Knock signal detection method, device, equipment and storage medium

Abstract

The application provides a knock signal detection method, device and equipment and a storage medium. The method comprises the following steps: determining time-frequency domain types corresponding to a knock signal under different working conditions; screening wavelet bases corresponding to the time-frequency domain types by using a preset constraint condition to obtain a wavelet base candidate library; determining a target wavelet base based on a plurality of evaluation indexes obtained by decomposing and reconstructing the knock signal based on a first candidate wavelet base, the first candidate wavelet base being any wavelet base in the wavelet base candidate library; and detecting the knock signal based on the target wavelet base to obtain a detection result. In the application, the wavelet bases corresponding to the time-frequency domain types are screened by using the preset constraint condition to establish the wavelet base candidate library, the target wavelet base matching the working condition is obtained from the wavelet base candidate library by analyzing the plurality of evaluation indexes obtained by decomposing and reconstructing the knock signal, and the target wavelet base is used to realize efficient and robust knock signal detection.

Description

Technical Field

[0001] This application belongs to the field of artificial intelligence technology, specifically relating to a method, device, and storage medium for detecting detonation signals. Background Technology

[0002] Hydrogen engines are characterized by rapid combustion and a strong tendency to detonate. Accurate and rapid identification of detonation signals can provide reliable control information for the ECU, achieving optimal combustion performance while protecting the engine. However, the unique characteristics of hydrogen engine detonation signals, which exhibit a coupling of high-frequency oscillations and transient impacts, pose a significant challenge to traditional detection methods.

[0003] Currently, the industry typically uses fixed wavelet bases (such as a specific wavelet basis function) for detonation signal processing. Most existing research focuses on optimal basis enumeration under a single evaluation index, lacking a systematic analysis of the rationality and completeness of the candidate basis function sources. When faced with the diversity (different working conditions) and time-varying nature of detonation signals, inaccurate detonation signal detection may occur. Summary of the Invention

[0004] This application proposes a method, apparatus, device, and storage medium for detonation signal detection, which can solve the technical problem in related technologies that rely on optimal basis enumeration to determine wavelet basis under a single evaluation index, lacking a systematic analysis of the rationality and completeness of the sources of candidate basis functions. When faced with the diversity (different working conditions) and time-varying nature of detonation signals, inaccurate detonation signal detection may occur.

[0005] A first aspect of this application provides a method for detecting detonation signals, the method comprising: Determine the time-frequency domain type of the knock signal under different operating conditions; By using preset constraints, wavelet bases corresponding to multiple time-frequency domain types are screened to obtain a candidate library of wavelet bases; Based on multiple evaluation indicators obtained by decomposing and reconstructing the detonation signal using the first candidate wavelet basis, a target wavelet basis is determined. The first candidate wavelet basis is any one of the multiple candidate wavelet bases in the wavelet basis candidate library. The detonation signal is detected based on the target wavelet basis, and the detection result is obtained.

[0006] A second aspect of this application provides a detonation signal detection device, comprising: The determination module is used to determine the time-frequency domain type of the knock signal under different operating conditions; The filtering module is used to filter the wavelet bases corresponding to the multiple time-frequency domain types using preset constraints to obtain a candidate library of wavelet bases. The determining module is further configured to determine the target wavelet basis based on multiple evaluation indicators obtained by decomposing and reconstructing the detonation signal using the first candidate wavelet basis, wherein the first candidate wavelet basis is any one of the multiple candidate wavelet bases in the wavelet base candidate library. The detection module is used to detect the detonation signal based on the target wavelet basis and obtain the detection result.

[0007] An embodiment of the third aspect of this application provides an electronic 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 method described in the first aspect above.

[0008] An embodiment of the fourth aspect of this application provides a computer-readable storage medium having a computer program stored thereon, the program being executed by a processor to implement the method described in the first aspect above.

[0009] The technical solutions provided in this application embodiment have at least the following technical effects or advantages: This application proposes a method, apparatus, device, and storage medium for detonation signal detection. The method includes: determining the time-frequency domain type corresponding to the detonation signal under different operating conditions; filtering wavelet bases corresponding to multiple time-frequency domain types using preset constraints to obtain a wavelet base candidate library; determining a target wavelet base based on multiple evaluation indicators obtained by decomposing and reconstructing the detonation signal using a first candidate wavelet base, wherein the first candidate wavelet base is any wavelet base among multiple candidate wavelet bases in the wavelet base candidate library; and detecting the detonation signal based on the target wavelet base to obtain a detection result. In this application, the embodiments utilize preset constraints to filter wavelet bases corresponding to multiple time-frequency domain types to establish a wavelet base candidate library. By analyzing multiple evaluation indicators obtained by decomposing and reconstructing the detonation signal, a target wavelet base matching the operating conditions is obtained from the wavelet base candidate library, thereby achieving efficient and robust detonation signal detection.

[0010] Additional aspects and advantages of this application will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of this application. Attached Figure Description

[0011] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the scope of this application. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings: Figure 1 A flowchart illustrating a knock signal detection method according to an embodiment of this application is shown. Figure 2 A schematic diagram of a process for determining a target wavelet basis is shown in an embodiment of this application; Figure 3 This illustration shows a schematic diagram of the structure of a knock signal detection device according to an embodiment of this application; Figure 4 This illustration shows a schematic diagram of the structure of an electronic device according to an embodiment of this application; Figure 5 A schematic diagram of a storage medium provided in one embodiment of this application is shown. Detailed Implementation

[0012] Exemplary embodiments of this application will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of this application are shown in the drawings, it should be understood that this application may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided to enable a more thorough understanding of this application and to fully convey the scope of this application to those skilled in the art.

[0013] It should be noted that, unless otherwise stated, the technical or scientific terms used in this application shall have the ordinary meaning as understood by one of ordinary skill in the art to which this application pertains.

[0014] Building upon the aforementioned background technology, related technologies typically use a fixed wavelet basis or select a wavelet basis through a single evaluation index to detect the knock signal of a hydrogen engine. However, the above-mentioned wavelet basis selection process lacks systematic theoretical guidance, and a single evaluation index cannot meet the multi-dimensional requirements of hydrogen knock, resulting in the problem of erroneous knock signal detection based on this wavelet basis.

[0015] To address the aforementioned technical problems, this application proposes a method, apparatus, device, and storage medium for detonation signal detection. The method includes: determining the time-frequency domain type corresponding to the detonation signal under different operating conditions; filtering wavelet bases corresponding to multiple time-frequency domain types using preset constraints to obtain a wavelet base candidate library; determining a target wavelet base based on multiple evaluation indicators obtained by decomposing and reconstructing the detonation signal using a first candidate wavelet base, wherein the first candidate wavelet base is any one of the multiple candidate wavelet bases in the wavelet base candidate library; and detecting the detonation signal based on the target wavelet base to obtain a detection result. This application utilizes preset constraints to filter wavelet bases corresponding to multiple time-frequency domain types to establish a wavelet base candidate library. By analyzing multiple evaluation indicators obtained from decomposing and reconstructing the detonation signal, a target wavelet base matching the operating conditions is obtained from the wavelet base candidate library, enabling efficient and robust detonation signal detection.

[0016] The detonation signal detection method of this application can be executed by a computing device, which can be a server, such as a single server, multiple servers, a server cluster, a cloud computing platform, etc. Optionally, the computing device can also be a terminal device, such as a mobile phone, tablet computer, game console, portable computer, desktop computer, advertising machine, all-in-one machine, etc. This application does not limit the type or number of computing devices.

[0017] The following describes a knock signal detection method according to an embodiment of this application, with reference to the accompanying drawings. See Figure 1 The method specifically includes the following steps: S101. Determine the time-frequency domain type of the knock signal under different operating conditions.

[0018] S102. Using preset constraints, the wavelet bases corresponding to multiple time-frequency domain types are screened to obtain a candidate library of wavelet bases.

[0019] S103. Based on the multiple evaluation indicators obtained by decomposing and reconstructing the detonation signal using the first candidate wavelet basis, determine the target wavelet basis.

[0020] The first candidate wavelet basis is any wavelet basis from the pool of candidate wavelet basis bases; S104. Detect the detonation signal based on the target wavelet basis corresponding to the target reconstructed signal to obtain the detection result.

[0021] In some embodiments, the hydrogen engine typically operates under multiple conditions, and the time-frequency domain type of the knock signal generated by the hydrogen engine operating under different conditions is different.

[0022] In some embodiments, the time-frequency domain types include: time-domain transient-dominant, frequency-domain oscillation-dominant, and time-frequency coupled complex types.

[0023] Generally, after acquiring the detonation signal, based on the time-frequency performance of the detonation signal under different operating conditions, it can be summarized into three typical modes: Mode I – Time-domain transient-dominated: Characterized by single or multiple sharp pulses, commonly seen in the initial stage of detonation in hydrogen engines or under low-speed, high-load conditions. The corresponding wavelet bases for this mode are generally Haar, DB2, DB3, etc.

[0024] Mode II – Frequency Domain Oscillation Dominant: Characterized by continuous oscillations around a certain center frequency, commonly seen in cylinder pressure resonance under high-speed operating conditions in hydrogen engines. The corresponding wavelet bases for this mode are generally DB6, DB8, Sym6, etc.

[0025] Mode III – Time-Frequency Coupled Complex: Impulses and oscillations coexist, and the background noise level is high. The corresponding wavelet bases for this mode are generally wavelet bases Coif2, Coif3, Coif4, Bior3.5, etc.

[0026] The wavelet bases corresponding to the above multiple time-frequency domain types can form an initial candidate pool {Haar, DB2, DB3, DB4, DB5, DB6, Sym4, Sym6, Sym8, Coif2, Coif3, Coif4, Bior3.5}. As can be seen, there are many wavelet bases in the initial candidate pool, and it is necessary to combine some methods to filter the wavelet bases in the initial candidate pool.

[0027] In some embodiments, wavelet bases corresponding to multiple time-frequency domain types can be screened using preset constraints to obtain a candidate wavelet base library.

[0028] In some embodiments, the preset constraints include mathematical constraints and engineering constraints. These preset constraints are used to screen wavelet bases corresponding to multiple time-frequency domain types to obtain a candidate wavelet base library. This includes: identifying multiple initial wavelet bases among the multiple time-frequency domain types that satisfy the mathematical constraints, which are: orthogonality or bioorthogonality, compact support, and first-order vanishing moments; identifying multiple candidate wavelet bases among the multiple initial wavelet bases that satisfy the engineering constraints, where the filter length is less than a preset length; and facilitating the creation of a candidate wavelet base library based on these multiple candidate wavelet bases.

[0029] In some embodiments, the wavelet basis can be screened in terms of mathematical constraints based on the characteristics of the detonation signal. First, in order to ensure the objectivity of the detonation signal detection results, that is, the extracted detonation feature energy is real and will not be amplified or reduced due to the transformation process itself, it is necessary to preset the wavelet basis to be orthogonal or biorthogonal. Orthogonal transformation ensures that the total energy of the signal remains unchanged during the "decomposition" and "reconstruction" process, and can ensure that the decomposed coefficients have no information redundancy, and fully express the original signal with the least amount of data.

[0030] Secondly, hydrogen detonation is an extremely short, transient impact. Only wavelets of finite length can precisely capture the exact moment of this impact, like a microscope. If the wavelet is too long, it will mix up the preceding and following signals, making it impossible to determine exactly which millisecond the detonation occurred. Furthermore, in digital signal processing, wavelet transform is essentially a convolution operation. Tight support means that the filter length is finite, and the ECU only needs to perform a finite number of multiplication and addition operations to complete the transform. If the support is too long, the computational load will increase exponentially, causing the ECU to be unable to process it and meet the engine's real-time control requirements.

[0031] Finally, the first-order vanishing moment means that the wavelet function integrates with a constant (DC component) to obtain zero. In simpler terms, the wavelet "ignores" slowly changing, flat parts of the signal (such as background noise and base cylinder pressure), only reacting to rapidly changing parts (such as knock). In engine cylinder pressure signals, the vast majority is a slowly changing background signal. If the wavelet has a first-order or higher vanishing moment, the coefficients generated after decomposition of these flat signals will be very small (approaching zero). In this way, we only need to retain a few large coefficients representing knock to recover the core features, thus achieving efficient signal compression and automatically filtering out slowly changing low-frequency noise.

[0032] Considering that the knock detection algorithm needs to be deployed on an embedded ECU, but the ECU's computing power and storage resources are extremely limited, it is necessary to consider whether the current wavelet basis can be deployed from an application perspective. The wavelet transform process requires caching the detailed coefficients of each layer, and the memory usage is related to the filter length and signal length. For hardware with limited caching, wavelet bases with excessively large filters should be further screened and eliminated.

[0033] After obtaining the wavelet basis candidate library, it is necessary to evaluate each wavelet basis in the candidate library from multiple aspects to determine the final target wavelet basis for detecting detonation signals.

[0034] Generally, to achieve a high energy ratio, wavelet bases need to have a narrow passband, which is accompanied by a long time domain support, causing the reconstructed signal to ring before the impulse, reducing the waveform matching degree. To highlight kurtosis, wavelet bases tend to enhance high-frequency details in the signal, which may simultaneously amplify high-frequency noise, thereby lowering the overall correlation coefficient. Both energy ratio and kurtosis reflect the non-stationary characteristics of the signal, but energy ratio emphasizes the steady-state energy of a specific frequency band, while kurtosis emphasizes the transient energy in the time domain. Over-focusing on a certain frequency band may weaken the high-frequency components of the transient pulse, which is not conducive to improving kurtosis.

[0035] In summary, each wavelet basis can be evaluated from the perspectives of bandwidth energy ratio and signal kurtosis. In addition, in order to ensure the waveform fidelity of the reconstructed signal, it is also necessary to evaluate each wavelet basis from the perspective of Pearson correlation coefficient.

[0036] Since the frequency band energy ratio is calculated based on the coefficients obtained during the detonation signal decomposition process, and the Pearson correlation coefficient and signal kurtosis are obtained based on the reconstructed signal obtained from the detonation signal reconstruction, the above three evaluation indicators can be obtained during the decomposition and reconstruction of the detonation signal by each wavelet basis in the wavelet basis candidate library. Based on the three evaluation indicators of each wavelet basis, the target wavelet basis can be determined from each wavelet basis.

[0037] After determining the target wavelet basis, the acquired detonation signal from the hydrogen engine is first decomposed using wavelet decomposition. Next, only the coefficients of the specific frequency band containing the detonation characteristics (i.e., the detonation characteristic frequency band) are extracted, while the coefficients of the remaining frequency bands (mainly background noise and irrelevant low-frequency trends) are set to zero. Then, wavelet reconstruction is performed. The reconstructed signal obtained in this way is one that has eliminated most of the interference and retains only the core detonation characteristics.

[0038] The reconstructed signal from the extracted detonation characteristic frequency band is first taken as its absolute value, then integrated to obtain the integrated voltage. This integrated voltage reflects the overall magnitude of the detonation characteristic energy within the current signal window. A larger integrated voltage indicates a stronger detonation characteristic in the signal.

[0039] To further acquire background noise and calculate knock intensity, simply looking at the integral voltage is insufficient because the overall signal amplitude will drift under different engine speeds and loads. To achieve accurate detection across all operating conditions, a first-order low-pass filter can be applied to the integral voltage. The low-pass filter removes rapidly changing components in the signal (i.e., sudden knock energy), leaving only the slowly changing portion. This slowly changing portion represents the stable background noise level under the current operating conditions.

[0040] The detonation intensity can be calculated using the following formula: Knock intensity = Integral voltage / Background noise level; Among them, the detonation intensity is the test result.

[0041] The real-time calculated detonation intensity is compared with the pre-calibrated detonation threshold in the system.

[0042] If the knock intensity exceeds the knock threshold, the system determines that knock has occurred. The ECU will immediately receive the signal and take measures such as retarding the ignition timing and enriching the air-fuel mixture to suppress knock.

[0043] If the knock intensity is less than or equal to the knock threshold: the system determines that the current combustion is normal, and the engine continues to operate under the original conditions.

[0044] This application proposes a method, apparatus, device, and storage medium for detonation signal detection. The method includes: determining the time-frequency domain type corresponding to the detonation signal under different operating conditions; filtering wavelet bases corresponding to multiple time-frequency domain types using preset constraints to obtain a wavelet base candidate library; determining a target wavelet base based on multiple evaluation indicators obtained by decomposing and reconstructing the detonation signal using a first candidate wavelet base, wherein the first candidate wavelet base is any wavelet base among multiple candidate wavelet bases in the wavelet base candidate library; and detecting the detonation signal based on the target wavelet base to obtain a detection result. In this application, the embodiments utilize preset constraints to filter wavelet bases corresponding to multiple time-frequency domain types to establish a wavelet base candidate library. By analyzing multiple evaluation indicators obtained by decomposing and reconstructing the detonation signal, a target wavelet base matching the operating conditions is obtained from the wavelet base candidate library, thereby achieving efficient and robust detonation signal detection.

[0045] In some embodiments, the target wavelet basis is determined based on multiple evaluation metrics obtained by decomposing and reconstructing the detonation signal using a first candidate wavelet basis. This includes: performing multi-level decomposition of the detonation signal using the first candidate wavelet basis to obtain detail coefficients and approximation coefficients corresponding to each level of decomposition; calculating the frequency band energy ratio corresponding to the first candidate wavelet basis based on the detail coefficients and approximation coefficients corresponding to each level of decomposition; reconstructing the detonation signal using the first candidate wavelet basis to obtain a reconstructed signal; calculating the Pearson correlation coefficient and signal kurtosis of the first candidate wavelet basis based on the reconstructed signal; and determining the target wavelet basis based on the frequency band energy ratio, Pearson correlation coefficient, and signal kurtosis corresponding to multiple candidate wavelet bases, respectively.

[0046] In determining multiple rating indicators, the detonation signal is first decomposed based on each candidate wavelet basis. This decomposition can involve performing J-level wavelet decomposition on the detonation signal to obtain detail coefficients and approximation coefficients at each level. Here, J is determined by the sampling rate and the characteristic frequency band of the hydrogen engine's detonation, typically ranging from 4 to 6. Further, the frequency band energy ratio is calculated based on the detail coefficients and approximation coefficients at each level. In some embodiments, the band power ratio can be calculated using equation (1): (1) in, The frequency band energy ratio, It is the sum of the energy of the wavelet coefficients corresponding to the detonation characteristic frequency band of the detonation signal. It is the total energy of the wavelet coefficients across all frequency bands of the detonation signal. It is the set of wavelet decomposition layers covered by the detonation characteristic frequency band. It is the wavelet decomposition level index. It is the maximum number of wavelet decomposition layers. It is the sequence of detail coefficients of the k-th wavelet basis at the j-th level. It is the approximation coefficient sequence of the Jth layer of the kth wavelet basis. This ratio measures the wavelet basis's ability to capture energy in the detonation characteristic frequency band, i.e., its energy concentration. It highlights the effective components from a frequency domain perspective; the larger the value, the more powerful the wavelet basis is in compressing detonation energy into the target frequency band and suppressing out-of-band mechanical noise.

[0047] Furthermore, the detonation signal is reconstructed using each candidate wavelet basis to obtain the reconstructed signal, and the Pearson correlation coefficient and signal kurtosis of each candidate wavelet basis are calculated based on the reconstructed signal.

[0048] The Pearson correlation coefficient can be calculated using equation (2): (2) in Here, is the Pearson correlation coefficient, k represents the k-th wavelet basis, N represents the total number of sampling points in a single cyclic signal, and n is the sampling point index. It is the original detonation signal. It is the signal reconstructed from the k-th wavelet basis. It is the original signal The arithmetic mean, It is a reconstructed signal The arithmetic mean of the values. This coefficient measures the degree to which the reconstructed signal retains the waveform of the original signal, i.e., the waveform matching degree. The closer the value is to 1, the less information is lost during the decomposition and reconstruction process of the wavelet basis, and the better the overall structure of the signal is preserved.

[0049] The signal kurtosis can be calculated using equation (3):

[0050] in, The value represents the signal kurtosis. Other symbols are interpreted in the same way as in the formula above. This index reflects the sharpness of the reconstructed signal and is sensitive to transient impact components. The larger the value, the more obvious the impact characteristics.

[0051] Furthermore, the target wavelet basis can be determined from multiple candidate wavelet basis bases based on the frequency band energy ratio, Pearson correlation coefficient, and signal kurtosis corresponding to the multiple candidate wavelet basis bases respectively.

[0052] In some embodiments, the target wavelet basis is determined based on multiple evaluation metrics obtained from decomposing and reconstructing the detonation signal using a first candidate wavelet basis. This includes: performing multi-level decomposition of the detonation signal using the first candidate wavelet basis to obtain detail coefficients and approximation coefficients corresponding to each level of decomposition; calculating the band power ratio corresponding to the first candidate wavelet basis based on the detail coefficients and approximation coefficients corresponding to each level of decomposition; arranging multiple candidate wavelet bases in descending order of their band power ratios; retaining a preset number of candidate wavelet bases at the top of the order to obtain multiple pre-selected wavelet bases; reconstructing the detonation signal using the multiple pre-selected wavelet bases to obtain multiple reconstructed signals; calculating the Pearson correlation coefficient and signal kurtosis of the multiple pre-selected wavelet bases based on the multiple reconstructed signals; and determining the target wavelet basis based on the band power ratio, Pearson correlation coefficient, and signal kurtosis corresponding to the multiple pre-selected wavelet bases.

[0053] In some embodiments, the applicant found that by performing multi-level decomposition on the detonation signal, the detail coefficients and approximation coefficients corresponding to each level of decomposition are obtained, and the frequency band energy ratio can be calculated based on the detail coefficients and approximation coefficients corresponding to each level of decomposition. In order to reduce the calculation pressure of subsequent indicators, multiple pre-selected wavelet bases can be screened based on the frequency band energy ratio.

[0054] In some embodiments, the screening process can be implemented as follows: Since the effective components are highlighted from the frequency domain perspective, the larger the value of the band energy ratio, the more the wavelet basis can compress the detonation energy into the target frequency band, thereby suppressing out-of-band mechanical noise. Therefore, multiple candidate wavelet bases are arranged in descending order of their band energy ratios; a preset number of candidate wavelet bases are retained at the top of the order to obtain multiple pre-selected wavelet bases. The number of the preset number of candidate wavelet bases can be set manually based on the actual situation, generally 2-4.

[0055] Furthermore, the detonation signal is reconstructed based on multiple pre-selected wavelet bases, and the Pearson correlation coefficient and signal kurtosis of each pre-selected wavelet base are calculated to determine the target wavelet base.

[0056] In some embodiments, determining the target wavelet basis based on the band energy ratio, Pearson correlation coefficient, and signal kurtosis corresponding to multiple candidate wavelet bases includes: normalizing the band energy ratio, Pearson correlation coefficient, and signal kurtosis of a first pre-selected wavelet basis to obtain normalized band energy ratio, Pearson correlation coefficient, and signal kurtosis, wherein the first pre-selected wavelet basis is any one of the multiple pre-selected wavelet bases; obtaining preset weight coefficients corresponding to the normalized band energy ratio, Pearson correlation coefficient, and signal kurtosis; calculating a comprehensive score of the first pre-selected wavelet basis based on the normalized band energy ratio, Pearson correlation coefficient, and signal kurtosis and the corresponding preset weight coefficients; and determining the pre-selected wavelet basis with the largest comprehensive score as the target wavelet basis.

[0057] In some embodiments, the three indicators mentioned above have different physical dimensions and values, for example... A comprehensive score cannot be directly calculated based on the above three indicators; therefore, it is necessary to normalize these three indicators.

[0058] In some embodiments, the range normalization method can be used to map all indices to the [0,1] interval. The normalized pairwise power ratio, Pearson correlation coefficient and signal kurtosis corresponding to the first preselected wavelet basis are as shown in equations (4)-(6): (4)

[0059] (6) in, The normalized Pearson correlation coefficient, The largest Pearson correlation coefficient among multiple pre-selected wavelet basis coefficients. The smallest Pearson correlation coefficient among multiple pre-selected wavelet basis coefficients. This is the original Pearson correlation coefficient.

[0060] The normalized frequency band energy ratio, The highest frequency band energy ratio among multiple pre-selected wavelet bases. This refers to the minimum frequency band energy ratio among multiple pre-selected wavelet bases. This represents the original frequency band energy ratio.

[0061] For the normalized signal kurtosis, The maximum signal kurtosis among the signal kurtosis corresponding to multiple pre-selected wavelet bases. The minimum signal kurtosis among multiple pre-selected wavelet basis signals. This represents the kurtosis of the original signal.

[0062] Furthermore, the preset weight coefficients corresponding to the normalized pairwise energy ratio, Pearson correlation coefficient, and signal kurtosis are obtained respectively; the comprehensive score of the first pre-selected wavelet basis is calculated based on the normalized pairwise energy ratio, Pearson correlation coefficient, signal kurtosis, and the corresponding preset weight coefficients; the pre-selected wavelet basis with the largest comprehensive score is determined as the target wavelet basis.

[0063] In some embodiments, a comprehensive score for the first pre-selected wavelet basis is calculated based on the normalized pairwise energy ratio, Pearson correlation coefficient, signal kurtosis, and corresponding preset weight coefficients. This includes: obtaining the signal-to-noise ratio (SNR) of the detonation signal; if the SNR is greater than or equal to a preset SNR threshold, determining the normalized pairwise energy ratio, Pearson correlation coefficient, signal kurtosis, and corresponding preset weight coefficients as the normalized pairwise energy ratio, Pearson correlation coefficient, signal kurtosis, and corresponding target weight coefficients; and based on the normalized pairwise energy ratio, Pearson correlation coefficient, signal kurtosis, and corresponding target weight coefficients, Calculate the comprehensive score of the first pre-selected wavelet basis; when the signal-to-noise ratio is less than the preset signal-to-noise ratio threshold, calculate the target weight coefficient corresponding to the normalized band energy ratio based on the preset weight coefficient corresponding to the normalized band energy ratio, the preset signal-to-noise ratio threshold, and the signal-to-noise ratio; based on the target weight coefficient corresponding to the normalized band energy ratio, calculate the normalized band energy ratio, Pearson correlation coefficient, signal kurtosis, and corresponding target weight coefficient; calculate the comprehensive score of the first pre-selected wavelet basis based on the normalized band energy ratio, Pearson correlation coefficient, signal kurtosis, and corresponding target weight coefficient.

[0064] In some embodiments, fixed weights cannot adapt to signal characteristic drift caused by drastic engine changes. Therefore, this invention proposes an adaptive weight adjustment mechanism based on signal-to-noise ratio (SNR) estimation. The SNR of the current frame is estimated using the energy of non-knock history segments or non-characteristic frequency bands of the signal frame.

[0065] Specifically, when the signal-to-noise ratio (SNR) exceeds a preset SNR threshold, the signal is considered sufficiently good. Therefore, the normalized pairwise energy ratio, Pearson correlation coefficient, signal kurtosis, and their corresponding preset weighting coefficients are not changed; instead, they are stabilized at the preset maximum fidelity weights. When the SNR is below the preset SNR threshold, the signal is not very good. In this case, when the SNR is high, the preset weighting coefficients of the normalized Pearson correlation coefficient and signal kurtosis should be increased to fully preserve signal details and impact characteristics; when the SNR is low, the preset weighting coefficient corresponding to the normalized pairwise energy ratio should be increased to enhance out-of-band noise suppression capabilities.

[0066] Specifically, a preset signal-to-noise ratio (SNR) threshold can be set. When the SNR is greater than or equal to the preset SNR threshold, the normalized pairwise energy ratio, Pearson correlation coefficient, and signal kurtosis, along with their corresponding preset weight coefficients, are determined as the normalized pairwise energy ratio, Pearson correlation coefficient, and signal kurtosis, along with their corresponding target weight coefficients. Based on the normalized pairwise energy ratio, Pearson correlation coefficient, and signal kurtosis, along with their corresponding target weight coefficients, a comprehensive score for the first pre-selected wavelet basis is calculated. When the SNR is less than the preset SNR threshold, based on the normalized pairwise energy ratio, the preset weight coefficients, the preset SNR threshold, and the SNR, the target weight coefficients corresponding to the normalized pairwise energy ratio are calculated. Based on the target weight coefficients corresponding to the normalized pairwise energy ratio, the normalized pairwise energy ratio, Pearson correlation coefficient, and signal kurtosis, along with their corresponding target weight coefficients, are calculated. Finally, based on the normalized pairwise energy ratio, Pearson correlation coefficient, and signal kurtosis, a comprehensive score for the first pre-selected wavelet basis is calculated.

[0067] The preset signal-to-noise ratio threshold can be flexibly set based on the actual situation, and will not be elaborated here.

[0068] In some embodiments, when the signal-to-noise ratio is greater than or equal to a preset signal-to-noise ratio threshold, the normalized pairwise power ratio, Pearson correlation coefficient, signal kurtosis, and corresponding target weight coefficients can be calculated using equations (7)-(8):

[0069]

[0070] in The target weighting coefficients corresponding to the normalized frequency band energy ratio, The target weighting coefficients for the normalized Pearson correlation coefficient. The target weighting coefficients for the normalized signal kurtosis. It is the signal-to-noise ratio threshold. Signal-to-noise ratio; Preset weighting coefficients for the normalized frequency band energy ratio; This represents the desired minimum signal-to-noise ratio (SNR). The SNR threshold and minimum SNR were determined through offline benchtop testing.

[0071] Finally, the comprehensive score of the first preselected wavelet basis is calculated based on the normalized band energy ratio, Pearson correlation coefficient, signal kurtosis, and corresponding target weight coefficients, as shown in Equation (9): (9) in, For comprehensive scoring, the interpretations of other symbols are the same as those in the formula above.

[0072] In some embodiments, to illustrate the determination of the target wavelet basis based on the band energy ratio, Pearson correlation coefficient, and signal kurtosis corresponding to multiple pre-selected wavelet bases, this application provides a flowchart for determining the target wavelet basis, as follows: Figure 2 As shown, the method includes the following steps: S201. Normalize the frequency band energy ratio, Pearson correlation coefficient, and signal kurtosis of multiple pre-selected wavelet bases to obtain normalized frequency band energy ratio, Pearson correlation coefficient, and signal kurtosis.

[0073] S202. Obtain the preset weighting coefficients corresponding to the normalized pairwise power ratio, Pearson correlation coefficient and signal kurtosis, respectively.

[0074] S203. Determine whether the signal-to-noise ratio of the detonation signal is greater than or equal to the signal-to-noise ratio threshold.

[0075] If the signal-to-noise ratio of the detonation signal is greater than or equal to the signal-to-noise ratio threshold, execute S204 to determine the normalized pairwise energy ratio, Pearson correlation coefficient, signal kurtosis, and corresponding preset weight coefficients as the normalized pairwise energy ratio, Pearson correlation coefficient, signal kurtosis, and corresponding target weight coefficients. When the signal-to-noise ratio of the detonation signal is less than the signal-to-noise ratio threshold, execute S205, and calculate the target weight coefficient corresponding to the normalized band energy ratio based on the preset weight coefficient, the signal-to-noise ratio threshold, and the signal-to-noise ratio. S206. Based on the target weight coefficient corresponding to the normalized pairband energy ratio, calculate the normalized pairband energy ratio, Pearson correlation coefficient, signal kurtosis, and corresponding target weight coefficient. S207. Calculate the comprehensive score of the first preselected wavelet basis based on the normalized pair band energy ratio, Pearson correlation coefficient, signal kurtosis, and corresponding target weight coefficient.

[0076] This application also provides a detonation signal detection device, which is used to execute the model training method provided in any of the above embodiments. Figure 3 As shown, the device includes: a determination module 301, a screening module 302, and a detection module 303.

[0077] The determination module 301 is used to determine the time-frequency domain type of the knock signal under different operating conditions; The filtering module 302 is used to filter the wavelet bases corresponding to the multiple time-frequency domain types using preset constraints to obtain a candidate library of wavelet bases. The determining module 301 is further configured to determine the target wavelet basis based on multiple evaluation indicators obtained by decomposing and reconstructing the detonation signal using the first candidate wavelet basis, wherein the first candidate wavelet basis is any one of the multiple candidate wavelet bases in the wavelet base candidate library. The detection module 303 is used to detect the detonation signal based on the target wavelet basis and obtain the detection result.

[0078] This application proposes a method, apparatus, device, and storage medium for detonation signal detection. The method includes: determining the time-frequency domain type corresponding to the detonation signal under different operating conditions; filtering wavelet bases corresponding to multiple time-frequency domain types using preset constraints to obtain a wavelet base candidate library; determining a target wavelet base based on multiple evaluation indicators obtained by decomposing and reconstructing the detonation signal using a first candidate wavelet base, wherein the first candidate wavelet base is any wavelet base among multiple candidate wavelet bases in the wavelet base candidate library; and detecting the detonation signal based on the target wavelet base to obtain a detection result. In this application, the embodiments utilize preset constraints to filter wavelet bases corresponding to multiple time-frequency domain types to establish a wavelet base candidate library. By analyzing multiple evaluation indicators obtained by decomposing and reconstructing the detonation signal, a target wavelet base matching the operating conditions is obtained from the wavelet base candidate library, thereby achieving efficient and robust detonation signal detection.

[0079] In some embodiments, the preset constraints include mathematical constraints and engineering constraints, and the filtering module 302 is specifically used for: Multiple initial wavelet bases that satisfy the mathematical constraints are determined from the wavelet bases corresponding to the multiple time-frequency domain types. The mathematical constraints are that they satisfy the following properties: orthogonality or bioorthogonality, compact support and first-order vanishing moment. Determine multiple candidate wavelet bases among the multiple initially screened wavelet bases that satisfy the engineering constraint condition, wherein the engineering constraint condition is that the filter length is less than a preset length; A candidate wavelet basis library is formed based on the aforementioned candidate wavelet basis bases.

[0080] In some embodiments, the time-frequency domain type includes: time-domain transient-dominant, frequency-domain oscillation-dominant, and time-frequency coupled complex types.

[0081] In some embodiments, the determining module 301 is specifically used for: The detonation signal is decomposed into multiple layers based on the first candidate wavelet basis to obtain the detail coefficients and approximation coefficients corresponding to each layer of decomposition. The frequency band energy ratio corresponding to the first candidate wavelet basis is calculated based on the detail coefficients and approximation coefficients corresponding to each decomposition level. The detonation signal is reconstructed based on the first candidate wavelet basis to obtain the reconstructed signal; The Pearson correlation coefficient and signal kurtosis of the first candidate wavelet basis are calculated based on the reconstructed signal. The target wavelet basis is determined based on the frequency band energy ratio, Pearson correlation coefficient, and signal kurtosis corresponding to multiple candidate wavelet bases.

[0082] In some embodiments, the determining module 301 is further specifically used for: The detonation signal is decomposed into multiple layers based on the first candidate wavelet basis to obtain the detail coefficients and approximation coefficients corresponding to each layer of decomposition. The frequency band energy ratio corresponding to the first candidate wavelet basis is calculated based on the detail coefficients and approximation coefficients corresponding to each decomposition level. Multiple candidate wavelet bases are arranged in descending order of their bandwidth energy ratio; Retain the first few candidate wavelet bases in the order of sorting to obtain multiple pre-selected wavelet bases; The detonation signal is reconstructed based on the multiple pre-selected wavelet bases to obtain multiple reconstructed signals; Based on the multiple reconstructed signals, calculate the Pearson correlation coefficient and signal kurtosis of the multiple preselected wavelet bases respectively; The target wavelet basis is determined based on the frequency band energy ratio, Pearson correlation coefficient, and signal kurtosis corresponding to the multiple pre-selected wavelet bases.

[0083] In some embodiments, the determining module 301 is further specifically used for: The frequency band energy ratio, Pearson correlation coefficient and signal kurtosis of the first preselected wavelet basis are normalized to obtain the normalized frequency band energy ratio, Pearson correlation coefficient and signal kurtosis. The first preselected wavelet basis is any one of the plurality of preselected wavelet bases. Obtain the preset weighting coefficients corresponding to the normalized pairwise power ratio, Pearson correlation coefficient, and signal kurtosis, respectively; The comprehensive score of the first preselected wavelet basis is calculated based on the normalized pairwise energy ratio, Pearson correlation coefficient, signal kurtosis, and corresponding preset weight coefficients. The pre-selected wavelet basis with the highest comprehensive score is determined as the target wavelet basis.

[0084] In some embodiments, the determining module 301 is further specifically used for: Obtain the signal-to-noise ratio of the detonation signal; When the signal-to-noise ratio is greater than or equal to the signal-to-noise ratio threshold, the normalized pairwise power ratio, Pearson correlation coefficient, signal kurtosis, and corresponding preset weight coefficients are determined as the normalized pairwise power ratio, Pearson correlation coefficient, signal kurtosis, and corresponding target weight coefficients. Based on the normalized pairwise energy ratio, Pearson correlation coefficient, signal kurtosis, and corresponding target weight coefficient, the comprehensive score of the first preselected wavelet basis is calculated. When the signal-to-noise ratio is less than the signal-to-noise ratio threshold, the target weight coefficient corresponding to the normalized band energy ratio is calculated based on the preset weight coefficient corresponding to the normalized band energy ratio, the signal-to-noise ratio threshold, and the signal-to-noise ratio. Based on the target weight coefficient corresponding to the normalized pairband energy ratio, calculate the normalized pairband energy ratio, Pearson correlation coefficient, signal kurtosis, and corresponding target weight coefficient; The comprehensive score of the first preselected wavelet basis is calculated based on the normalized pairwise energy ratio, Pearson correlation coefficient, signal kurtosis, and corresponding target weight coefficient.

[0085] This application also provides an electronic device for performing the above-described knock signal detection method. Please refer to... Figure 4 It illustrates a schematic diagram of an electronic device provided by some embodiments of this application. For example... Figure 4 As shown, the electronic device 7 includes: a processor 700, a memory 701, a bus 702, and a communication interface 703. The processor 700, the communication interface 703, and the memory 701 are connected via the bus 702. The memory 701 stores a computer program that can run on the processor 700. When the processor 700 runs the computer program, it executes the detonation signal detection method provided in any of the foregoing embodiments of this application.

[0086] The memory 701 may include high-speed random access memory (RAM) or non-volatile memory, such as at least one disk storage device. Communication between the device network element and at least one other network element is achieved through at least one communication interface 703 (which can be wired or wireless), such as the Internet, wide area network, local area network, metropolitan area network, etc.

[0087] Bus 702 can be an ISA bus, PCI bus, or EISA bus, etc. The bus can be divided into address bus, data bus, control bus, etc. The memory 701 is used to store programs. After receiving an execution instruction, the processor 700 executes the program. The detonation signal detection method disclosed in any of the aforementioned embodiments of this application can be applied to the processor 700, or implemented by the processor 700.

[0088] The processor 700 may be an integrated circuit chip with signal processing capabilities. In implementation, each step of the above method can be completed by the integrated logic circuitry in the hardware of the processor 700 or by instructions in software form. The processor 700 may be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it may also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), an off-the-shelf programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. It can implement or execute the methods, steps, and logic block diagrams disclosed in the embodiments of this application. The general-purpose processor may be a microprocessor or any conventional processor. The steps of the methods disclosed in the embodiments of this application can be directly embodied in the execution of a hardware decoding processor, or executed by a combination of hardware and software modules in the decoding processor. The software modules may reside in random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, registers, or other mature storage media in the art. The storage medium is located in memory 701. Processor 700 reads the information in memory 701 and, in conjunction with its hardware, completes the steps of the above method.

[0089] The electronic device provided in this application embodiment and the knock signal detection method provided in this application embodiment are based on the same inventive concept and have the same beneficial effects as the methods they adopt, operate or implement.

[0090] This application also provides a computer-readable storage medium corresponding to the detonation signal detection method provided in the foregoing embodiments. Please refer to... Figure 5 The computer-readable storage medium shown is an optical disc 30, on which a computer program (i.e., a program product) is stored. When the computer program is run by a processor, it executes the knock signal detection method provided in any of the aforementioned embodiments.

[0091] It should be noted that examples of computer-readable storage media may also include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other optical and magnetic storage media, which will not be elaborated here.

[0092] The computer-readable storage medium provided in the above embodiments of this application and the knock signal detection method provided in the embodiments of this application are based on the same inventive concept and have the same beneficial effects as the methods adopted, run or implemented by the application programs stored therein.

[0093] It should be noted that: Numerous specific details are set forth in the specification provided herein. However, it will be understood that embodiments of this application may be practiced without these specific details. In some instances, well-known structures and techniques have not been shown in detail so as not to obscure the understanding of this specification.

[0094] Similarly, it should be understood that, for the sake of brevity and to aid in understanding one or more of the various inventive aspects, in the above description of exemplary embodiments of this application, various features of this application are sometimes grouped together in a single embodiment, figure, or description thereof. However, this disclosure should not be construed as reflecting a schematic diagram in which the claimed application requires more features than expressly recited in each claim. Rather, as reflected in the following claims, inventive aspects lie in fewer than all features of a single foregoing disclosed embodiment. Therefore, the claims following the detailed description are hereby expressly incorporated into that detailed description, wherein each claim itself is a separate embodiment of this application.

[0095] Furthermore, those skilled in the art will understand that although some embodiments herein include certain features included in other embodiments but not others, combinations of features from different embodiments are intended to be within the scope of this application and form different embodiments. For example, in the following claims, any of the claimed embodiments can be used in any combination.

[0096] The above are merely preferred embodiments of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for detecting detonation signals, characterized in that, The method includes: Determine the time-frequency domain type of the knock signal under different operating conditions; By using preset constraints, wavelet bases corresponding to multiple time-frequency domain types are screened to obtain a candidate library of wavelet bases; Based on multiple evaluation indicators obtained by decomposing and reconstructing the detonation signal using the first candidate wavelet basis, a target wavelet basis is determined. The first candidate wavelet basis is any one of the multiple candidate wavelet bases in the wavelet basis candidate library. The detonation signal is detected based on the target wavelet basis, and the detection result is obtained.

2. The method according to claim 1, characterized in that, The preset constraints include mathematical constraints and engineering constraints. The step of using these preset constraints to filter wavelet bases corresponding to multiple time-frequency domain types to obtain a candidate wavelet base library includes: Multiple initial wavelet bases that satisfy the mathematical constraints are determined from the wavelet bases corresponding to the multiple time-frequency domain types. The mathematical constraints are that they satisfy the following properties: orthogonality or bioorthogonality, compact support and first-order vanishing moment. Determine multiple candidate wavelet bases among the multiple initially screened wavelet bases that satisfy the engineering constraint condition, wherein the engineering constraint condition is that the filter length is less than a preset length; A candidate wavelet basis library is formed based on the aforementioned candidate wavelet basis bases.

3. The method according to claim 1, characterized in that, The time-frequency domain types include: time-domain transient-dominant, frequency-domain oscillation-dominant, and time-frequency coupled complex types.

4. The method according to claim 1, characterized in that, The determination of the target wavelet basis based on multiple evaluation indicators obtained by decomposing and reconstructing the detonation signal using the first candidate wavelet basis includes: The detonation signal is decomposed into multiple layers based on the first candidate wavelet basis to obtain the detail coefficients and approximation coefficients corresponding to each layer of decomposition. The frequency band energy ratio corresponding to the first candidate wavelet basis is calculated based on the detail coefficients and approximation coefficients corresponding to each decomposition level. The detonation signal is reconstructed based on the first candidate wavelet basis to obtain the reconstructed signal; The Pearson correlation coefficient and signal kurtosis of the first candidate wavelet basis are calculated based on the reconstructed signal. The target wavelet basis is determined based on the frequency band energy ratio, Pearson correlation coefficient, and signal kurtosis corresponding to multiple candidate wavelet bases.

5. The method according to claim 1, characterized in that, The determination of the target wavelet basis based on multiple evaluation indicators obtained by decomposing and reconstructing the detonation signal using the first candidate wavelet basis includes: The detonation signal is decomposed into multiple layers based on the first candidate wavelet basis to obtain the detail coefficients and approximation coefficients corresponding to each layer of decomposition. The frequency band energy ratio corresponding to the first candidate wavelet basis is calculated based on the detail coefficients and approximation coefficients corresponding to each decomposition level. Multiple candidate wavelet bases are arranged in descending order of their bandwidth energy ratio; Retain the first few candidate wavelet bases in the order of sorting to obtain multiple pre-selected wavelet bases; The detonation signal is reconstructed based on the multiple pre-selected wavelet bases to obtain multiple reconstructed signals; Based on the multiple reconstructed signals, calculate the Pearson correlation coefficient and signal kurtosis of the multiple preselected wavelet bases respectively; The target wavelet basis is determined based on the frequency band energy ratio, Pearson correlation coefficient, and signal kurtosis corresponding to the multiple pre-selected wavelet bases.

6. The method according to claim 5, characterized in that, The determination of the target wavelet basis based on the frequency band energy ratio, Pearson correlation coefficient, and signal kurtosis corresponding to multiple candidate wavelet bases includes: The frequency band energy ratio, Pearson correlation coefficient and signal kurtosis of the first preselected wavelet basis are normalized to obtain the normalized frequency band energy ratio, Pearson correlation coefficient and signal kurtosis. The first preselected wavelet basis is any one of the plurality of preselected wavelet bases. Obtain the preset weighting coefficients corresponding to the normalized pairwise power ratio, Pearson correlation coefficient, and signal kurtosis, respectively; The comprehensive score of the first preselected wavelet basis is calculated based on the normalized pairwise energy ratio, Pearson correlation coefficient, signal kurtosis, and corresponding preset weight coefficients. The pre-selected wavelet basis with the highest comprehensive score is determined as the target wavelet basis.

7. The method according to claim 1, characterized in that, The calculation of the comprehensive score of the first preselected wavelet basis based on the normalized band power ratio, Pearson correlation coefficient, signal kurtosis, and corresponding preset weight coefficients includes: Obtain the signal-to-noise ratio of the detonation signal; When the signal-to-noise ratio is greater than or equal to a preset signal-to-noise ratio threshold, the normalized pairwise power ratio, Pearson correlation coefficient, signal kurtosis, and corresponding preset weight coefficients are determined as the normalized pairwise power ratio, Pearson correlation coefficient, signal kurtosis, and corresponding target weight coefficients. Based on the normalized pairwise energy ratio, Pearson correlation coefficient, signal kurtosis, and corresponding target weight coefficient, the comprehensive score of the first preselected wavelet basis is calculated. When the signal-to-noise ratio is less than the preset signal-to-noise ratio threshold, the target weight coefficient corresponding to the normalized band energy ratio is calculated based on the preset weight coefficient corresponding to the normalized band energy ratio, the preset signal-to-noise ratio threshold, and the signal-to-noise ratio. Based on the target weight coefficient corresponding to the normalized pairband energy ratio, calculate the normalized pairband energy ratio, Pearson correlation coefficient, signal kurtosis, and corresponding target weight coefficient; The comprehensive score of the first preselected wavelet basis is calculated based on the normalized pairwise energy ratio, Pearson correlation coefficient, signal kurtosis, and corresponding target weight coefficient.

8. A detonation signal detection device, characterized in that, include: The determination module is used to determine the time-frequency domain type of the knock signal under different operating conditions; The filtering module is used to filter the wavelet bases corresponding to the multiple time-frequency domain types using preset constraints to obtain a candidate wavelet base library. The determining module is further configured to determine the target wavelet basis based on multiple evaluation indicators obtained by decomposing and reconstructing the detonation signal using the first candidate wavelet basis, wherein the first candidate wavelet basis is any one of the multiple candidate wavelet bases in the wavelet base candidate library. The detection module is used to detect the detonation signal based on the target wavelet basis and obtain the detection result.

9. An electronic 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 method as described in any one of claims 1-7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, The program is executed by a processor to implement the method as described in any one of claims 1-7.

Images