A vehicle-mounted hydrogen system leakage identification method, system, device, medium and product

Abstract

The application discloses a kind of vehicle-mounted hydrogen system leakage identification method, system, equipment, medium and product, it is related to hydrogen fuel cell vehicle safety monitoring field, the method includes obtaining the leakage response signal of vehicle-mounted hydrogen system pipeline including vibration signal and deformation strain signal and carries out preprocessing operation;The preprocessed leakage response signal is respectively subjected to frequency domain feature extraction and time domain feature extraction, and frequency domain feature data and time domain feature data are obtained;If frequency domain feature data is located in the preset leakage energy proportion interval, and time domain feature data meets preset pulse discriminant leakage condition, then determine that vehicle-mounted hydrogen system occurs leakage and carries out alarm.The application is monitored based on pipeline vibration and deformation signal, is judged by frequency domain and time domain double feature fusion, realizes the leakage identification of fast, accurate, non-invasive of vehicle-mounted hydrogen system, effectively overcome the problem, such as slow response, easily disturbed, difficult to locate and the risk brought by invasive measurement, improves the real-time and reliability of monitoring.

Description

Technical Field

[0001] This application relates to the field of hydrogen fuel cell vehicle safety monitoring technology, and in particular to a method, system, device, medium and product for identifying leaks in on-board hydrogen systems. Background Technology

[0002] Against the backdrop of a global shift towards a low-carbon energy structure, hydrogen energy, with its zero-carbon emissions, high energy density, and versatility, has become one of the most promising clean energy sources. Hydrogen fuel cell vehicles represent a crucial direction for hydrogen energy applications, and their large-scale deployment is accelerating. However, the unique physicochemical properties of hydrogen also present significant safety challenges. Its extremely small molecular weight allows for rapid diffusion in the air after leakage. Furthermore, hydrogen has a very wide flammability range and low ignition energy, meaning it can be ignited by even a weak electrostatic spark. Therefore, the large-scale promotion of hydrogen fuel cell vehicles always faces the potential risk of leakage and explosion. Rapid and accurate monitoring of hydrogen leakage in onboard hydrogen systems is a crucial means of ensuring the safe application of fuel cell vehicles.

[0003] The monitoring solutions in related technologies mainly focus on two aspects: First, monitoring the atmospheric hydrogen concentration above the vehicle's hydrogen system, such as using multiple hydrogen concentration sensors for joint monitoring and combining them with temperature sensors for comprehensive evaluation to determine the leakage situation in the target space. However, this method requires hydrogen to diffuse to the sensor location to achieve monitoring. During driving, it is greatly affected by ambient wind interference, has a slow response, and is difficult to locate leaks. Second, monitoring the physical parameters of hydrogen inside the pipeline to determine whether hydrogen leakage has occurred, such as monitoring abnormal changes in hydrogen pressure, flow rate, and temperature, and using neural network models to determine whether the vehicle's hydrogen system has leaked. However, this method is greatly affected by gas fluctuations, has low accuracy, and requires a large number of temperature, pressure, and flow sensors for invasive measurements, increasing the risk of gas blockage or leakage.

[0004] Therefore, there is an urgent need to propose a method for identifying leaks in vehicle-mounted hydrogen systems to address issues such as slow monitoring response, susceptibility to interference, difficulty in locating leaks, and additional risks associated with invasive measurements, thereby achieving rapid and accurate leak monitoring during vehicle operation. Summary of the Invention

[0005] The purpose of this application is to provide a method, system, device, medium and product for identifying leaks in vehicle-mounted hydrogen systems. It can utilize the physical signals of pipeline vibration and deformation caused by leaks and make judgments through dual feature fusion to achieve rapid, interference-resistant and non-invasive leak identification and location of leaks in vehicle-mounted hydrogen systems. This solves the problems of slow monitoring response, susceptibility to environmental interference, difficulty in location, and additional risks brought by invasive measurements in the prior art.

[0006] To achieve the above objectives, this application provides the following solution:

[0007] Firstly, this application provides a method for identifying leaks in an on-board hydrogen system, including:

[0008] The leakage response signal of the on-board hydrogen system pipeline is acquired; the leakage response signal includes vibration signal and deformation strain signal.

[0009] The leakage response signal is preprocessed to obtain a preprocessed leakage response signal.

[0010] Frequency domain feature extraction and time domain feature extraction are performed on the preprocessed leakage response signal to obtain frequency domain feature data and time domain feature data. The frequency domain feature data includes the energy proportion of the preprocessed leakage response signal in a preset frequency band. The time domain feature data includes the difference sequence of adjacent pulse peak amplitudes and the pulse period sequence extracted based on the preprocessed leakage response signal.

[0011] If the frequency domain feature data is within a preset leakage energy percentage range, and the time domain feature data meets a preset pulse discrimination leakage condition, then the on-board hydrogen system is determined to have leaked, and an alarm is triggered. The preset pulse discrimination leakage condition includes that the statistical characteristics of the difference sequence of amplitudes of adjacent pulse peaks are greater than or equal to a preset difference threshold, and the statistical characteristics of the pulse period sequence are less than a preset period threshold. The statistical characteristics include mean, median, or a set percentile value.

[0012] Secondly, this application provides an on-board hydrogen system leak detection system, including:

[0013] The signal acquisition module is used to acquire the leakage response signal of the on-board hydrogen system pipeline; the leakage response signal includes vibration signal and deformation strain signal.

[0014] The signal preprocessing module is used to preprocess the leakage response signal to obtain a preprocessed leakage response signal.

[0015] The feature extraction module is used to perform frequency domain feature extraction and time domain feature extraction on the preprocessed leakage response signal to obtain frequency domain feature data and time domain feature data. The frequency domain feature data includes the energy ratio of the preprocessed leakage response signal in a preset frequency band. The time domain feature data includes the difference sequence of adjacent pulse peak amplitudes and the pulse period sequence extracted based on the preprocessed leakage response signal.

[0016] The leakage identification module is used to determine that the on-board hydrogen system has leaked and to issue an alarm if the frequency domain feature data is within a preset leakage energy percentage range and the time domain feature data meets a preset pulse discrimination leakage condition. The preset pulse discrimination leakage condition includes that the statistical characteristics of the difference sequence of amplitudes of adjacent pulse peaks are greater than or equal to a preset difference threshold, and the statistical characteristics of the pulse period sequence are less than a preset period threshold. The statistical characteristics include mean, median, or a set percentile value.

[0017] Thirdly, this application provides a computer device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the vehicle hydrogen system leak identification method described in any one of the above.

[0018] Fourthly, this application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the vehicle hydrogen system leak identification method described above.

[0019] Fifthly, this application provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the vehicle hydrogen system leak identification method described above.

[0020] According to the specific embodiments provided in this application, this application has the following technical effects:

[0021] This application provides a method, system, device, medium, and product for identifying leaks in an on-board hydrogen system. It acquires leak response signals from the on-board hydrogen system pipeline; these signals include vibration and deformation strain signals. This step completely abandons the traditional passive mode that relies on gas diffusion monitoring or invasive measurements inside the pipeline, directly acquiring information from the source of the physical effects caused by the leak. This solves the monitoring delay problem caused by waiting for gas diffusion in existing technologies, as well as the additional leakage risks and increased system complexity caused by installing sensors inside high-pressure pipelines. It achieves rapid, direct, and non-invasive physical signal perception of leak events. By preprocessing the leak response signals, a preprocessed leak response signal is obtained. This preprocessing effectively suppresses strong background noise such as continuous road bumps and engine vibrations in the on-board environment, solving the technical problem that the original leak signal is easily submerged and has a very low signal-to-noise ratio. This creates conditions for subsequent accurate feature extraction, enabling effective separation of potential leak components from complex mixed signals. The preprocessed leak response signal is further processed by frequency... Frequency domain feature extraction and time domain feature extraction are used to obtain frequency domain feature data and time domain feature data. The inherent fingerprint characteristics of the leakage signal are deeply mined from two independent dimensions: frequency and time domain. Frequency domain feature data (such as the energy ratio within a preset frequency band) captures the high-frequency vibration characteristics caused by the leakage jet, and time domain feature data (adjacent peak difference sequence and pulse period sequence) captures the transient pulse characteristics of the leakage impact. This overcomes the defects of incomplete information and limited characterization ability of a single feature dimension, and realizes a comprehensive and three-dimensional feature characterization of the leakage signal. If the frequency domain feature data is located within a preset leakage energy ratio range and the time domain feature data meets the preset pulse discrimination leakage condition, a leakage is determined and an alarm is triggered. The strict fusion decision logic that the frequency domain feature and the time domain feature must be satisfied simultaneously effectively filters out common interference scenarios such as only high-frequency noise (such as motor noise) or only a single impact (such as stone impact). It solves the key problems of high false alarm rate and poor reliability under single criteria or simple logic combination. It realizes intelligent identification and real-time alarm of hydrogen leakage with high accuracy and low false alarm rate under complex real vehicle working conditions. In summary, this application establishes a complete, efficient, and reliable active monitoring technology system for on-board hydrogen system leaks, providing crucial technical support for the safe operation of hydrogen fuel cell vehicles. Attached Figure Description

[0022] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0023] Figure 1This is an application environment diagram of a method for identifying leaks in an on-board hydrogen system according to an embodiment of this application.

[0024] Figure 2 This is a flowchart illustrating a method for identifying leaks in an on-board hydrogen system, provided as an embodiment of this application.

[0025] Figure 3 This is a flowchart illustrating a method for identifying leaks in an on-board hydrogen system, provided as another embodiment of this application.

[0026] Figure 4 This is a schematic diagram of a leakage source localization process based on the time difference of arrival of multi-sensor signals, provided as an embodiment of this application.

[0027] Figure 5 This is a schematic diagram of a sensor arrangement scheme for a monitoring system provided in an embodiment of this application.

[0028] Figure 6 This is a schematic diagram of the functional modules of an on-board hydrogen system leak detection system provided in an embodiment of this application.

[0029] Figure 7 This is a schematic diagram of the structure of a computer device provided in an embodiment of this application. Detailed Implementation

[0030] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of this application.

[0031] To make the above-mentioned objectives, features and advantages of this application more apparent and understandable, the application will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0032] The vehicle-mounted hydrogen system leakage identification method provided in this application embodiment can be applied to, for example... Figure 1In the application environment shown, terminal 101 communicates with server 102 via a network. A data storage system can store the data that server 102 needs to process. The data storage system can be set up independently, integrated into server 102, or placed in the cloud or on another server. Terminal 101 can send the leakage response signal of the on-board hydrogen system pipeline to server 102. After receiving the leakage response signal, server 102 performs preprocessing on the leakage response signal to obtain a preprocessed leakage response signal. Frequency domain feature extraction and time domain feature extraction are performed on the preprocessed leakage response signal to obtain frequency domain feature data and time domain feature data. The frequency domain feature data includes the energy percentage of the preprocessed leakage response signal within a preset frequency band. The time domain feature data includes the difference sequence of adjacent pulse peak amplitudes and the pulse period sequence extracted based on the preprocessed leakage response signal. If the frequency domain feature data is within a preset leakage energy percentage range and the time domain feature data meets a preset pulse leakage discrimination condition, then the on-board hydrogen system is determined to have leaked, and an alarm is triggered. The preset pulse leakage discrimination condition includes that the statistical characteristics of the difference sequence of adjacent pulse peak amplitudes are greater than or equal to a preset difference threshold, and the statistical characteristics of the pulse period sequence are less than a preset period threshold. The statistical characteristics include the mean, median, or a set percentile value. Server 102 can feed back the leak detection result of the on-board hydrogen system to terminal 101. Furthermore, in some embodiments, the on-board hydrogen system leak detection method can also be implemented by either server 102 or terminal 101 independently. For example, terminal 101 can directly perform leak detection processing on the leak response signal of the on-board hydrogen system pipeline, or server 102 can obtain the leak response signal of the on-board hydrogen system pipeline from the data storage system and perform leak detection processing on the leak response signal.

[0033] The terminal 101 can be, but is not limited to, various desktop computers, laptops, smartphones, tablets, IoT devices, and portable wearable devices. The server 102 can be implemented using a standalone server or a server cluster consisting of multiple servers, or it can be a cloud server.

[0034] In one exemplary embodiment, such as Figure 2 As shown, a method for identifying leaks in an on-board hydrogen system is provided. This method is executed by a computer device, specifically a terminal or server, or both. In this embodiment, the method is applied to... Figure 1 Taking server 102 as an example, the explanation includes the following steps 201 to 204. Wherein:

[0035] Step 201: Obtain the leakage response signal of the on-board hydrogen system pipeline; the leakage response signal includes vibration signal and deformation strain signal.

[0036] Step 202: Perform preprocessing on the leakage response signal to obtain a preprocessed leakage response signal.

[0037] Step 203: Perform frequency domain feature extraction and time domain feature extraction on the preprocessed leakage response signal to obtain frequency domain feature data and time domain feature data; the frequency domain feature data includes the energy ratio of the preprocessed leakage response signal in a preset frequency band; the time domain feature data includes the difference sequence of adjacent pulse peak amplitudes and the pulse period sequence extracted based on the preprocessed leakage response signal.

[0038] Step 204: If the frequency domain feature data is within a preset leakage energy percentage range and the time domain feature data meets a preset pulse leakage detection condition, then the on-board hydrogen system is determined to have leaked, and an alarm is triggered. The preset pulse leakage detection condition includes that the statistical characteristics of the difference sequence of amplitudes of adjacent pulse peaks are greater than or equal to a preset difference threshold, and the statistical characteristics of the pulse period sequence are less than a preset period threshold. The statistical characteristics include mean, median, or a set percentile value.

[0039] By implementing steps 201 to 204 above, this application achieves rapid, in-situ detection of leak events by directly acquiring the physical response signals (vibration and deformation) caused by the leak, avoiding the diffusion delay and additional risks of invasive measurements inherent in traditional hydrogen concentration monitoring. Secondly, preprocessing the signal effectively suppresses onboard environmental noise and improves the signal-to-noise ratio. Furthermore, complementary features are extracted from two independent dimensions—frequency and time domains—to construct a comprehensive leak signal fingerprint. Finally, through a dual-feature fusion decision-making mechanism in the frequency and logical time domains, false alarms caused by a single interference source are significantly reduced. This enables highly accurate and reliable intelligent identification and real-time early warning of onboard hydrogen system leaks in complex driving environments, effectively improving the overall safety level of hydrogen fuel cell vehicles.

[0040] In another exemplary embodiment of this application, the preprocessed leakage response signal is preprocessed to obtain a preprocessed leakage response signal, specifically including:

[0041] Based on the vibration signal and deformation strain signal in the leakage response signal, the dominant frequency of the vibration signal and the dominant frequency of the deformation strain signal are extracted through real-time spectrum analysis.

[0042] The dominant frequencies of the vibration signal and the deformation strain signal are dynamically filtered using an adaptive Kalman filter to obtain a preprocessed leakage response signal. The preprocessed leakage response signal includes the filtered vibration signal and the filtered deformation strain signal. The adaptive Kalman filter method includes increasing the process noise covariance matrix of the Kalman filter method using the following formula when the dominant frequency is greater than or equal to a first frequency threshold and less than or equal to a second frequency threshold:

[0043] .

[0044] in, Represents the process noise covariance matrix; Represents the basic process noise covariance matrix; Indicates an indicator function; Indicates the adjustment factor; Indicates the dominant frequency; Indicates the first frequency threshold; This indicates the second frequency threshold.

[0045] If the dominant frequency is greater than the second frequency threshold and less than or equal to the third frequency threshold, the process noise covariance matrix is ​​adjusted to the basic process noise covariance matrix, and the measurement noise covariance matrix is ​​adjusted to be less than or equal to the basic measurement noise covariance matrix.

[0046] In another exemplary embodiment of this application, frequency domain feature extraction is performed on the preprocessed leakage response signal to obtain frequency domain feature data, specifically including:

[0047] The filtered vibration signal and the filtered deformation strain signal are respectively subjected to Hanning windows using the following formulas to obtain the corresponding windowed signals:

[0048] .

[0049] .

[0050] in, , The time indicates the filtered vibration signal. Time represents the filtered deformation strain signal; Represents the nth sampling point. Windowed signals of the same type; Represents the nth sampling point. Class of signals; The nth sampling point represents the Hanning window function; N represents the total number of sampling points.

[0051] Based on the windowed signals corresponding to the filtered vibration signal and the filtered deformation strain signal, frequency domain transformation is performed using the Fast Fourier Transform method to obtain the corresponding frequency domain complex number sequence.

[0052] Based on the frequency domain complex sequence, the frequency domain feature data is calculated using the following formula:

[0053] .

[0054] .

[0055] in, Indicates the first The percentage of energy of a signal type within a preset frequency band; This represents the k-th frequency component after the Fourier transform; This represents the actual frequency value corresponding to the k-th frequency component; Indicates the sampling frequency; Indicates the lower limit frequency of the preset frequency band; Indicates the upper limit frequency of the preset frequency band; Indicates the first Frequency domain complex sequences of signals; Indicates the first The power spectral density of a complex sequence of signals in the frequency domain.

[0056] In another exemplary embodiment of this application, time-domain feature extraction is performed on the preprocessed leakage response signal to obtain time-domain feature data, specifically including:

[0057] Based on the filtered vibration signal and filtered deformation strain signal in the preprocessed leakage response signal, the wave peaks are detected by the sliding window maxima method to obtain the corresponding wave peak amplitude sequences.

[0058] Based on the peak amplitude sequence, the difference sequence of adjacent pulse peak amplitudes is calculated using the following formula:

[0059] .

[0060] in, Indicates the first The difference in amplitude of the peak of the i-th adjacent pulse of a signal; Indicates the first The amplitude of the (i+1)th pulse peak of the signal type; Indicates the first The amplitude of the i-th pulse peak of the signal class; i={1,2,....., -1}; Indicates the first The total number of pulse peaks detected in the signal type.

[0061] The pulse period sequence is calculated using the following formula:

[0062] .

[0063] in, Indicates the first The i-th pulse period of the signal type; Indicates the first The moment of the (i+1)th pulse peak of the signal type; Indicates the first The moment of the i-th pulse peak of the signal.

[0064] In another exemplary embodiment of this application, after determining that the on-board hydrogen system has leaked, the method further includes:

[0065] The actual arrival time of the vibration signal in the leak response signal detected by each acceleration sensor in the sensor array arranged axially along the on-board hydrogen system pipeline is obtained.

[0066] Based on the actual arrival time, the pipe length between each accelerometer, and the material vibration wave velocity, calculate the arrival time difference of the vibration signal between any two accelerometers.

[0067] Based on the arrival time difference of vibration signals between any two accelerometers and the inherent propagation time difference of the pipeline length between the corresponding accelerometers, the orientation of the leak point relative to each accelerometer is determined, and the preliminary position coordinates of the leak point along the pipeline axis are calculated to obtain the preliminary positioning positions of multiple accelerometer pairs.

[0068] Based on the preliminary location of multiple sets of acceleration sensor pairs, the final location of the leak point is determined through a weighted fusion algorithm.

[0069] In another exemplary embodiment of this application, based on the preliminary positioning positions of multiple sets of acceleration sensor pairs, the final leak point location is determined through a weighted fusion algorithm, specifically including:

[0070] Based on the preliminary positioning positions of multiple sets of accelerometer pairs, the preliminary positioning positions that meet the preset validity conditions are taken as valid preliminary positioning positions, thus obtaining a set of valid preliminary positioning positions. The preset validity conditions include that the absolute value of the difference between the arrival time difference of the vibration signal of the accelerometer pair and the inherent propagation time difference of the pipeline length between the corresponding accelerometers is less than or equal to a preset time difference threshold, and the signal-to-noise ratio of the vibration signal of the accelerometer pair is greater than or equal to a preset signal-to-noise ratio threshold.

[0071] The weight of the accelerometer pair corresponding to each valid preliminary positioning position in the set of valid preliminary positioning positions is determined by the following formula:

[0072] .

[0073] in, This represents the weight of the q-th accelerometer pair in the set of valid preliminary positioning locations; This represents the signal-to-noise ratio of the vibration signal of the q-th accelerometer pair in the effective preliminary positioning location set. ; Indicates the first position in the set of valid preliminary locations. The signal-to-noise ratio of the vibration signals from a pair of accelerometers. p represents the total number of accelerometer pairs in the effective preliminary positioning location set.

[0074] The final location of the leak can be determined using the following formula:

[0075] .

[0076] in, This indicates the final location of the leak; This represents the effective preliminary positioning position of the q-th accelerometer pair in the effective preliminary positioning position set.

[0077] The following example illustrates this application using a specific process for identifying leaks in an onboard hydrogen system.

[0078] This application proposes a method for monitoring leaks in an on-board hydrogen system based on pipeline vibration and deformation signals, to achieve rapid and accurate leak monitoring during vehicle operation, specifically including:

[0079] 1. Monitoring system architecture.

[0080] The sensor arrangement of the monitoring system in this application is as follows: Figure 5 As shown, the core design is based on the physical response characteristics (vibration and deformation) of an on-board hydrogen system after leakage. On one hand, it monitors vibration and deformation signals caused by leakage: On-board hydrogen system leaks often occur at critical locations such as pipe joints and valve interfaces. During a leak, the jet backlash of high-pressure hydrogen gas causes pipe vibration, with the maximum vibration displacement concentrated around the leak point and in the middle of long, straight pipes. Therefore, piezoelectric accelerometers are placed in these highly sensitive vibration displacement areas to collect vibration signals caused by leakage in real time, indirectly obtaining the displacement change characteristics of these areas. On the other hand, it monitors deformation signals caused by leakage: Pipe joints, pipe-to-hydrogen tank interfaces, and vehicle body connections experience the most concentrated deformation during leakage due to their structural connection characteristics. Therefore, anti-hydrogen embrittlement strain gauges are placed in these areas of concentrated deformation to accurately monitor pipe strain changes caused by leakage, directly reflecting the local deformation state of the pipe.

[0081] 2. Leakage signal identification.

[0082] In real-world vehicle operation, onboard hydrogen systems suffer from interference from road bumps and vibrations from rotating vehicle components, which can mask leak signals, necessitating the design of targeted identification strategies. Interference vibrations in the real-world vehicle environment are characterized by "low frequency and high energy": the vibration frequency of vehicle transmission components is concentrated between 10Hz and 200Hz, and the vibration frequency of road bumps is concentrated between 5Hz and 50Hz; both are low-frequency interferences with relatively high vibration energy, achieving acceleration amplitudes of 5g-10g. In contrast, the jet vibrations caused by hydrogen leaks exhibit "high frequency and low energy," with a frequency range concentrated between 2kHz and 5kHz and an acceleration amplitude of only 0.1g-0.5g. Based on the different frequencies and amplitudes of leak and noise signals, a three-level process of "preprocessing - feature extraction - multi-feature fusion judgment" is employed to eliminate environmental vibration interference and achieve accurate identification of leak signals. A flowchart of an onboard hydrogen system leak identification method is shown below. Figure 3 As shown.

[0083] 2.1 Signal preprocessing.

[0084] The noise signal is mainly low-frequency vibration, which is dynamically filtered out using an adaptive Kalman filter. The fixed parameters of the traditional Kalman filter (mainly including the process noise covariance matrix Q and the measurement noise covariance matrix R) cannot adapt to the dynamic interference in the vehicle environment. Therefore, an adaptive Kalman filter based on frequency characteristics is designed to dynamically adjust the filter parameters by identifying the dominant frequency of environmental vibration in real time, thereby achieving accurate filtering of low-frequency interference.

[0085] For sensor signals containing environmental noise, the dominant frequency of the current signal is extracted through real-time spectrum analysis. When the dominant frequency f dom When the frequency range is ∈ [10Hz, 200Hz], it is determined that environmental vibration is dominant. The process noise covariance Q is increased to enhance the filtering strength and suppress low-frequency components. The formula is as follows:

[0086] .

[0087] in, Represents the process noise covariance matrix; Represents the basic process noise covariance matrix; Indicates an indicator function; Indicates the adjustment factor; Indicates the dominant frequency; This represents the first frequency threshold, with a value of 10Hz. This represents the second frequency threshold, with a value of 200Hz.

[0088] When f dom When ∈ (200Hz, 10kHz), it is determined that there is a leakage signal, Q is reduced to Q0, and the measurement noise covariance matrix R is reduced at the same time.

[0089] After adaptive Kalman filtering, low-frequency signals are significantly reduced while high-frequency signals are retained, ensuring the effectiveness of feature extraction.

[0090] 2.2 Leakage feature identification.

[0091] The core of leakage feature extraction is to screen out features from the pre-processed vibration / strain signals that are generated solely by hydrogen leakage and are significantly different from environmental vibrations. These features are divided into two categories: frequency domain (high-frequency main peak) and time domain (pulsatility). The two work together to achieve accurate identification of leakage signals.

[0092] (1) Leakage signal identification based on frequency domain characteristics.

[0093] First, the time-domain signal is converted to the frequency domain based on the Fast Fourier Transform (FFT), and the energy proportion of the high-frequency band is extracted as the core frequency domain criterion for leakage.

[0094] Let the preprocessed vibration signal be (n=1,2,...,N, where N is the total number of sampling points and the data length), the sampling frequency is =10kHz (satisfies Nyquist's theorem (also known as the Nyquist sampling theorem or Shannon sampling theorem), covering the target frequency band of 2kHz-5kHz). To reduce spectral leakage, first... Apply Hanning window To avoid low-frequency energy interference from environmental vibrations in high-frequency band analysis:

[0095] .

[0096] .

[0097] For the windowed signal Perform an N-point FFT to obtain a complex sequence in the frequency domain. (k=1,2,...,N), the corresponding actual frequencies are:

[0098] .

[0099] Taking N=1024 (balancing computational efficiency and frequency resolution), then the frequency resolution... It can accurately distinguish frequency details within the range of 2kHz-5kHz.

[0100] The quantitative index of frequency domain characteristics is "the proportion of energy in the leakage characteristic frequency band to the total energy", and the formula is as follows: .

[0101] in, Indicates the first The energy proportion of the signal type within the preset frequency band (2kHz-5kHz) (high-frequency energy proportion) has a value range of [0,1]. This represents the k-th frequency component after the Fourier transform; This represents the actual frequency value corresponding to the k-th frequency component; Indicates the sampling frequency; Indicates the lower limit frequency of the preset frequency band; Indicates the upper limit frequency of the preset frequency band; Indicates the first Frequency domain complex sequences of signals; Indicates the first The power spectral density of a complex sequence of signals in the frequency domain. The numerator on the right side of the formula represents the energy of the leakage characteristic frequency band (2kHz-5kHz), and the denominator represents the total energy of the 0kHz-5kHz frequency band (due to...). The Nyquist frequency is 5kHz. Signals above 5kHz have no physical meaning, so only 0kHz-5kHz are calculated.

[0102] The energy proportions of the two types of signals within a preset frequency band are fused to obtain the fused energy proportion.

[0103] Whether a leak has occurred is determined based on whether the fusion energy percentage is within a preset leakage energy percentage range.

[0104] Real-vehicle comparative experiments were conducted (including multiple sets of normal operating conditions and multiple sets of leakage operating conditions). The proportion of high-frequency energy in the 2kHz-5kHz range was statistically obtained when hydrogen was not leaking and when it was leaking. The measured fusion energy proportion was then used to... Comparing the results with experimental results can serve as a basis for determining whether a leak has occurred.

[0105] Wherein, the fused energy ratio is the arithmetic mean of the energy ratios of the vibration signal and the deformation strain signal within a preset frequency band, and its calculation formula is as follows:

[0106] .

[0107] in, Indicates the percentage of fused energy; and These represent the energy percentages of vibration signals and deformation strain signals within a preset frequency band, respectively.

[0108] like If a leak occurs, it is determined that a leak has occurred; otherwise, it is determined that no leak has occurred. and These are the lower and upper limits of the preset leakage energy percentage range obtained through leakage condition statistics, respectively.

[0109] (2) Leakage signal identification based on time domain characteristics.

[0110] Environmental vibrations (such as idling vibrations of transmission components) are continuous and stable signals (small fluctuations in peak amplitude and stable period), while the jet recoil force of hydrogen leakage has an "intermittent impact" characteristic—turbulent disturbances during gas leakage cause pulse-like vibrations / deformations in the pipeline, manifested as "large differences between adjacent peaks and short pulse periods." These two pulse characteristics are extracted as supplementary criteria through time-domain peak analysis.

[0111] Preprocessed time-domain signal (t is time), the peak is detected using the "sliding window maxima method". The window length is set to T. w =20ms (covering half a cycle of environmental vibration to avoid missed detections), if at a certain moment signal value If the value is greater than all other values ​​within the window, and also greater than the "noise threshold" (set to 0.05g / 25με, based on the maximum noise amplitude under normal operating conditions), then it is determined that... The peak moment corresponds to a peak amplitude of [value missing]. .

[0112] Let the continuously detected peak amplitude be , ( Indicates the first (Total number of pulse peaks detected in the signal class), the difference between the i-th adjacent peaks is defined as:

[0113] .

[0114] in, Indicates the first The difference in amplitude of the peak of the i-th adjacent pulse of a signal; Indicates the first The amplitude of the (i+1)th pulse peak of the signal type; Indicates the first The amplitude of the i-th pulse peak of the signal class; i={1,2,....., -1}; Indicates the first The total number of pulse peaks detected in the signal type.

[0115] The pulsatility of the leak is also reflected in the "short time intervals between wave crests and large fluctuations" (unlike the stable wave crest intervals of environmental vibrations). The "pulse period" is defined as the time difference between two consecutive wave crests.

[0116] .

[0117] in, Indicates the first The i-th pulse period of the signal type; Indicates the first The moment of the (i+1)th pulse peak of the signal type; Indicates the first The moment of the i-th pulse peak of the signal.

[0118] Calculate the mean of the adjacent peak difference sequence and the mean of the pulse period sequence for the vibration signal and the deformation strain signal, respectively:

[0119] .

[0120] .

[0121] in, Indicates the first Mean value of the sequence of adjacent peak differences of a signal; Indicates the first The mean of the pulse period sequence of the signal type.

[0122] The arithmetic mean of the two types of signals is fused to obtain the difference between adjacent peaks. and fusion pulse period :

[0123] .

[0124] .

[0125] Based on real-vehicle comparative test data, the distribution range of fused features under normal and leakage conditions was statistically analyzed, and a time-domain leakage judgment threshold was set; if and If the data falls within the characteristic range corresponding to the leakage condition, it is considered temporal evidence of a suspected leakage.

[0126] Both frequency domain characteristics and time domain characteristics must be satisfied simultaneously for a signal to be definitively identified as a leakage signal. The specific logic is as follows:

[0127] ① If only the frequency domain meets the requirements (such as high-frequency noise from a vehicle motor): it is determined to be an "interference signal" and no alarm is triggered.

[0128] ② If only the time domain meets the requirement (such as a single pulse from a stone hitting a pipeline): it is determined as an "occasional impact" and no alarm is triggered.

[0129] ③ If both conditions are met: confirm the leak and trigger an audible and visual alarm.

[0130] The above content presents a leakage monitoring method based on frequency domain and time domain characteristics. Furthermore, the installed accelerometer and resistance strain gauge can be used in conjunction with pressure sensors and hydrogen concentration sensors to improve accuracy. In addition, structural damage and loose connections in the vehicle hydrogen system can be monitored by comparing the vibration signals during use with those at the factory.

[0131] 3. Methods for locating the leakage source.

[0132] like Figure 4 The diagram illustrates a leak source localization process based on the time difference of arrival (TDOA) of multi-sensor signals. The core of leak source localization is to calculate coordinates using the TDOA of sensor array signals. Each sensor is numbered S1, S2, ... Record the pipe length between any two sensors as d. xy (0≤x≤ , 0≤y≤ Establish a three-dimensional relationship of "number-location-spacing" as a quantitative benchmark for positioning. This indicates the total number of sensors.

[0133] Drawing inspiration from the binaural effect, the vibration waves generated by the leak propagate to both ends of the pipeline. Sensors with different numbers receive signals in order of distance from the leak point. The signal arrival time difference between sensors located on the same side of the leak point is fixed and denoted as Δt. xy，const , Δt xy，const The arrival time difference of the sensor signals located on both sides of the leak point depends on the location of the leak point. The leak point can be located using the signals from two sensors located on either side of the leak point.

[0134] 3.1 Dual-sensor positioning method.

[0135] Two sensors, S1 and S2, receive vibration signals with frequencies and amplitudes consistent with leakage characteristics. The cross-correlation function characterizes the similarity between the two sensor signals over a given time offset, with the peak position corresponding to the time difference between the signals. The cross-correlation function between S1 and S2 is:

[0136] .

[0137] in, The cross-correlation function value of the signals from sensors S1 and S2. T For signal duration, set to , This is the time offset.

[0138] when When taking the maximum value, the corresponding That is (If signal S2 is later than signal S1,) ). Δt 12 With Δt 12，const In comparison, if Δt 12 =Δt 12，const This indicates that the two sensors are located on the same side of the leak point, and the sensor group cannot be used for localization. If Δt 12 <Δt 12，const This indicates that the leak point is located between the two sensors and can be used for localization. Let the distance from the leak point to S1 be d and the distance to S2 be d', then:

[0139] .

[0140] in, L 12 denoted as , where is the straight-line distance in the pipeline between sensor S1 and sensor S2 (i.e., the sensor spacing); v is the speed of sound in the pipeline.

[0141] By solving the system of equations simultaneously, the distance between the leak point and the two sensors can be obtained. Since the sensor positions are known, the specific location of the leak point along the pipeline axis can be determined.

[0142] 3.2 Cross-localization of multi-sensor networks.

[0143] The above describes a dual-sensor localization method. When the vibration signal propagates along the hydrogen supply pipeline, this method can be used to locate the leak by any two sensors located on either side of the leak point. Therefore, a multi-sensor network can be used for cross-validation to solve problems such as measurement errors and noise interference that exist in sensors during actual operation, thereby improving the reliability of leak localization.

[0144] Based on the arrival time difference of any two sensor signals and their inherent positions on the pipeline, determine which side (upstream or downstream) each sensor is located on at the leak point. First, for sensor signals on the same side, check if the following conditions are met; if not, discard them:

[0145] ① Time difference rationality: Calculate the time difference Δt between every two sensors. Its absolute value should be the same as or close to the predetermined time difference (less than or equal to the preset time difference threshold). If the time difference between a certain sensor and other sensors is significantly different, the result of that sensor shall be discarded.

[0146] ② Signal quality rationality: The signal-to-noise ratio of the vibration signal of the sensor pair must be greater than the preset signal-to-noise ratio threshold (20dB). If it is lower, it means that the signal is seriously interfered with and should be rejected.

[0147] For the multiple sets of valid positioning results after screening, a weighted fusion algorithm is used to calculate the final positioning value. The core idea is to give higher weights to sensors with high signal quality to reduce random errors.

[0148] Assign weights to each group of valid sensor pairs. ( ), calculated based on the signal-to-noise ratio (SNR) of the sensor pair, with higher SNR carrying greater weight:

[0149] .

[0150] in, This represents the weight of the q-th accelerometer pair in the set of valid preliminary positioning locations; This represents the signal-to-noise ratio of the vibration signal of the q-th accelerometer pair in the effective preliminary positioning location set. ; Indicates the first position in the set of valid preliminary locations. The signal-to-noise ratio of the vibration signals from a pair of accelerometers. p represents the total number of accelerometer pairs in the effective preliminary positioning location set.

[0151] Let the positioning result of each effective sensor pair be: Final positioning value for:

[0152] .

[0153] This allows for the reliable location of the leak point.

[0154] Based on the same inventive concept, this application also provides a vehicle-mounted hydrogen system leak detection system for implementing the above-described vehicle-mounted hydrogen system leak detection method. The solution provided by this system is similar to the implementation described in the above method; therefore, the specific limitations of one or more vehicle-mounted hydrogen system leak detection system embodiments provided below can be found in the limitations of the vehicle-mounted hydrogen system leak detection method described above, and will not be repeated here.

[0155] In one exemplary embodiment, such as Figure 6 As shown, an on-board hydrogen system leak detection system is provided, comprising:

[0156] The signal acquisition module 301 is used to acquire the leakage response signal of the on-board hydrogen system pipeline; the leakage response signal includes vibration signal and deformation strain signal.

[0157] The signal preprocessing module 302 is used to perform preprocessing operations on the leakage response signal to obtain a preprocessed leakage response signal.

[0158] The feature extraction module 303 is used to perform frequency domain feature extraction and time domain feature extraction on the preprocessed leakage response signal to obtain frequency domain feature data and time domain feature data; the frequency domain feature data includes the energy ratio of the preprocessed leakage response signal in a preset frequency band; the time domain feature data includes the difference sequence of adjacent pulse peak amplitudes and the pulse period sequence extracted based on the preprocessed leakage response signal.

[0159] The leakage identification module 304 is used to determine that the on-board hydrogen system has leaked and to issue an alarm if the frequency domain feature data is within a preset leakage energy ratio range and the time domain feature data meets a preset pulse discrimination leakage condition; the preset pulse discrimination leakage condition includes that the statistical characteristics of the difference sequence of amplitudes of adjacent pulse peaks are greater than or equal to a preset difference threshold, and the statistical characteristics of the pulse period sequence are less than a preset period threshold; the statistical characteristics include mean, median or set percentile value.

[0160] Compared to existing methods that use pressure and hydrogen sensors to monitor hydrogen leaks, this application has the following advantages:

[0161] 1. Vibration signals propagate quickly in pipelines, significantly reducing response time and improving real-time monitoring, thus meeting the need for rapid alarm of leaks in vehicle hydrogen systems.

[0162] 2. The sensor placement is more flexible, requiring only the selection of easily installable points in areas with significant deformation and displacement, without restrictions on the hydrogen supply pipeline layout.

[0163] 3. This application is a non-contact monitoring method, which is easy to set up, does not require changes to the existing pipeline structure, and does not interfere with the airflow inside the pipeline.

[0164] 4. It can achieve rapid and accurate leak location, which facilitates subsequent maintenance.

[0165] In one exemplary embodiment, a computer device is provided, which may be a server or a terminal, and its internal structure diagram may be as follows. Figure 7As shown, the computer device includes a processor, memory, input / output (I / O) interfaces, and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is also connected to the system bus via the I / O interfaces. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides the environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The database stores data related to the identification and processing of leaks in the on-board hydrogen system. The I / O interfaces are used for exchanging information between the processor and external devices. The communication interface is used for communication with external terminals via a network connection. When the computer program is executed by the processor, it implements a method for identifying leaks in an on-board hydrogen system.

[0166] Those skilled in the art will understand that Figure 7 The structures shown are merely block diagrams of some structures related to the present application and do not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than shown in the figures, or combine certain components, or have different component arrangements. In an exemplary embodiment, a computer device is provided, including a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the steps in the above-described method embodiments.

[0167] In one exemplary embodiment, a computer-readable storage medium is provided storing a computer program that, when executed by a processor, implements the steps in the above-described method embodiments.

[0168] In one exemplary embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps in the above-described method embodiments.

[0169] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of the relevant data must comply with relevant regulations.

[0170] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments described above. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM).

[0171] The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0172] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0173] This document uses specific examples to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the methods and core ideas of this application. Furthermore, those skilled in the art will recognize that, based on the ideas of this application, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A method for identifying leaks in an on-board hydrogen system, characterized in that, The method for identifying leaks in the on-board hydrogen system includes: The leakage response signal of the on-board hydrogen system pipeline is acquired; the leakage response signal includes vibration signal and deformation strain signal; The leakage response signal is preprocessed to obtain a preprocessed leakage response signal; the preprocessed leakage response signal includes a filtered vibration signal and a filtered deformation strain signal. Frequency domain feature extraction and time domain feature extraction are performed on the preprocessed leakage response signal to obtain frequency domain feature data and time domain feature data. The frequency domain feature data includes the energy proportion of the preprocessed leakage response signal within a preset frequency band. The time domain feature data includes the difference sequence of adjacent pulse peak amplitudes and the pulse period sequence extracted based on the preprocessed leakage response signal. Specifically, the frequency domain feature extraction of the preprocessed leakage response signal to obtain the frequency domain feature data includes: The filtered vibration signal and the filtered deformation strain signal are respectively subjected to Hanning windows using the following formulas to obtain the corresponding windowed signals: ； ； in, , The time indicates the filtered vibration signal. Time represents the filtered deformation strain signal; Represents the nth sampling point. Windowed signals of the same type; Represents the nth sampling point. Class of signals; This represents the Hanning window function for the nth sampling point; N represents the total number of sampling points. Based on the windowed signals corresponding to the filtered vibration signal and the filtered deformation strain signal, frequency domain transformation is performed using the fast Fourier transform method to obtain the corresponding frequency domain complex number sequence. Based on the frequency domain complex sequence, the frequency domain feature data is calculated using the following formula: ； ； in, Indicates the first The percentage of energy of a signal type within a preset frequency band; This represents the k-th frequency component after the Fourier transform; This represents the actual frequency value corresponding to the k-th frequency component; Indicates the sampling frequency; Indicates the lower limit frequency of the preset frequency band; Indicates the upper limit frequency of the preset frequency band; Indicates the first Frequency domain complex sequences of signals; Indicates the first Power spectral density of a complex sequence of signals in the frequency domain; The preprocessed leakage response signal is subjected to time-domain feature extraction to obtain time-domain feature data, specifically including: Based on the filtered vibration signal and filtered deformation strain signal in the preprocessed leakage response signal, the peaks are detected by the sliding window maximum method to obtain the corresponding peak amplitude sequence. Based on the peak amplitude sequence, the difference sequence of adjacent pulse peak amplitudes is calculated using the following formula: ； in, Indicates the first The difference in amplitude of the peak of the i-th adjacent pulse of a signal; Indicates the first The amplitude of the (i+1)th pulse peak of the signal type; Indicates the first The amplitude of the i-th pulse peak of the signal class; i={1,2,....., -1}; Indicates the first The total number of pulse peaks detected for this type of signal; The pulse period sequence is calculated using the following formula: ； in, Indicates the first The i-th pulse period of the signal type; Indicates the first The moment of the (i+1)th pulse peak of the signal type; Indicates the first The moment of the peak of the i-th pulse of the signal; If the frequency domain feature data is within a preset leakage energy percentage range, and the time domain feature data meets a preset pulse discrimination leakage condition, then the on-board hydrogen system is determined to have leaked, and an alarm is triggered. The preset pulse discrimination leakage condition includes that the statistical characteristics of the difference sequence of amplitudes of adjacent pulse peaks are greater than or equal to a preset difference threshold, and the statistical characteristics of the pulse period sequence are less than a preset period threshold. The statistical characteristics include mean, median, or a set percentile value.

2. The method for identifying leaks in an on-board hydrogen system according to claim 1, characterized in that, The leakage response signal is preprocessed to obtain a preprocessed leakage response signal, specifically including: Based on the vibration signal and deformation strain signal in the leakage response signal, the dominant frequency of the vibration signal and the dominant frequency of the deformation strain signal are extracted by real-time spectrum analysis. The dominant frequencies of the vibration signal and the deformation strain signal are dynamically filtered using an adaptive Kalman filter to obtain a preprocessed leakage response signal. The adaptive Kalman filter method includes increasing the process noise covariance matrix of the Kalman filter method using the following formula when the dominant frequency is greater than or equal to a first frequency threshold and less than or equal to a second frequency threshold: ； in, Represents the process noise covariance matrix; Represents the basic process noise covariance matrix; Indicates an indicator function; Indicates the adjustment factor; Indicates the dominant frequency; Indicates the first frequency threshold; Indicates the second frequency threshold; If the dominant frequency is greater than the second frequency threshold and less than or equal to the third frequency threshold, the process noise covariance matrix is ​​adjusted to the basic process noise covariance matrix, and the measurement noise covariance matrix is ​​adjusted to be less than or equal to the basic measurement noise covariance matrix.

3. The method for identifying leaks in an on-board hydrogen system according to claim 1, characterized in that, After determining that the on-board hydrogen system has leaked, the following steps are also included: The actual arrival time of the vibration signal in the leak response signal detected by each acceleration sensor in the sensor array arranged axially along the on-board hydrogen system pipeline is obtained; Based on the actual arrival time, the pipe length between each accelerometer, and the material vibration wave velocity, calculate the vibration signal arrival time difference between any two accelerometers. Based on the arrival time difference of vibration signals between any two accelerometers and the inherent propagation time difference of the pipeline length between the corresponding accelerometers, the orientation of the leak point relative to each accelerometer is determined, and the preliminary position coordinates of the leak point along the pipeline axis are calculated to obtain the preliminary positioning positions of multiple accelerometer pairs. Based on the preliminary location of multiple sets of acceleration sensor pairs, the final location of the leak point is determined through a weighted fusion algorithm.

4. The method for identifying leaks in an on-board hydrogen system according to claim 3, characterized in that, Based on the preliminary location data from multiple accelerometer pairs, a weighted fusion algorithm is used to determine the final location of the leak point, specifically including: Based on the preliminary positioning positions of multiple accelerometer pairs, the preliminary positioning positions that meet the preset validity conditions are taken as valid preliminary positioning positions, thus obtaining a set of valid preliminary positioning positions. The preset validity conditions include that the absolute value of the difference between the arrival time difference of the vibration signal of the accelerometer pair and the inherent propagation time difference of the pipeline length between the corresponding accelerometers is less than or equal to a preset time difference threshold, and the signal-to-noise ratio of the vibration signal of the accelerometer pair is greater than or equal to a preset signal-to-noise ratio threshold. The weight of the accelerometer pair corresponding to each valid preliminary positioning position in the set of valid preliminary positioning positions is determined by the following formula: ； in, This represents the weight of the q-th accelerometer pair in the set of valid preliminary positioning locations; This represents the signal-to-noise ratio of the vibration signal of the q-th accelerometer pair in the effective preliminary positioning location set. ; Indicates the first position in the set of valid preliminary locations. The signal-to-noise ratio of the vibration signals from a pair of accelerometers. ;p represents the total number of accelerometer pairs in the set of valid preliminary positioning locations; The final location of the leak can be determined using the following formula: ； in, This indicates the final location of the leak; This represents the effective preliminary positioning position of the q-th accelerometer pair in the effective preliminary positioning position set.

5. A vehicle-mounted hydrogen system leak detection system, characterized in that, The vehicle-mounted hydrogen system leak detection system applies the vehicle-mounted hydrogen system leak detection method according to any one of claims 1-4, and the vehicle-mounted hydrogen system leak detection system includes: The signal acquisition module is used to acquire the leakage response signal of the on-board hydrogen system pipeline; the leakage response signal includes vibration signal and deformation strain signal; The signal preprocessing module is used to preprocess the leakage response signal to obtain a preprocessed leakage response signal. The feature extraction module is used to perform frequency domain feature extraction and time domain feature extraction on the preprocessed leakage response signal to obtain frequency domain feature data and time domain feature data; the frequency domain feature data includes the energy ratio of the preprocessed leakage response signal in a preset frequency band; the time domain feature data includes the difference sequence of adjacent pulse peak amplitudes and the pulse period sequence extracted based on the preprocessed leakage response signal. The leakage identification module is used to determine that the on-board hydrogen system has leaked and to issue an alarm if the frequency domain feature data is within a preset leakage energy percentage range and the time domain feature data meets a preset pulse discrimination leakage condition. The preset pulse discrimination leakage condition includes that the statistical characteristics of the difference sequence of amplitudes of adjacent pulse peaks are greater than or equal to a preset difference threshold, and the statistical characteristics of the pulse period sequence are less than a preset period threshold. The statistical characteristics include mean, median, or a set percentile value.

6. A computer device, comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that the processor executes the computer program to implement the vehicle-mounted hydrogen system leak identification method according to any one of claims 1-4.

7. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the computer program implements the vehicle hydrogen system leak identification method as described in any one of claims 1-4.

8. A computer program product, comprising a computer program, characterized in that, When executed by a processor, the computer program implements the vehicle hydrogen system leak identification method as described in any one of claims 1-4.

Images