Water turbine audio signal separation method, device, equipment and medium

Abstract

The invention provides a water turbine audio signal separation method, device and equipment and a medium. The method comprises the following steps: acquiring an original audio signal of a water turbine in a running state; preprocessing the original audio signal to obtain an intermediate audio signal; performing short-time Fourier transform on the intermediate audio signal to obtain a Fourier transform result; performing inverse short-time Fourier transform on the time-frequency points of which the energy values are higher than a preset threshold value in the Fourier transform result, and constructing a final audio signal; performing modal decomposition on the final audio signal to obtain a series of modal functions; performing independent component analysis on the modal function, and separating an independent signal based on non-Gaussian property; the audio signal generated by cavitation of the water turbine can be quickly separated from the audio signal in normal operation, and whether the water turbine has cavitation or not can be judged.

Description

Technical Field

[0001] The present invention relates to the technical field of water turbines, and in particular to a water turbine audio signal separation method, device, equipment and medium. Background Art

[0002] The hydro-turbine generator set is the core equipment of the hydropower station. The stable operation of the unit is of great significance for avoiding safety hazards and improving the economic benefits of power generation enterprises. Cavitation is a common destructive phenomenon inside hydraulic machinery. The existence of cavitation will reduce the efficiency of the unit, increase the vibration of the unit and the wear of the blades, resulting in a significant reduction in the working life of the turbine unit and causing huge economic losses. Therefore, the determination of cavitation is particularly important. However, in the prior art, audio signals are usually used to determine whether a turbine has cavitation. However, it is difficult to identify cavitation signals from audio signals, resulting in low accuracy in cavitation determination. Summary of the invention

[0003] In response to the above-mentioned problems, the present invention proposes a method, device, equipment and medium for separating audio signals of a water turbine, which solves the technical problem that in the prior art, audio signals are used to judge whether a water turbine is cavitating, but it is difficult to identify cavitation signals from audio signals, resulting in low accuracy of cavitation judgment. The method can quickly separate the audio signals generated by water turbine cavitation from the audio signals of normal operation, which is conducive to judging whether a water turbine is cavitating.

[0004] An embodiment of the present invention provides a method for separating a turbine audio signal, comprising:

[0005] Collect the original audio signal of the turbine in operation;

[0006] Preprocessing the original audio signal to obtain an intermediate audio signal;

[0007] Performing a short-time Fourier transform on the intermediate audio signal to obtain a Fourier transform result;

[0008] Perform inverse short-time Fourier transform on the time-frequency points whose energy values ​​in the Fourier transform results are higher than a preset threshold to construct the final audio signal;

[0009] Performing modal decomposition on the final audio signal to obtain a series of modal functions;

[0010] Independent component analysis is performed on the modal functions, and independent signals are separated based on non-Gaussianity.

[0011] In some embodiments, preprocessing the original audio signal to obtain the intermediate audio signal includes:

[0012] Assume that the original audio signal is x(t), and the original audio signal x(t) is denoised by a bandpass filter, as shown in the following formula:

[0013] y(t)=x(t)*h(t);

[0014] Where y(t) is the intermediate audio signal and h(t) is the impulse response of the bandpass filter.

[0015] In some embodiments, performing a short-time Fourier transform on the intermediate audio signal to obtain a Fourier transform result includes:

[0016] Perform short-time Fourier transform on the intermediate audio signal y(t) to obtain a Fourier transform result:

[0017]

[0018] Where S(ω,τ) is expressed in the time-frequency domain, t is time, ω is the angular frequency, τ is the time offset, ω(t-τ) is the window function, and j is the imaginary unit.

[0019] In some embodiments, performing an inverse short-time Fourier transform on the time-frequency points in the Fourier transform result whose energy values ​​are higher than a preset threshold to construct the final audio signal includes:

[0020] Calculate the energy value of each time-frequency point of the Fourier transform result:

[0021] E(ω,τ)=|S(ω,τ)| 2 ;

[0022] In the formula, E(ω,τ) is the energy value corresponding to the time-frequency point;

[0023] Arrange all calculated energy values ​​E(ω,τ) in the order of time and frequency to construct a time-frequency energy graph;

[0024] Compare the energy value E(ω,τ) of each time-frequency point with the preset threshold, retain the time-frequency points with energy values ​​higher than the preset threshold, and record the time-frequency points higher than the preset threshold as S′(ω,τ);

[0025] Perform inverse short-time Fourier transform on the time-frequency points whose energy values ​​in the Fourier transform results are higher than the preset threshold to construct the final audio signal:

[0026]

[0027] Where z(t) is the final audio signal.

[0028] In some embodiments, the performing modal decomposition on the final audio signal to obtain a series of modal functions includes:

[0029] The final audio signal z(t) is subjected to empirical mode decomposition, as shown in the following formula:

[0030]

[0031] Among them, IFM i (t) is the i-th mode function, r(t) is the residual term, and N is the number of modes.

[0032] In some embodiments, performing independent component analysis on the modal function and separating independent signals based on non-Gaussianity includes:

[0033] For the i-th mode function IFM i (t) is centralized to make its mean zero. The centralization is carried out using the following formula:

[0034] IFM centered (t) = IFM i (t)-E[IFM i (t)];

[0035] In the formula, IFM centered (t) is the modal function after centralization, E[IFM i (t)] is the expected value;

[0036] The centered mode function is whitened so that its covariance matrix becomes the identity matrix:

[0037] IFM whitened (t) = BIFM centered (t);

[0038] In the formula, IFM whitened (t) is the whitened mode function, and B is the whitening matrix;

[0039] The Fast ICA algorithm is used to find the separation matrix W that maximizes the non-Gaussianity of the signal by iteratively optimizing the negative entropy:

[0040]

[0041] In the formula, J(y) represents negative entropy. The larger the value of negative entropy J(y), the stronger the non-Gaussianity of signal y is. i ) is the separation signal y i The probability density function, q(y i ) is the probability density function of the standard normal distribution;

[0042] Use the gradient descent method to maximize the negative entropy J(y) and iteratively update the separation matrix W;

[0043] Using the estimated separation matrix W, separate the independent signals:

[0044] s(t)=WIFMM whitened (t);

[0045] Where s(t) is an independent signal.

[0046] The embodiment of the present invention provides a hydraulic turbine audio signal separation device, comprising:

[0047] An acquisition module, used for acquiring the original audio signal of the turbine in operation;

[0048] A preprocessing module, used for preprocessing the original audio signal to obtain an intermediate audio signal;

[0049] A short-time Fourier transform module, used for performing a short-time Fourier transform on the intermediate audio signal to obtain a Fourier transform result;

[0050] An inverse short-time Fourier transform module is used to perform an inverse short-time Fourier transform on the time-frequency points whose energy values ​​in the Fourier transform results are higher than a preset threshold value to construct a final audio signal;

[0051] A modal decomposition module, used for performing modal decomposition on the final audio signal to obtain a series of modal functions;

[0052] The independent component analysis module is used to perform independent component analysis on the modal function and separate independent signals based on non-Gaussianity.

[0053] An embodiment of the present invention provides an electronic device, including:

[0054] A memory and a processor, wherein a computer program is stored in the memory, and when the computer program is executed by the processor, any one of the above methods is executed.

[0055] An embodiment of the present invention provides a storage medium, and a computer program stored in the storage medium can be executed by one or more processors and can be used to implement any of the above methods.

[0056] Compared with the prior art, the present invention has the following beneficial effects:

[0057] The original audio signal of a running turbine is collected, the original audio signal is preprocessed to obtain an intermediate audio signal, the intermediate audio signal is short-time Fourier transformed to obtain a Fourier transform result, the time-frequency points with energy values ​​higher than a preset threshold in the Fourier transform result are inversely short-time Fourier transformed to construct a final audio signal, the final audio signal is modally decomposed to obtain a series of modal functions, the modal functions are independently component analyzed, and independent signals are separated based on non-Gaussianity; the audio signal generated by turbine cavitation can be quickly separated from the audio signal of normal operation, which is conducive to judging whether cavitation exists in the turbine. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] The embodiments of the present invention are further described below with reference to the accompanying drawings:

[0059] Figure 1 A schematic diagram of a flow chart of a method for separating audio signals of a hydraulic turbine provided by an embodiment of the present invention;

[0060] Figure 2 A schematic structural diagram of a hydraulic turbine audio signal separation device provided by an embodiment of the present invention;

[0061] Figure 3 A schematic diagram of the structure of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0062] In order to make the purpose, technical solutions and advantages of the present invention clearer, the present invention will be further described in detail below in conjunction with the accompanying drawings. The described embodiments should not be regarded as limiting the present invention. All other embodiments obtained by ordinary technicians in the field without making creative work are within the scope of protection of the present invention.

[0063] In the following description, reference is made to “some embodiments”, which describe a subset of all possible embodiments, but it will be understood that “some embodiments” may be the same subset or different subsets of all possible embodiments and may be combined with each other without conflict.

[0064] If similar descriptions of "first\second\third" appear in the application documents, the following instructions are added. In the following description, the terms "first\second\third" involved are only used to distinguish similar objects and do not represent a specific order for the objects. It can be understood that "first\second\third" can be interchanged in a specific order or sequence where permitted, so that the embodiments of the present invention described herein can be implemented in an order other than that illustrated or described herein.

[0065] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art to which the present invention belongs. The terms used herein are only for the purpose of describing the embodiments of the present invention and are not intended to limit the present invention.

[0066] Based on the problems existing in the related art, an embodiment of the present invention provides a method for separating audio signals of a turbine, and the execution subject of the separation method can be an electronic device. The electronic device can be various types of terminals such as a laptop computer, a tablet computer, a desktop computer, a set-top box, a mobile device (for example, a mobile phone, a portable music player, a personal digital assistant, a dedicated messaging device, a portable gaming device), and can also be implemented as a server. The server can be an independent physical server, or a server cluster or distributed system composed of multiple physical servers, or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, content delivery networks (Content Delivery Network, CDN), and big data and artificial intelligence platforms.

[0067] In some embodiments, the functions implemented by the separation method provided in the embodiments of the present invention may be implemented by a processor of an electronic device calling a program code, wherein the program code may be stored in a computer storage medium.

[0068] The embodiment of the present invention provides a method for separating audio signals of a water turbine. Figure 1 A schematic diagram of the implementation flow of a method for separating a water turbine audio signal provided by an embodiment of the present invention is shown in FIG. Figure 1 As shown, including:

[0069] Step S1: collecting the original audio signal of the turbine in operation;

[0070] In the embodiment of the present invention, an acoustic sensor suitable for the turbine environment can be selected, such as a microphone, a sound pressure meter, a microphone or a hydrophone. These sensors can capture the sound waves generated when the turbine is running. The sensors can be reasonably arranged according to the structure of the turbine and the expected sound source location. The sensor should be installed in a position where the sound signal generated by cavitation and other phenomena can be captured to the maximum extent, and at the same time, as far away from strong noise sources as possible to reduce the interference of background noise. According to the sound frequency range that may be generated when the turbine is running, a suitable sampling frequency is set to collect and store the audio signal.

[0071] Step S2: preprocessing the original audio signal to obtain an intermediate audio signal;

[0072] In some embodiments, the step S2 comprises:

[0073] Step S21: Assume that the original audio signal is x(t), and perform denoising on the original audio signal x(t) through a bandpass filter, as shown in the following formula:

[0074] y(t)=x(t)*h(t);

[0075] Where y(t) is the intermediate audio signal and h(t) is the impulse response of the bandpass filter.

[0076] In the embodiment of the present invention, the original audio signal x(t) is denoised by a bandpass filter. By selecting appropriate filter coefficients and design parameters, background noise and interference in the original audio signal x(t) can be effectively removed, thereby improving the signal-to-noise ratio and sound quality of the original audio signal x(t).

[0077] Step S3: performing short-time Fourier transform on the intermediate audio signal to obtain a Fourier transform result;

[0078] In some embodiments, step S3 includes:

[0079] Step S31: Perform short-time Fourier transform on the intermediate audio signal y(t) to obtain a Fourier transform result:

[0080]

[0081] Where S(ω,τ) is expressed in the time-frequency domain, t is time, ω is the angular frequency, τ is the time offset, ω(t-τ) is the window function, and j is the imaginary unit.

[0082] In the embodiment of the present invention, a local Fourier transform is performed on the intermediate audio signal y(t) by using a windowing function to obtain time-frequency features, thereby facilitating subsequent time-frequency point screening.

[0083] Step S4: performing an inverse short-time Fourier transform on the time-frequency points in the Fourier transform result whose energy values ​​are higher than a preset threshold value to construct a final audio signal;

[0084] In some embodiments, the step S4 comprises:

[0085] Step S41: Calculate the energy value of each time-frequency point of the Fourier transform result:

[0086] E(ω,τ)=|S(ω,τ)| 2 ;

[0087] In the formula, E(ω,τ) is the energy value corresponding to the time-frequency point;

[0088] Step S42: Arrange all calculated energy values ​​E(ω,τ) in the order of time and frequency to construct a time-frequency energy graph;

[0089] Step S43: Compare the energy value E(ω,τ) of each time-frequency point with a preset threshold, retain the time-frequency points with energy values ​​higher than the preset threshold, and record the time-frequency points higher than the preset threshold as S′(ω,τ);

[0090] Step S44: Perform inverse short-time Fourier transform on the time-frequency points in the Fourier transform result whose energy values ​​are higher than the preset threshold value to construct the final audio signal:

[0091]

[0092] Where z(t) is the final audio signal.

[0093] In an embodiment of the present invention, the energy value E(ω, τ) of each time-frequency point of the Fourier transform result can be obtained by calculating the square of the modulus value of the Fourier transform result, and then the calculated energy value E(ω, τ) is sorted in the order of time and frequency, so as to construct a time-frequency energy graph, thereby comparing the energy value E(ω, τ) of each time-frequency point with a preset threshold, retaining the time-frequency point S′(ω, τ) with an energy value higher than the preset threshold, and then performing an inverse short-time Fourier transform on the time-frequency point S′(ω, τ) with an energy value higher than the preset threshold to construct the final audio signal. The energy of the cavitation signal is relatively high. By removing the low-energy noise of the intermediate audio signal y(t), the main characteristics of the signal are retained, which can reduce the total amount of subsequent signal processing and improve processing efficiency.

[0094] Step S5: performing modal decomposition on the final audio signal to obtain a series of modal functions;

[0095] In some embodiments, the step S5 comprises:

[0096] Step S51: performing empirical mode decomposition on the final audio signal z(t), as shown in the following formula:

[0097]

[0098] Among them, IFM i (t) is the i-th mode function, r(t) is the residual term, and N is the number of modes.

[0099] In the embodiment of the present invention, a complex signal is decomposed into a series of simple modal functions through empirical mode decomposition (EMD), each function representing an inherent mode of the signal, which helps to further analyze and identify specific features in the final audio signal.

[0100] Step S6: Perform independent component analysis on the modal function and separate independent signals based on non-Gaussianity.

[0101] In some embodiments, step S6 includes:

[0102] Step S61: For the i-th mode function IFM i (t) is centralized to make its mean zero. The centralization is carried out using the following formula:

[0103] IFM centered (t) = IFM i (t)-E[IFM i (t)];

[0104] In the formula, IFM centered (t) is the modal function after centralization, E[IFM i (t)] is the expected value;

[0105] Step S62: whitening the centered mode function so that its covariance matrix becomes the unit matrix:

[0106] IFM whitened (t) = BIFM centered (t);

[0107] In the formula, IFM whitened (t) is the whitened mode function, and B is the whitening matrix;

[0108] Step S63: Using the Fast ICA algorithm to iteratively optimize the negative entropy to find the separation matrix W that maximizes the non-Gaussianity of the signal:

[0109]

[0110] In the formula, J(y) represents negative entropy. The larger the value of negative entropy J(y), the stronger the non-Gaussianity of signal y is. i ) is the separation signal y i The probability density function, q(y i ) is the probability density function of the standard normal distribution;

[0111] Step S64: Use the gradient descent method to maximize the negative entropy J(y) and iteratively update the separation matrix W;

[0112] Step S65: Use the estimated separation matrix W to separate the independent signals:

[0113] s(t)=WIFM whitened (t);

[0114] Where s(t) is an independent signal.

[0115] In an embodiment of the present invention, the independent signal s(t) can be separated by identifying the modal function with the highest non-Gaussianity. This can be achieved by comparing the negative entropy J(y) of each modal function, and selecting the modal function with the highest negative entropy J(y) as the independent signal s(t). After obtaining the independent signal s(t), to determine whether it is a cavitation signal, it can be done by analyzing the characteristics of these independent signals s(t). For example, a cavitation signal may have specific spectral characteristics, amplitude modulation characteristics, or kurtosis and skewness anomalies. By comparing these characteristics with known cavitation signal characteristics, it can be determined whether the separated independent signal s(t) is a cavitation signal. In summary, the audio signal generated by turbine cavitation can be quickly separated from the audio signal of normal operation, which is conducive to determining whether cavitation exists in the turbine.

[0116] Based on the foregoing embodiments, an embodiment of the present invention provides a fault detection device for a hydraulic turbine. The modules included in the device and the units included in each module can be implemented by a processor in a computer device; of course, they can also be implemented by a specific logic circuit; in the implementation process, the processor can be a central processing unit (CPU, Central Processing Unit), a microprocessor (MPU, Microprocessor Unit), a digital signal processor (DSP, Digital Signal Processing) or a field programmable gate array (FPGA, Field Programmable Gate Array), etc.

[0117] The embodiment of the present invention provides a device for separating audio signals of a water turbine. Figure 2 A schematic diagram of the structure of a hydraulic turbine audio signal separation device provided by an embodiment of the present invention is shown in FIG. Figure 2 As shown, including:

[0118] An acquisition module, used for acquiring the original audio signal of the turbine in operation;

[0119] A preprocessing module, used for preprocessing the original audio signal to obtain an intermediate audio signal;

[0120] In some embodiments, the preprocessing module comprises:

[0121] Assume that the original audio signal is x(t), and the original audio signal x(t) is denoised by a bandpass filter, as shown in the following formula:

[0122] y(t)=x(t)*h(t);

[0123] Where y(t) is the intermediate audio signal and h(t) is the impulse response of the bandpass filter.

[0124] A short-time Fourier transform module, used for performing a short-time Fourier transform on the intermediate audio signal to obtain a Fourier transform result;

[0125] In some embodiments, the short-time Fourier transform module includes:

[0126] Perform short-time Fourier transform on the intermediate audio signal y(t) to obtain a Fourier transform result:

[0127]

[0128] Where S(ω,τ) is expressed in the time-frequency domain, t is time, ω is the angular frequency, τ is the time offset, ω(t-τ) is the window function, and j is the imaginary unit.

[0129] An inverse short-time Fourier transform module is used to perform an inverse short-time Fourier transform on the time-frequency points whose energy values ​​in the Fourier transform results are higher than a preset threshold value to construct a final audio signal;

[0130] In some embodiments, the inverse short-time Fourier transform module includes:

[0131] Calculate the energy value of each time-frequency point of the Fourier transform result:

[0132] E(ω,τ)=|S(ω,τ)| 2 ;

[0133] In the formula, E(ω,τ) is the energy value corresponding to the time-frequency point;

[0134] Arrange all calculated energy values ​​E(ω,τ) in the order of time and frequency to construct a time-frequency energy graph;

[0135] Compare the energy value E(ω,τ) of each time-frequency point with the preset threshold, retain the time-frequency points with energy values ​​higher than the preset threshold, and record the time-frequency points higher than the preset threshold as S′(ω,τ);

[0136] Perform inverse short-time Fourier transform on the time-frequency points whose energy values ​​in the Fourier transform results are higher than the preset threshold to construct the final audio signal:

[0137]

[0138] Where z(t) is the final audio signal.

[0139] A modal decomposition module, used for performing modal decomposition on the final audio signal to obtain a series of modal functions;

[0140] In some embodiments, the modal decomposition module comprises:

[0141] The final audio signal z(t) is subjected to empirical mode decomposition, as shown in the following formula:

[0142]

[0143] Among them, IFM i (t) is the i-th mode function, r(t) is the residual term, and N is the number of modes.

[0144] The independent component analysis module is used to perform independent component analysis on the modal function and separate independent signals based on non-Gaussianity.

[0145] In some embodiments, the independent component analysis module includes:

[0146] For the i-th mode function IFM i (t) is centralized to make its mean zero. The centralization is carried out using the following formula:

[0147] IFM centered (t) = IFM i (t)-E[IFM i (t)];

[0148] In the formula, IFM centered (t) is the modal function after centralization, E[IFM i (t)] is the expected value;

[0149] The centered mode function is whitened so that its covariance matrix becomes the identity matrix:

[0150] IFM whitened (t) = BIFM centered (t);

[0151] In the formula, IFM whitened (t) is the whitened mode function, and B is the whitening matrix;

[0152] The Fast ICA algorithm is used to find the separation matrix W that maximizes the non-Gaussianity of the signal by iteratively optimizing the negative entropy:

[0153]

[0154] In the formula, J(y) represents negative entropy. The larger the value of negative entropy J(y), the stronger the non-Gaussianity of signal y is. i ) is the separation signal y i The probability density function, q(y i ) is the probability density function of the standard normal distribution;

[0155] Use the gradient descent method to maximize the negative entropy J(y) and iteratively update the separation matrix W;

[0156] Using the estimated separation matrix W, separate the independent signals:

[0157] s(t)=WIFM whitened (t);

[0158] Where s(t) is an independent signal.

[0159] It should be noted that, in the embodiment of the present invention, if the above separation method is implemented in the form of a software function module and sold or used as an independent product, it can also be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the embodiment of the present invention can be essentially or partly reflected in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium, including a number of instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the methods described in each embodiment of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as a U disk, a mobile hard disk, a read-only memory (ROM, Read Only Memory), a magnetic disk or an optical disk. In this way, the embodiment of the present invention is not limited to any specific combination of hardware and software.

[0160] Accordingly, an embodiment of the present invention provides a storage medium on which a computer program is stored, characterized in that when the computer program is executed by a processor, the steps in the separation method provided in the above embodiment are implemented.

[0161] An embodiment of the present invention provides an electronic device; Figure 3 A schematic diagram of the structure of an electronic device provided by an embodiment of the present invention is shown in FIG. Figure 3 As shown, the electronic device 400 includes: a processor 401, at least one communication bus 402, a user interface 403, at least one external communication interface 404, and a memory 405. The communication bus 402 is configured to realize the connection and communication between these components. The user interface 403 may include a display screen, and the external communication interface 404 may include a standard wired interface and a wireless interface. The processor 401 is configured to execute the program of the separation method stored in the memory to implement the steps in the separation method provided in the above embodiment.

[0162] It should be noted that the description of the above storage medium and electronic device embodiments is similar to the description of the above method embodiments, and has similar beneficial effects as the method embodiments. For technical details not disclosed in the storage medium and device embodiments of the present invention, please refer to the description of the method embodiments of the present invention for understanding.

[0163] It should be understood that "one embodiment" or "an embodiment" mentioned throughout the specification means that specific features, structures or characteristics related to the embodiment are included in at least one embodiment of the present invention. Therefore, "in one embodiment" or "in an embodiment" appearing throughout the specification does not necessarily refer to the same embodiment. In addition, these specific features, structures or characteristics can be combined in one or more embodiments in any suitable manner. It should be understood that in various embodiments of the present invention, the size of the serial number of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiment of the present invention. The serial numbers of the above-mentioned embodiments of the present invention are only for description and do not represent the advantages and disadvantages of the embodiments.

[0164] It should be noted that, in this article, the terms "include", "comprises" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, object or device including a series of elements includes not only those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such process, method, object or device. In the absence of further restrictions, an element defined by the sentence "comprises a ..." does not exclude the existence of other identical elements in the process, method, object or device including the element.

[0165] In the several embodiments provided by the present invention, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as: multiple units or components can be combined, or can be integrated into another system, or some features can be ignored or not executed. In addition, the coupling, direct coupling, or communication connection between the components shown or discussed can be through some interfaces, and the indirect coupling or communication connection of the devices or units can be electrical, mechanical or other forms.

[0166] The units described above as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units; they may be located in one place or distributed on multiple network units; some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0167] In addition, all functional units in the embodiments of the present invention may be integrated into one processing unit, or each unit may be separately used as a unit, or two or more units may be integrated into one unit; the above-mentioned integrated units may be implemented in the form of hardware or in the form of hardware plus software functional units.

[0168] A person of ordinary skill in the art can understand that: all or part of the steps of implementing the above-mentioned method embodiment can be completed by hardware related to program instructions, and the aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it executes the steps of the above-mentioned method embodiment; and the aforementioned storage medium includes: mobile storage devices, read-only memories (ROM, Read Only Memory), disks or optical disks, etc. Various media that can store program codes.

[0169] Alternatively, if the above-mentioned integrated unit of the present invention is implemented in the form of a software function module and sold or used as an independent product, it can also be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the embodiment of the present invention can be essentially or partly reflected in the form of a software product, which is stored in a storage medium and includes several instructions for a controller to execute all or part of the methods described in various embodiments of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as mobile storage devices, ROMs, magnetic disks or optical disks.

[0170] The above is only an embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. A method for separating audio signals of a water turbine, characterized in that: include: Collect the original audio signal of the turbine in operation; Preprocessing the original audio signal to obtain an intermediate audio signal; Performing a short-time Fourier transform on the intermediate audio signal to obtain a Fourier transform result; Perform inverse short-time Fourier transform on the time-frequency points whose energy values ​​in the Fourier transform results are higher than a preset threshold to construct the final audio signal; Performing modal decomposition on the final audio signal to obtain a series of modal functions; Independent component analysis is performed on the modal functions, and independent signals are separated based on non-Gaussianity.

2. A method for separating audio signals of a water turbine according to claim 1, characterized in that: The preprocessing of the original audio signal to obtain an intermediate audio signal includes: Assume that the original audio signal is x(t), and the original audio signal x(t) is denoised by a bandpass filter, as shown in the following formula: y(t)=x(t)*h(t); Where y(t) is the intermediate audio signal and h(t) is the impulse response of the bandpass filter.

3. A method for separating audio signals of a water turbine according to claim 2, characterized in that: The performing a short-time Fourier transform on the intermediate audio signal to obtain a Fourier transform result includes: Perform short-time Fourier transform on the intermediate audio signal y(t) to obtain a Fourier transform result: Where S(ω,τ) is expressed in the time-frequency domain, t is time, ω is the angular frequency, τ is the time offset, ω(t-τ) is the window function, and j is the imaginary unit.

4. A method for separating audio signals of a water turbine according to claim 3, characterized in that: The step of performing an inverse short-time Fourier transform on the time-frequency points in the Fourier transform result whose energy values ​​are higher than a preset threshold to construct a final audio signal includes: Calculate the energy value of each time-frequency point of the Fourier transform result: E(ω,τ)=|S(ω,τ)| 2 ; In the formula, E(ω,τ) is the energy value corresponding to the time-frequency point; Arrange all calculated energy values ​​E(ω,τ) in the order of time and frequency to construct a time-frequency energy graph; Compare the energy value E(ω,τ) of each time-frequency point with the preset threshold, retain the time-frequency points with energy values ​​higher than the preset threshold, and record the time-frequency points higher than the preset threshold as S′(ω,τ); Perform inverse short-time Fourier transform on the time-frequency points whose energy values ​​in the Fourier transform results are higher than the preset threshold to construct the final audio signal: Where z(t) is the final audio signal.

5. A method for separating audio signals of a water turbine according to claim 4, characterized in that: The final audio signal is subjected to modal decomposition to obtain a series of modal functions, including: The final audio signal z(t) is subjected to empirical mode decomposition, as shown in the following formula: Among them, IFM i (t) is the i-th mode function, r(t) is the residual term, and N is the number of modes.

6. A method for separating audio signals of a water turbine according to claim 5, characterized in that: The performing independent component analysis on the modal function and separating independent signals based on non-Gaussianity includes: For the i-th mode function IFM i (t) is centralized to make its mean zero. The centralization is carried out using the following formula: IFM centered (t)=IFM i (t)-E[IFM i (t)]; In the formula, IFM centered (t) is the modal function after centralization, E[IFM i (t)] is the expected value; The centered mode function is whitened so that its covariance matrix becomes the identity matrix: IFM whitened (t)=BIFM centered (t); In the formula, IFM whitened (t) is the whitened mode function, and B is the whitening matrix; The FastICA algorithm is used to find the separation matrix W that maximizes the non-Gaussianity of the signal by iteratively optimizing the negative entropy: In the formula, J(y) represents negative entropy. The larger the value of negative entropy J(y), the stronger the non-Gaussianity of signal y is. i ) is the separation signal y i The probability density function, q(y i ) is the probability density function of the standard normal distribution; Use the gradient descent method to maximize the negative entropy J(y) and iteratively update the separation matrix W; Using the estimated separation matrix W, separate the independent signals: s(t)=WIFM whitened (t); Where s(t) is an independent signal.

7. A water turbine audio signal separation device, characterized in that: include: An acquisition module, used for acquiring the original audio signal of the turbine in operation; A preprocessing module, used for preprocessing the original audio signal to obtain an intermediate audio signal; A short-time Fourier transform module, used for performing a short-time Fourier transform on the intermediate audio signal to obtain a Fourier transform result; An inverse short-time Fourier transform module is used to perform an inverse short-time Fourier transform on the time-frequency points whose energy values ​​in the Fourier transform results are higher than a preset threshold value to construct a final audio signal; A modal decomposition module, used for performing modal decomposition on the final audio signal to obtain a series of modal functions; The independent component analysis module is used to perform independent component analysis on the modal function and separate independent signals based on non-Gaussianity.

8. An electronic device, characterized in that: include: A memory and a processor, wherein a computer program is stored in the memory, and when the computer program is executed by the processor, the method according to any one of claims 1 to 6 is executed.

9. A storage medium, characterized in that: The computer program stored in the storage medium can be executed by one or more processors and can be used to implement the method as claimed in any one of claims 1 to 6.

Images