, this vibration sensor is suitable for high-temperature and high-pressure environments (such as near the turbine end) or for installation in areas with strong electromagnetic interference (due to the inherent electromagnetic interference resistance of optical fiber). For example, this vibration sensor can be deployed at the root of a turbine blade to monitor blade vibration and crack growth at high temperatures.

[0063] The core operating principle of a capacitive vibration sensor is to detect changes in the spacing between capacitor plates caused by vibration and output an electrical signal. This vibration sensor features high resolution (capable of detecting nanometer-level displacements), a wide dynamic range (0.1Hz to 1kHz), high stability, and minimal long-term drift. Therefore, this vibration sensor can be used for low-frequency micro-vibration monitoring (such as during engine idling) or installed in applications requiring high linearity and high resolution. For example, this vibration sensor can be used to monitor low-frequency vibration anomalies during engine startup and shutdown.

[0064] Based on the different types of vibration sensors mentioned above, when determining the vibration sensors installed in different parts, the following characteristics of the sensors can be combined to determine: first, the frequency response range, that is, the selected vibration sensor can cover the characteristic frequency band of the engine vibration signal (such as 0.1Hz~20kHz); second, environmental adaptability, that is, determining whether the corresponding part requires the vibration sensor to have characteristics such as high temperature resistance, oil pollution resistance or impact resistance; third, sensitivity and range, that is, the selected vibration sensor can match the vibration amplitude range (such as 0.1g~1000g).

[0065] In this embodiment, the vibration sensors installed at various parts of the engine are shown in Table 1 below:

[0066] Table 1

[0067] Engine parts Sensor Type Monitoring targets compressor rotor Piezoelectric accelerometer Blade passing frequency, rotor imbalance turbine disc Fiber Bragg Grating (FBG) Sensors Blade vibration and thermal fatigue at high temperatures Main bearing seat MEMS accelerometers Bearing wear and lubrication failure Engine mounting bracket Capacitive sensors Analysis of vibration transfer characteristics of the whole machine

[0068] In this invention, the vibration sensors installed in aircraft engines are not of a single type. Instead, a combination of sensors is used based on monitoring requirements and environmental conditions, forming a multimodal data acquisition network. By integrating piezoelectric, MEMS, fiber optic, and capacitive sensors, the full frequency range of vibration signals is covered, ensuring reliable monitoring even under complex operating conditions.

[0069] Step S202: Obtain vibration signals of various parts of the aircraft engine during operation. Figure 1 Step S101 of the illustrated embodiment will not be described in detail here.

[0070] Step S203: using a wavelet threshold denoising algorithm to remove high-frequency noise and low-frequency drift in the vibration signal; and performing normalization processing and signal segmentation processing on the vibration signal.

[0071] Specifically, among the wavelet bases, the Daubechies wavelet family (such as db6) has the characteristics of tight support and approximate symmetry, which makes it suitable for transient signal analysis. Therefore, this implementation uses the Daubechies wavelet family as the wavelet basis. After determining the wavelet basis, the number of decomposition levels is determined based on the signal's main frequency range. Specifically, the number of decomposition levels can be determined using the following formula:

[0072]

[0073] Where J represents the number of decomposition layers, f s represents the sampling rate, f max Indicates the highest effective frequency of the signal).

[0074] Then the selected wavelet basis is used to perform J-layer discrete wavelet transform on the original signal, i.e. the obtained vibration signal, to obtain the approximate coefficients a of each layer. j and detail coefficient d j Because the obtained detail coefficients contain the transient details of the real signal and the random fluctuations of the noise, it is necessary to denoise the detail coefficients. Specifically, the following soft threshold function can be used to denoise the detail coefficients:

[0075]

[0076] In the formula, λ represents the universal threshold, which can be expressed as Determine, where σ is the noise standard deviation (through the highest layer detail coefficient d J Estimate), where N is the signal length. This universal threshold can distinguish between noise and signal details. Therefore, after processing the detail coefficients using the above formula, denoising is achieved. Finally, an inverse wavelet transform is performed using the denoised detail coefficients and the approximate coefficients obtained by layering to denoise the vibration signal.

[0077] Normalizing the denoised vibration signal can eliminate the effects of sensor range differences and amplitude fluctuations, facilitating unified processing by subsequent algorithms. During signal normalization, the signal is first linearly scaled. Specifically, the following formula is used to normalize the signal amplitude to the [-1, 1] range:

[0078]

[0079] If the signal dynamic range is too large, the following formula can be used for standard deviation normalization:

[0080]

[0081] Where μ is the mean and σ is the standard deviation. After linear amplitude scaling, a threshold (e.g., ±3σ) is set to forcibly truncate out-of-range amplitudes to the boundary value, i.e., perform outlier truncation to prevent normalization distortion.

[0082] In addition, the acquired vibration signal is usually a continuous signal, so it is further divided into analysis windows of fixed length to meet the needs of real-time monitoring and algorithm batch processing. First, the window is divided using a preset window length and overlap rate. The window length is set according to the characteristic frequency of the signal, for example, it can be 1 second, and the corresponding number of sampling points in a window is N = f s × T = 20 kHz × 1 s = 20,000 points. Furthermore, to reduce edge effects, a 50% overlap ratio can be used, meaning adjacent windows overlap by N / 2 points. After windowing, a Hanning window is applied to the signal in each window based on the weight w(n) calculated using the following formula to suppress spectral leakage:

[0083]

[0084] After that, the amplitude of the signal processed by the Hanning window is corrected using the window energy compensation coefficient calculated by the following formula:

[0085]

[0086] Specifically, in practical applications, a ring buffer can be used to store the latest N points of data, triggering segment processing every time N / 2 points are reached, thereby achieving real-time processing. Alternatively, historical data can be segmented using a fixed-step sliding window to achieve offline processing.

[0087] Step S204 : classifying the vibration signal into a steady-state signal, a periodic impact signal, or a non-steady-state transient signal based on the short-time energy and the zero-crossing rate.

[0088] Specifically, the above step S204 includes:

[0089] Step S2041 : when the short-time energy is less than or equal to the short-time energy baseline and the zero-crossing rate is less than or equal to a first zero-crossing rate threshold, classify the vibration signal as a steady-state signal.

[0090] Step S2042: When the short-time energy is greater than the short-time energy baseline and meets the periodic detection condition, the vibration signal is classified as a periodic impact signal.

[0091] Step S2043: When the short-time energy is greater than the transient event energy threshold and the zero-crossing rate is greater than the second zero-crossing rate threshold, the vibration signal is classified as a non-steady-state transient signal, the transient event energy threshold is greater than the short-time energy baseline, and the second zero-crossing rate threshold is greater than the first zero-crossing rate threshold.

[0092] Different types of vibration signals have different short-term energy and zero-crossing rates. Specifically, steady-state signals are characterized by low-frequency steady-state vibrations and a dominant low-frequency component, resulting in low short-term energy and a low zero-crossing rate. Periodic shock signals, as shock events, contain high-frequency components, resulting in periodic surges in short-term energy and a moderate zero-crossing rate. Non-steady-state transient signals, as transient events containing broadband noise, exhibit sudden surges in short-term energy and a high zero-crossing rate. Therefore, vibration signals can be classified based on their short-term energy and zero-crossing rate within each window.

[0093] Before classification, it is necessary to determine the thresholds for dividing short-time energy and zero-crossing rate. For short-time energy, two thresholds need to be determined: the short-time energy baseline (steady-state signal energy baseline) and the transient event energy threshold. Among them, the 90% quantile of the historical data (historical vibration signal) STE can be taken as the steady-state signal energy baseline STE. base ; 3×STE base As the transient event energy threshold STE thresh_high For the zero crossing rate, the first zero crossing rate ZCR is determined low and the second zero-crossing rate ZCR high The two thresholds, the sizes of the two zero-crossing rate thresholds can be determined according to the actual situation, such as ZCR low =0.1, ZCR high =0.3. In addition, the threshold value determined here can be updated in real time according to the acquired data, for example, the STE can be recalculated every 24 hours according to the latest data. base With ZCR low 、ZCR high , adapt to changes in working conditions.

[0094] In determining the STE of the current window k With ZCR k After that, the following process can be used to divide the signal:

[0095] If STE k ≤STE base And ZCR k ≤ZCR low , the vibration signal of the current window is classified as Class A signal (steady-state signal).

[0096] If STE k >STE baseIf the periodicity detection passes, the vibration signal in the current window is classified as a Class B signal (periodic impact signal): wherein the method for periodicity detection of the signal includes: performing autocorrelation analysis on the STE sequence, and if there is a significant periodic peak (peak height > 0.7 times the maximum autocorrelation value), it is determined to be periodic.

[0097] If STE k >STE thresh_high And ZCR k >ZCR high , the vibration signal of the current window is classified as a Class C signal (non-steady transient signal).

[0098] It should be noted that the length of the window division in step S203 can also be determined based on considerations of STE and ZCR. If the window time division is short, such as 10ms, ZCR statistics will be unstable and STE will be susceptible to noise interference. If the window time division is too long, such as 200ms, transient events (such as surge duration of approximately 50ms) will not be captured. Based on this, this embodiment sets the window time division length to T = 50ms, thereby achieving a balance between characteristic stability and transient response.

[0099] Step S205: When the vibration signal is a steady-state signal, use time domain feature extraction and classification model to determine whether there is a fault. Figure 1 Step S103 of the illustrated embodiment will not be described in detail here.

[0100] Step S206: When the vibration signal is a periodic impact signal, Fourier transform and envelope spectrum analysis are used to determine whether there is a fault. Figure 1 Step S104 of the illustrated embodiment will not be described in detail here.

[0101] Step S207: When the vibration signal is a non-steady transient signal, time-frequency analysis is used to determine whether there is a fault. Figure 1 Step S105 of the illustrated embodiment will not be described in detail here.

[0102] Step S208: When it is determined that there is a fault, the buzzer is controlled to alarm at a preset frequency and the red light is controlled to flash; when it is determined that there is no fault, the buzzer is controlled to be silent and the green light is controlled to be always on.

[0103] Specifically, in order to realize the fault alarm function of the present embodiment, it is necessary to design hardware and software. Among them, the hardware components used include a microcontroller (such as an embedded development platform such as Arduino, STM32 or Raspberry Pi), a buzzer (supporting PWM control, frequency 800Hz), a red LED and a green LED (with a current limiting resistor of 220Ω), a signal input interface (such as GPIO or ADC interface, for receiving abnormal signals) and a communication module (such as a USB serial port, Wi-Fi / Bluetooth module, for communicating with a host computer). For the above-mentioned hardware components, it is specifically connected in the following manner: the buzzer is connected to the PWM output pin of the microcontroller (such as Pin9 of Arduino). The red LED is connected to the GPIO output pin (such as Pin10), and a 220Ω resistor is connected in series to ground. The green LED is connected to the GPIO output pin (such as Pin11), and a 220Ω resistor is connected in series to ground. The abnormal detection signal input is connected to the GPIO input pin (such as Pin2) or the ADC pin (analog signal).

[0104] In addition to the above hardware components, software design is also required. Specifically, this software design includes microcontroller firmware development and host computer interface development. The development environment for microcontroller firmware development can be ArduinoIDE, PlatformIO, or STM32CubeIDE. The specific development logic process includes:

[0105] while True:

[0106] If an abnormal signal is detected:

[0107] Turn off the green LED

[0108] Start buzzer (800Hz, 0.2 seconds sound, 0.5 seconds silent cycle)

[0109] Start red LED (0.25 seconds on, 0.25 seconds off, cycle 0.5 seconds)

[0110] Send abnormal data to the host computer (including abnormal level)

[0111] else:

[0112] Turn off the buzzer and red LED

[0113] Turn on the green LED (steady on)

[0114] Send normal data to the host computer

[0115] The host computer interface is developed using Python (PyQt5 / Tkinter + Matplotlib) or LabVIEW. The developed host computer features the following functions: First, displaying time-frequency graphs, such as using FFT to transform real-time signal data and dynamically updating it using Matplotlib. Second, providing abnormality levels and action suggestions, matching received data with pre-set rules (e.g., Level 1 - Minor abnormality → Recommendation to inspect equipment). Furthermore, the host computer receives data via the serial port (UART) during operation, using the following format: {status:"abnormal",level:2}.

[0116] For the hardware structure and software design determined above, debugging and testing are required to ensure their normal operation. The debugging process specifically includes the following steps:

[0117] Step a1: Hardware setup. Solder the buzzer, LED, and current-limiting resistor to the microcontroller development board. Also connect the abnormal signal input source (such as a sensor or analog signal generator).

[0118] Step a2: Firmware debugging. Test the buzzer frequency (verify the 800Hz waveform with an oscilloscope) and calibrate the LED flashing period (adjust delay() or use a timer interrupt).

[0119] Step a3: Develop the host computer. Write a serial data parsing module. Implement a dynamic time-frequency graph (using Matplotlib animation). Design the UI layout (displaying charts, anomaly levels, and handling suggestions by region).

[0120] Step a4: System integration. Simulate abnormal signals to verify the synchronous response of the buzzer, LED, and interface. Optimize communication delays (to ensure data real-time).

[0121] After completing the above debugging process, any issues that arise during debugging should be resolved before testing and verification. For issues with buzzer and LED synchronization, use hardware timer interrupts (non-blocking code) to avoid lags caused by delay() calls. For lags in time-frequency graphs, reduce the data sampling rate or use a high-efficiency plotting library (such as PyQtGraph). For packet loss, add a checksum (such as CRC) and a retransmission mechanism.

[0122] After resolving any issues during debugging, you can perform testing and verification. First, perform a normal state test. This involves monitoring the vibration signal to confirm that no faults exist. Then, determine if the green light remains on, the buzzer is silent, and the interface displays "Normal." Next, perform an abnormal state test. This involves monitoring the vibration signal to confirm that a fault exists. Then, determine if the red light flashes, the buzzer sounds at a set frequency, and the interface updates the time-frequency graph and abnormality information. Additionally, a stress test can be performed, which continuously triggers abnormal states to verify system stability.

[0123] In this embodiment, a method for monitoring faults of an aircraft engine is provided, the method comprising the following steps:

[0124] Step S301: Obtain vibration signals of various parts of the aircraft engine during operation; see Figure 1 Step S101 of the illustrated embodiment will not be described in detail here.

[0125] Step S302: classify the vibration signal into a steady-state signal, a periodic impact signal, or a non-steady-state transient signal based on the short-time energy and the zero-crossing rate; Figure 1 Step S102 of the illustrated embodiment will not be described in detail here.

[0126] Step S303: When the vibration signal is a steady-state signal, time domain feature extraction and classification model are used to determine whether a fault exists.

[0127] Specifically, the above step S303 includes:

[0128] Step S3031: extract the peak value, kurtosis, impulse factor, and root mean square of the vibration signal as time domain features. The peak value of the vibration signal x(t) is extracted using the following formula:

[0129] Peak=max(|x(t)|)

[0130] The following formula is used to extract the kurtosis K of the vibration signal:

[0131]

[0132] Where N is the number of sampling points in the window, xi is the i-th signal in the window, μ is the mean of the signal in the window, and σ is the standard deviation of the signal in the window. This kurtosis further characterizes the signal pulse strength. Under normal conditions, K is approximately 0, while under abnormal conditions, K is greater than 3.

[0133] The pulse factor IF of the vibration signal is extracted using the following formula:

[0134]

[0135] The root mean square is extracted using the following formula:

[0136]

[0137] Step S3032: normalize the time domain features. Specifically, the following formula may be used to perform Z-score normalization on the extracted time domain features F = [Peak, K, IF, RMS]:

[0138]

[0139] In step S3033, the normalized time-domain features are input into a support vector machine classifier to obtain an output indicating whether a fault is present. Specifically, before using the support vector machine classifier for classification, the kernel function employed is determined and trained. This embodiment employs the radial basis function (RBF) kernel shown in the following formula as the kernel function, which can adapt to nonlinear classification boundaries.

[0140] K(F i , F j )=exp(-γ||F i -F j || 2 )

[0141] When training the support vector machine, the penalty coefficient and kernel parameters are mainly optimized. The penalty coefficient can be determined by hyperparameter search. The kernel parameter γ can be selected based on cross-validation. In practical applications, the extracted time domain features can be input into the trained support vector machine classifier to output the probability of whether there is a fault. The fault probability P fault ∈[0,1], if P fault >0.8, it is judged as abnormal, that is, there is a fault.

[0142] Step S304: When the vibration signal is a periodic impact signal, Fourier transform and envelope spectrum analysis are used to determine whether a fault exists.

[0143] Specifically, the above step S304 includes:

[0144] Step S3041, perform Fourier transform on the vibration signal to determine the characteristic frequency; specifically, perform fast Fourier transform (FFT) on the vibration signal to convert the time domain signal into the frequency domain to obtain the spectrum X(f), which can intuitively show the distribution of different frequency components in the signal. The theoretical characteristic frequency is then determined for possible fault conditions of different structures. For example, for turbine blades, the blade passing frequency (i.e., the theoretical characteristic frequency) can be determined by its fundamental frequency and the number of blades. Then, based on the theoretical characteristic frequency, observe in the spectrum X(f) whether there is an obvious peak near the frequency. If so, determine that the peak corresponds to the blade passing frequency. In addition, for the frequency resolution of the spectrum, the calculation formula is Δf=f s / N, where f s In practical applications, the frequency resolution can be adjusted based on this formula so that the peak corresponding to the theoretical characteristic frequency can be accurately observed in the spectrum.

[0145] Step S3042: Extract the envelope of the vibration signal and perform Fourier transform. Obtain the envelope spectrum. Specifically, when a fault occurs, such as a bearing or gear crack, there will be an impact component in the vibration signal, but a direct FFT may be masked by other frequencies. Therefore, envelope extraction is performed first, and then the envelope signal is FFTed to highlight the periodicity of the impact. The Hilbert transform can be used to extract the signal envelope. Where, represents the Hilbert transform; this step extracts the amplitude variation pattern of the impulse signal and filters out carrier frequency interference. The extracted envelope is then Fourier transformed, meaning that the envelope signal e(t) is subjected to an FFT to obtain the envelope spectrum E(f). The peaks in the spectrum correspond to the periodic frequencies of the impulse signal.

[0146] Step S3043, based on the relationship between the peak value that exceeds 3 times the standard deviation of the baseline and the characteristic frequency and its multiples in the envelope spectrum, determine whether there is a fault. Specifically, for the obtained envelope spectrum, first determine the baseline of the envelope spectrum based on the mean and standard deviation of the amplitude of all frequency points in the envelope spectrum, where μ+3σ (μ represents the mean and σ represents the standard deviation) can be used to represent the baseline; then determine the peak value whose amplitude exceeds the baseline in the envelope spectrum. Determine whether the peak value corresponds to the peak value of the theoretical characteristic frequency and the multiples of the theoretical characteristic frequency (i.e., perform harmonic matching). For example, if the theoretical characteristic frequency is 4000Hz, and there are peak values ​​exceeding the baseline at both 4000Hz and 8000Hz in the envelope spectrum, it is determined that there is a fault.

[0147] Step S305: When the vibration signal is a non-steady-state transient signal, time-frequency analysis is used to determine whether a fault exists.

[0148] Specifically, the above step S305 includes:

[0149] Step S3051: Continuous wavelet transform is used to determine the wavelet coefficients of the vibration signal at different scales and time points. Specifically, when using continuous wavelet transform, wavelet basis selection and scale selection are first performed. In this embodiment, Morlet wavelet is used as the wavelet basis, thereby taking into account both time and frequency resolution. The Morlet wavelet is expressed using the following formula:

[0150]

[0151] For scale selection, it needs to cover the signal main frequency range. For example, selecting scale a∈[0.1,10] corresponds to f=f c / (aΔt),f c is the wavelet center frequency). Then, based on the selected wavelet basis and scale, the wavelet coefficients are calculated using the following formula:

[0152]

[0153] Where b represents time and x(t) represents the vibration signal.

[0154] Step S3052: Determine the dynamic threshold based on the modulus square, mean, and standard deviation of the wavelet coefficients. Specifically, transient anomalies such as surge are short-term and sudden, and can cause energy mutations in multiple frequency ranges simultaneously. Based on this, this embodiment uses the modulus square of the wavelet coefficients |W(a,b)| 2 The energy distribution of scale a-time b is obtained, i.e., the energy matrix. Then, the local mean and standard deviation of each time point in the energy distribution are calculated using a sliding window, and the dynamic threshold is determined using the following formula:

[0155] Threshold(a,b)=μ |W| +k·σ |W|

[0156] In the formula, k is a coefficient, which can be 3, μ ∣W∣ represents the local mean and σ ∣W∣ Represents standard deviation.

[0157] Step S3053 , determining whether a fault exists based on whether multiple scales at the same time point exceed the dynamic threshold; specifically, when the energy corresponding to multiple different scales a simultaneously exceeds the threshold at a certain time point, such as b0, it is determined to be a transient event, i.e., a fault exists.

[0158] In step S3054, the time-frequency energy distribution obtained by the Hilbert-Huang transform is used to verify the existence of the fault. Specifically, when using the Hilbert-Huang transform, the signal is first decomposed into IMF (Intrinsic Mode Function) components through empirical mode decomposition (EMD), where each IMF is the fluctuation of a single frequency component. Afterwards, a Hilbert transform is performed on each IMF to obtain the instantaneous frequency-time-energy distribution (horizontal axis time, vertical axis frequency, color represents energy), i.e., the Hilbert spectrum. In the Hilbert spectrum, if the energy concentration area (color highlight area) of multiple IMFs near time b0 covers the multi-scale frequency range, it means that the time-frequency characteristics of the two are consistent, and it is confirmed to be a real transient anomaly.

[0159] This embodiment also provides an aircraft engine fault monitoring device for implementing the aforementioned embodiments and preferred implementations. Details already described will not be repeated. As used below, the term "module" may refer to a combination of software and / or hardware that implements a predetermined function. While the devices described in the following embodiments are preferably implemented using software, implementation using hardware, or a combination of software and hardware, is also possible and contemplated.

[0160] This embodiment provides a fault monitoring device for an aircraft engine, such as Figure 2 As shown, including:

[0161] The signal acquisition module 21 is used to acquire vibration signals of various parts of the aircraft engine during operation;

[0162] a signal classification module 22 for classifying the vibration signal into a steady-state signal, a periodic impact signal, or a non-steady-state transient signal based on short-time energy and zero-crossing rate;

[0163] A first fault judgment module 23 is used to determine whether there is a fault by using time domain feature extraction and classification model when the vibration signal is a steady-state signal;

[0164] A second fault judgment module 24 is used to determine whether a fault exists by using Fourier transform and envelope spectrum analysis when the vibration signal is a periodic impact signal;

[0165] The third fault judgment module 25 is configured to use time-frequency analysis to determine whether a fault exists when the vibration signal is a non-steady-state transient signal.

[0166] The further functional description of each of the above modules is the same as that of the above corresponding embodiments and will not be repeated here.

[0167] As a specific application example of the embodiment of the present invention, Figure 3As shown in the figure, the fault monitoring device of this aero-engine specifically includes the following modules:

[0168] The signal acquisition module is used to install high-precision vibration sensors at key parts of the aero-engine, such as the compressor, turbine, and bearing housing, etc., to collect vibration data (vibration signals).

[0169] The preprocessing module is used to input the original vibration signal. First, wavelet threshold denoising is adopted, and then the denoised signal is normalized, that is, the signal amplitude is scaled to the interval [-1, 1]. After that, the normalized signal is segmented, that is, the signal window is divided according to a fixed duration (such as 1 second), and the cleaned segmented signal is output.

[0170] The signal classification module is used to calculate the short-time energy and zero-crossing rate (ZCR) through feature extraction, and classify the signal by comparing ZCR with the threshold. For example, the ZCR thresholds are 1 and 2. Among them, if 1 < ZCR < 2, the vibration signal is classified as a Class B signal, that is, a periodic impact signal, which has a periodic sudden increase in energy; if ZCR < 1, the vibration signal is classified as a Class A signal, that is, a steady-state signal, which has low-energy fluctuations; if ZCR > 2, the vibration signal is classified as a Class C signal, that is, a non-steady-state transient signal, which has a high-energy short-time burst.

[0171] The data mining module is used to perform algorithm matching on the classified signals. For Class A signals, Algorithm 1 is adopted, that is, time-domain features (such as peak value and kurtosis, etc.) are extracted and input into the SVM classifier to determine whether there is a fault; for Class B signals, Algorithm 2 is adopted, that is, FFT and envelope spectrum analysis are used to extract and determine whether there is a fault by marking the fault frequency; for Class C signals, Algorithm 3 is adopted, that is, wavelet transform time-frequency is used, such as marking energy mutation detection; finally, an abnormal score (such as a 0-1 probability value) and a fault type label are output through the judgment result.

[0172] The alarm and display module is used to trigger an 800Hz buzzing sound (beeping for 0.2 seconds and cycling with a 0.5-second interval) and the red light flashes synchronously for 0.5 seconds when there is a fault, that is, an anomaly. When there is no fault, that is, it is normal, the green light is always on. For the display interface, a real-time time-frequency diagram (the horizontal axis is time and the vertical axis is frequency) is displayed in Area 1, and an LED status indicator (green / red) is in Area 2. An alarm level and maintenance suggestions (such as blade crack alarm, code B-203) are pop-up displayed in Area 3.

[0173] In an optional implementation manner, such as Figure 4As shown, the signal acquisition module is specifically used to collect vibration data by installing high-precision vibration sensors at aircraft engine vibration monitoring locations, such as the compressor, turbine, rotor bearing seat, combustion chamber, tail nozzle, engine mounting bracket, and other locations requiring vibration monitoring. These high-precision vibration sensors include piezoelectric accelerometers, micro-electromechanical system accelerometers, fiber Bragg grating (FBG) vibration sensors, capacitive vibration sensors, and other high-precision sensors. During installation, the sensor is determined based on the characteristics of the high-precision sensor and the location of the aircraft engine being monitored. After installation, the vibration sensor collects real-time vibration signals from key engine components at a sampling frequency of no less than 20kHz, and outputs the raw vibration signals after acquisition.

[0174] In an optional embodiment, as Figure 5 As shown in the figure, the preprocessing module is specifically used to: perform wavelet threshold denoising through wavelet basis selection, signal decomposition, threshold processing and signal reconstruction; then use linear amplitude scaling and outlier truncation processing to perform signal normalization processing, where ±3σ is used as the threshold. When the threshold is exceeded, the signal is truncated, and if it does not exceed the threshold, the signal is not truncated; finally, window division, windowing function and segmented data management (real-time mode and offline mode) are used to perform signal segmentation processing.

[0175] Among them, when the relevant parameter values ​​in the preprocessing module are as shown in Table 2 below, the implementation effect is evaluated and verified.

[0176] Table 2

[0177]

[0178]

[0179] When evaluating, the noise reduction effect is evaluated first. Specifically, the signal-to-noise ratio improvement is calculated using the following formula:

[0180]

[0181] Where x noisy represents the vibration signal before denoising, x clean It represents the vibration signal after denoising. When the signal-to-noise ratio improvement is greater than or equal to 10dB, it means that the noise reduction effect meets the requirements.

[0182] In addition, for the window divided by the vibration signal, a simulated fault pulse signal can be injected to verify whether the segmented window can complete the capture of transient events and perform segmented testing on them (if the pulse position deviation is less than 5ms, it means the test has passed).

[0183] In an optional embodiment, as Figure 6As shown, the signal classification module is specifically used to: input the segmented signal of the preprocessing module, calculate the short-time energy and zero-crossing rate, wherein the steady-state signal (type A signal) has the characteristics of being dominated by low-frequency steady-state vibration and low-frequency components, so the short-time energy is small and the zero-crossing rate is low; the periodic impact signal (type B signal) is an impact event, and contains high-frequency components in the periodic impact, so the short-time energy has a periodic surge, and the zero-crossing rate is medium; the non-steady-state transient signal (type C signal) is a transient event and contains wide-band noise, so the short-time energy has a sudden surge, and the zero-crossing rate is high. Therefore, the signal can be classified according to the size of the short-time energy and zero-crossing rate of the vibration signal in each window. For the short-time energy, it is necessary to determine the short-time energy baseline (steady-state signal energy baseline TE base ) and transient event energy threshold. base As the transient event energy threshold STE thresh_high For the zero crossing rate, the first zero crossing rate ZCR is determined low and the second zero-crossing rate ZCR high The two thresholds, the sizes of the two zero-crossing rate thresholds can be determined according to the actual situation, such as ZCR low =0.1, ZCR high =0.3. In addition, the threshold value determined here can be updated in real time according to the acquired data, for example, the STE can be recalculated every 24 hours according to the latest data. base With ZCR low 、ZCR high , adapt to changes in working conditions. If STE k ≤STE base And ZCR k ≤ZCR low , then the vibration signal of the current window is classified as Class A signal (steady-state signal). k >STE base If the periodicity test is passed, the vibration signal in the current window is classified as a Class B signal (periodic shock signal): wherein the method for periodicity detection of the signal includes: performing autocorrelation analysis on the STE sequence, and if there is a significant periodic peak (peak height > 0.7 times the maximum autocorrelation value), it is determined to be periodic. k >STE thresh_high And ZCR k >ZCR high , the vibration signal of the current window is classified as a Class C signal (non-steady transient signal).

[0184] In practical applications, the values ​​of various parameters in the signal classification module are shown in Table 3 below:

[0185] Table 3

[0186]

[0187] Specifically, the classification process of the signal classification module can be explained and verified by taking the aircraft engine bearing wear monitoring scenario as an example: first, the pre-processed segmented signal (Class A signal, 50ms window) is input; then the current window STE is calculated. k =0.05×10 -3 , ZCR k =0.08; then determine STE k <STE base (0.05×10 -3 <0.1×10 -3 ); ZCR k <ZCR low (0.08<0.1); thus, the signal is classified as Class A, triggering SVM time-domain feature analysis. This shows that the classification process is relatively accurate.

[0188] The classification effect of the above signal classification module can also be verified. The specific classification results are shown in Table 4 below:

[0189] Table 4

[0190]

[0191] In addition, the classification results can be used to draw a ROC (Receiver Operating Characteristic) curve. When the AUC (Area Under Curve) is greater than 0.95, the classification results are proven to be effective.

[0192] In an optional embodiment, as Figure 7 As shown, the data mining module is specifically used for: for Class A signals, the fault probability is determined based on time domain feature extraction, feature standardization, SVM classifier training and reasoning. If the fault probability is greater than 0.8, the fault type alarm is triggered, otherwise it is judged to be normal. The processing of this signal is suitable for scenarios such as low-frequency steady-state faults such as bearing wear, shaft misalignment, and poor lubrication; for Class B signals, whether to trigger the fault type alarm is determined based on fast Fourier transform, envelope spectrum analysis, fault frequency (characteristic frequency) identification, peak detection, and harmonic matching. The processing of this signal is suitable for periodic impact fault scenarios such as blade cracks, gear tooth breakage, and rolling element peeling. For Class C signals, whether there is a fault, that is, whether to trigger the fault type alarm, is determined based on continuous wavelet transform, time-frequency energy mutation detection, and Hilbert-Huang transform assisted verification (HHT verification).

[0193] Specifically, the processing can be explained with different application scenarios. When the vibration signal after classification is a Class A signal (such as bearing vibration data), the time domain features are extracted with kurtosis K = 4.2 and impulse factor IF = 5.0; then the features are input into the SVM classifier and the fault P is output. fault =0.92; then the bearing wear alarm is triggered. When the classified vibration signal is a Class B signal (turbine blade vibration data), Fourier transform analysis is performed to determine the fundamental frequency f rotor =200Hz, number of blades N blade =20, theoretical f BPF =4kHz; then perform envelope spectrum analysis and detect 2×f BPF =8kHz, the amplitude increases by 40%, triggering a blade crack alarm. When the classified vibration signal is a Class C signal (compressor surge data), a continuous wavelet transform is first performed to detect multiple scale energy surges (amplitude exceeding 5 times the threshold) within a 50ms window. HHT verification is then performed. If the first IMF component energy concentration area is consistent with the CWT result, the "surge warning" alarm is triggered.

[0194] The parameters of the data mining module are shown in Table 5 below:

[0195] Table 5

[0196]

[0197] Furthermore, various signal processing algorithms can be validated. For the Class A algorithm, using a full-life bearing data set, the SVM classification accuracy is ≥95% (F1-score); the fault warning time is ≥2 hours (based on kurtosis trend analysis). For the Class B algorithm, the sensitivity of envelope spectrum detection for gear tooth breakage faults is 98% (missing rate ≤2%); the frequency resolution error is ≤0.5Hz. The surge event detection delay is ≤10ms; and the time-frequency positioning accuracy (synchronized with high-speed video) is ≤1ms.

[0198] In an optional embodiment, as Figure 8 As shown, the alarm and display module is specifically used for: hardware design and software design. Hardware design includes determining the hardware component list and circuit connection, and software design includes microcontroller firmware development and host computer interface development. Based on the designed hardware and software, hardware construction, firmware debugging, and system joint debugging are carried out. Problems that arise during debugging, such as buzzer and LED synchronization, time-frequency graph freezes, and communication packet loss, are resolved before re-debugging. After debugging is completed, normal state testing, abnormal state testing, and stress testing are performed. If no problems are found in the tests, debugging is completed. The system is then used for alarm and display.

[0199] The embodiment of the present invention also provides a computer device having the above Figure 2 The fault monitoring device of the aircraft engine is shown.

[0200] See also Figure 9 , Figure 9 is a structural diagram of a computer device provided by an optional embodiment of the present invention, such as Figure 9 As shown, the computer device includes: one or more processors 10, memory 20, and interfaces for connecting various components, including high-speed interfaces and low-speed interfaces. Various components utilize different buses to communicate with each other and can be installed on a common mainboard or installed in other ways as needed. The processor can process the instructions executed in the computer device, including instructions stored in the memory or on the memory to display the graphical information of the GUI on an external input / output device (such as, a display device coupled to the interface). In some optional embodiments, if necessary, multiple processors and / or multiple buses can be used together with multiple memories and multiple memories. Equally, multiple computer devices can be connected, and each device provides part of the necessary operations (for example, as a server array, a group of blade servers, or a multi-processor system). Figure 9 A processor 10 is taken as an example.

[0201] The processor 10 may be a central processing unit, a network processor, or a combination thereof. The processor 10 may further include a hardware chip. The hardware chip may be an application-specific integrated circuit, a programmable logic device, or a combination thereof. The programmable logic device may be a complex programmable logic device, a field programmable gate array, a general purpose array logic, or any combination thereof.

[0202] The memory 20 stores instructions that can be executed by at least one processor 10, so as to enable at least one processor 10 to execute the method shown in the above embodiment.

[0203] The memory 20 may include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application required for at least one function; the data storage area may store data created based on the use of a computer device for displaying a small program landing page, etc. In addition, the memory 20 may include a high-speed random access memory, and may also include a non-transient memory, such as at least one disk storage device, a flash memory device, or other non-transient solid-state storage device. In some optional embodiments, the memory 20 may optionally include a memory remotely located relative to the processor 10, and these remote memories may be connected to the computer device via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and a combination thereof.

[0204] The memory 20 may include a volatile memory, such as a random access memory; the memory may also include a non-volatile memory, such as a flash memory, a hard disk or a solid-state drive; the memory 20 may also include a combination of the above types of memory.

[0205] The computer device further includes a communication interface 30 for the computer device to communicate with other devices or a communication network.

[0206] The embodiment of the present invention also provides a computer-readable storage medium. The above-mentioned method according to the embodiment of the present invention can be implemented in hardware, firmware, or implemented as a computer code that can be recorded in a storage medium, or implemented as a computer code that is originally stored in a remote storage medium or a non-temporary machine-readable storage medium and downloaded through a network and will be stored in a local storage medium, so that the method described herein can be stored in such software processing on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. Among them, the storage medium can be a magnetic disk, an optical disk, a read-only storage memory, a random access memory, a flash memory, a hard disk or a solid-state drive, etc.; further, the storage medium can also include a combination of the above-mentioned types of memory. It can be understood that a computer, a processor, a microprocessor controller or programmable hardware includes a storage component that can store or receive software or computer code. When the software or computer code is accessed and executed by a computer, a processor or hardware, the method shown in the above embodiment is implemented.

[0207] A portion of the present invention may be applied as a computer program product, such as a computer program instruction, which, when executed by a computer, can call or provide the method and / or technical solution according to the present invention through the operation of the computer. Those skilled in the art should understand that the form in which the computer program instruction exists in a computer-readable medium includes, but is not limited to, a source file, an executable file, an installation package file, etc. Accordingly, the way in which the computer program instruction is executed by the computer includes, but is not limited to: the computer directly executes the instruction, or the computer compiles the instruction and then executes the corresponding compiled program, or the computer reads and executes the instruction, or the computer reads and installs the instruction and then executes the corresponding installed program. Here, the computer-readable medium may be any available computer-readable storage medium or communication medium that can be accessed by the computer.

[0208] Although the embodiments of the present invention have been described with reference to the accompanying drawings, those skilled in the art may make various modifications and variations without departing from the spirit and scope of the present invention. Such modifications and variations are all within the scope defined by the appended claims.

Claims

1. A method for monitoring faults of an aircraft engine, characterized in that: The method comprises: Obtain vibration signals of various parts of the aircraft engine during operation; Classifying the vibration signal into a steady-state signal, a periodic impact signal, or a non-steady-state transient signal based on short-time energy and zero-crossing rate; When the vibration signal is a steady-state signal, time domain feature extraction and classification model are used to determine whether there is a fault; When the vibration signal is a periodic impact signal, Fourier transform and envelope spectrum analysis are used to determine whether there is a fault; When the vibration signal is a non-steady transient signal, time-frequency analysis is used to determine whether a fault exists.

2. The method according to claim 1, characterized in that Classifying the vibration signal into a steady-state signal, a periodic impact signal, or a non-steady-state transient signal based on short-time energy and zero-crossing rate includes: When the short-time energy is less than or equal to the short-time energy baseline and the zero-crossing rate is less than or equal to a first zero-crossing rate threshold, classifying the vibration signal as a steady-state signal; When the short-time energy is greater than the short-time energy baseline and the periodic detection condition is met, the vibration signal is classified as a periodic impact signal; When the short-time energy is greater than a transient event energy threshold and the zero-crossing rate is greater than a second zero-crossing rate threshold, the vibration signal is classified as a non-steady-state transient signal, the transient event energy threshold is greater than the short-time energy baseline, and the second zero-crossing rate threshold is greater than the first zero-crossing rate threshold.

3. The method according to claim 1, characterized in that When the vibration signal is a steady-state signal, time domain feature extraction and classification models are used to determine whether a fault exists, including: Extract the peak value, kurtosis, impulse factor and root mean square of the vibration signal as time domain features; performing normalization processing on the time domain features; The normalized time domain features are input into the support vector machine classifier to obtain the output result of whether there is a fault.

4. The method according to claim 1, wherein When the vibration signal is a periodic impact signal, Fourier transform and envelope spectrum analysis are used to determine whether a fault exists, including: Perform Fourier transform on the vibration signal to determine the characteristic frequency; Extract the envelope of the vibration signal and perform Fourier transform to obtain the envelope spectrum; Determine whether a fault exists based on the relationship between the peak value exceeding 3 times the standard deviation of the baseline and the characteristic frequency and its multiples in the envelope spectrum.

5. The method according to claim 1, wherein When the vibration signal is a non-steady-state transient signal, time-frequency analysis is used to determine whether a fault exists, including: Continuous wavelet transform is used to determine the wavelet coefficients of vibration signals at different scales and time points; Determine dynamic thresholds based on the modulus square, mean, and standard deviation of wavelet coefficients; determining whether a fault exists based on whether multiple metrics at the same time point exceed the dynamic threshold; The time-frequency energy distribution obtained by Hilbert-Huang transform is used to verify the existence of faults.

6. The method according to claim 1, characterized in that Before obtaining vibration signals of various parts of the aircraft engine during operation, the method further includes: Obtain the parts of the aircraft engine to be monitored and the types of sensors used for vibration signal acquisition. The parts include the compressor, turbine, rotor bearings, combustion chamber, tail nozzle, and engine mounting bracket. The sensor types include piezoelectric accelerometers, micro-electromechanical system accelerometers, fiber Bragg grating vibration sensors, and capacitive vibration sensors. The type of sensor installed at each part is determined based on the environment in which each part is located and the frequency band of the vibration signal generated by each part.

7. The method according to claim 1, characterized in that After obtaining vibration signals of various parts of the aircraft engine during operation, the method further includes: Wavelet threshold denoising algorithm is used to remove high-frequency noise and low-frequency drift in vibration signals; The vibration signal is normalized and segmented.

8. The method according to claim 1, characterized in that The method further comprises: When a fault is detected, the buzzer will sound an alarm at a preset frequency and the red light will flash. When it is determined that there is no fault, the control buzzer is silent and the control green light is always on.

9. A fault monitoring device for an aircraft engine, characterized in that: The device comprises: A signal acquisition module is used to obtain vibration signals of various parts of the aircraft engine during operation; A signal classification module, configured to classify the vibration signal into a steady-state signal, a periodic impact signal, or a non-steady-state transient signal based on short-time energy and zero-crossing rate; The first fault judgment module is used to determine whether there is a fault by using time domain feature extraction and classification model when the vibration signal is a steady-state signal; The second fault judgment module is used to determine whether there is a fault by using Fourier transform and envelope spectrum analysis when the vibration signal is a periodic impact signal; The third fault judgment module is used to use time-frequency analysis to determine whether there is a fault when the vibration signal is a non-steady-state transient signal.

10. A computer-readable storage medium, characterized in that The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a computer to execute the aircraft engine fault monitoring method according to any one of claims 1 to 8.

Images