Power transformer sound signal blind source separation method based on adaptive threshold repeated extraction mode

Abstract

The invention provides a blind source separation method for a sound signal of a power transformer based on an adaptive threshold repeated extraction mode. The method comprises the following steps: determining a repetition period of a mixed signal; the mixed signal comprises a power transformer sound signal and an environment noise signal; dividing the mixed signal into a plurality of mixed signal units according to the repetition period of the mixed signal, and determining a blind source separation adaptive threshold corresponding to each mixed signal unit; and separating a power transformer sound signal from the mixed signal according to the blind source separation adaptive threshold value corresponding to each mixed signal unit. According to the method provided by the embodiment of the invention, effective separation of the operation sound signals of the transformer is realized, and the signal separation precision in a noisy environment is remarkably improved.

Description

Technical Field

[0001] The present invention relates to the technical field of electric power, and in particular, to a method for blind source separation of sound signals of a power transformer based on an adaptive threshold repeated extraction mode. Background Art

[0002] The subway power transformer is a key device to ensure the stable operation of the subway system, responsible for converting high-voltage electrical energy into a voltage level suitable for the subway system. Due to its importance, it is crucial to monitor and diagnose the transformer in real time. Voiceprint recognition is an effective monitoring means. By analyzing the sound signals emitted by the transformer, the sound characteristics under different fault states can be detected.

[0003] In the related art, since the sound signals of the transformer are often mixed with environmental noise, it seriously affects the accuracy of fault recognition based on voiceprint. Therefore, how to accurately separate the pure sound signals of the power transformer from the mixed signals to improve the diagnostic efficiency and accuracy of the power transformer is a technical problem that those skilled in the art urgently need to solve. Summary of the Invention

[0004] The present invention provides a method for blind source separation of sound signals of a power transformer based on an adaptive threshold repeated extraction mode, which realizes the effective separation of the running sound signals of the transformer and significantly improves the signal separation accuracy in a noisy environment.

[0005] The present invention provides a method for blind source separation of sound signals of a power transformer based on an adaptive threshold repeated extraction mode, including the following steps.

[0006] Determine the repetition period of the mixed signal; the mixed signal includes the sound signal of the power transformer and the environmental noise signal; According to the repetition period of the mixed signal, divide the mixed signal into multiple mixed signal units, and determine the blind source separation adaptive threshold corresponding to each mixed signal unit; According to the blind source separation adaptive threshold corresponding to each mixed signal unit, separate the sound signal of the power transformer from the mixed signal.

[0007] According to the method for blind source separation of sound signals of a power transformer based on an adaptive threshold repeated extraction mode provided by the present invention, the determination of the blind source separation adaptive threshold corresponding to each mixed signal unit includes: Determine the energy information of each mixed signal unit; According to the energy information of each mixed signal unit, determine the blind source separation adaptive threshold corresponding to each mixed signal unit.

[0008] A method for blind source separation of power transformer sound signals based on an adaptive threshold repeated extraction pattern provided by the present invention, said determining the energy information of each of the mixed signal units; determining the blind source separation adaptive threshold corresponding to each of the mixed signal units according to the energy information of each of the mixed signal units, including: Determining the energy mean and energy standard deviation of each of the mixed signal units; Taking the arithmetic value between the energy mean of each of the mixed signal units and the energy standard deviation as the blind source separation adaptive threshold corresponding to each of the mixed signal units.

[0009] A method for blind source separation of power transformer sound signals based on an adaptive threshold repeated extraction pattern provided by the present invention, the arithmetic value between the energy mean and the energy standard deviation of each of the mixed signal units, including: A + 2×B; wherein, A represents the energy mean of each of the mixed signal units; B represents the energy standard deviation of each of the mixed signal units.

[0010] A method for blind source separation of power transformer sound signals based on an adaptive threshold repeated extraction pattern provided by the present invention, said separating the power transformer sound signals from the mixed signals according to the blind source separation adaptive threshold corresponding to each of the mixed signal units, including: Based on the soft time-frequency mask algorithm and the blind source separation adaptive threshold corresponding to each of the mixed signal units, separating the power transformer sound signals from the mixed signals.

[0011] A method for blind source separation of power transformer sound signals based on an adaptive threshold repeated extraction pattern provided by the present invention, after separating the power transformer sound signals from the mixed signals according to the blind source separation adaptive threshold corresponding to each of the mixed signal units, the method further includes: Taking the difference between the mixed signals and the power transformer sound signals as the environmental noise signals in the mixed signals.

[0012] The present invention also provides a blind source separation device for power transformer sound signals based on an adaptive threshold repeated extraction pattern, including the following modules: A first determination module, configured to determine the repetition period of the mixed signals; the mixed signals include power transformer sound signals and environmental noise signals; A second determination module, according to the repetition period of the mixed signals, dividing the mixed signals into a plurality of mixed signal units, and determining the blind source separation adaptive threshold corresponding to each of the mixed signal units; A separation module, configured to separate the sound signal of the power transformer from the mixed signals according to the blind source separation adaptive thresholds corresponding to the respective mixed signal units.

[0013] The present invention also provides an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the program, the method for blind source separation of the sound signal of the power transformer based on the adaptive threshold repeated extraction mode as described in any one of the above is implemented.

[0014] The present invention also provides a non-transitory computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the method for blind source separation of the sound signal of the power transformer based on the adaptive threshold repeated extraction mode as described in any one of the above is implemented.

[0015] The present invention also provides a computer program product, including a computer program. When the computer program is executed by a processor, the method for blind source separation of the sound signal of the power transformer based on the adaptive threshold repeated extraction mode as described in any one of the above is implemented.

[0016] The method for blind source separation of the sound signal of the power transformer based on the adaptive threshold repeated extraction mode provided by the present invention effectively separates the operation sound signal of the transformer by utilizing the periodic repetition pattern in the mixed signals and determining the dynamic blind source separation threshold, significantly improving the signal separation accuracy in a noisy environment. Compared with the traditional fixed threshold method, it can better cope with noises of different intensities and types, improving the accuracy and robustness of blind source separation. Description of the Drawings

[0017] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0018] Figure 1 It is one of the flow diagrams of the method for blind source separation of the sound signal of the power transformer based on the adaptive threshold repeated extraction mode provided by the present invention.

[0019] Figure 2 It is another flow diagram of the method for blind source separation of the sound signal of the power transformer based on the adaptive threshold repeated extraction mode provided by the present invention.

[0020] Figure 3 It is a schematic diagram of the device for blind source separation of the sound signal of the power transformer based on the adaptive threshold repeated extraction mode provided by the present invention.

[0021] Figure 4 It is a schematic structural diagram of the electronic device provided by the present invention. Specific embodiments

[0022] To make the objectives, technical solutions and advantages of the present invention clearer, the technical solutions in the present invention will be clearly and completely described below with reference to the accompanying drawings in the present invention. Obviously, the described embodiments are some but not all of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in the present invention without making creative efforts shall fall within the protection scope of the present invention.

[0023] The following will be combined with Figures 1-4 Describe the blind source separation method for the sound signal of the power transformer based on the adaptive threshold repeated extraction mode of the present invention.

[0024] To facilitate a clearer understanding of the technical solutions of the embodiments of the present application, some technical contents related to the embodiments of the present application will be introduced first.

[0025] The subway power transformer is a key device to ensure the stable operation of the subway system. It is responsible for converting high-voltage electrical energy into a voltage level suitable for the subway system. Due to its importance, it is crucial to monitor and diagnose the transformer in real time. Voiceprint recognition is an effective monitoring means. By analyzing the sound signal emitted by the transformer, the sound characteristics in different fault states can be detected.

[0026] In the actual environment, the sound signal of the transformer is often mixed with environmental noise, which seriously affects the accuracy of fault recognition based on voiceprint. Therefore, separating the pure fault voiceprint from the mixed signal has become a key issue. Traditional blind source separation algorithms such as independent component analysis (ICA) have limitations in processing such signals, especially when dealing with signals with periodic repetition patterns. For example, the prior art has the following disadvantages when dealing with signals containing periodic repetition patterns: (1) Limitation of fixed threshold: Traditional algorithms usually rely on a preset threshold to distinguish background and foreground signals. This method lacks flexibility and has poor separation effect when facing noises with different intensities and types.

[0027] (2) Poor adaptability to complex noise environments: When there are many types of interfering sounds in the environment and they change randomly, traditional methods are difficult to effectively separate the target signal.

[0028] (3) Failure to fully utilize the periodic structure: For signals containing obvious periodic structures, the prior art has not made good use of these characteristics for more accurate separation.

[0029] Figure 1It is a schematic flowchart of the blind source separation method for power transformer sound signals based on the adaptive threshold repeated extraction mode provided by the present invention. The method includes the following: Step 101: Determine the repetition period of the mixed signal; the mixed signal includes the power transformer sound signal and the environmental noise signal.

[0030] Specifically, the sound signal during the operation of the transformer is periodically repeated. Traditional blind source separation algorithms such as independent component analysis (ICA) have limitations when dealing with such signals with periodic repetition patterns and the effect is not good.

[0031] To solve the above problems, in the embodiments of the present application, the repetition period of the mixed signal is first determined. Among them, the mixed signal is a mixed signal containing the power transformer sound signal and the environmental noise signal. Optionally, the short-time Fourier transform (STFT) can be performed on the mixed signal, and the repetition period of the repeated signal can be determined by analyzing the peaks in the autocorrelation matrix.

[0032] Step 102: According to the repetition period of the mixed signal, divide the mixed signal into multiple mixed signal units, and determine the blind source separation adaptive threshold corresponding to each mixed signal unit.

[0033] Specifically, after determining the repetition period of the mixed signal, in the embodiments of the present application, the mixed signal is divided into multiple mixed signal units according to the repetition period of the mixed signal.

[0034] For example, based on the identified repetition period of the mixed signal, the amplitude spectrogram of the mixed signal is divided into multiple segments, each segment corresponding to a repetition period, that is, corresponding to a mixed signal unit.

[0035] Optionally, after determining multiple mixed signal units according to the repetition period of the mixed signal, in the embodiments of the present application, the blind source separation adaptive threshold corresponding to each mixed signal unit is further determined to separate the mixed signal, thereby effectively avoiding the problems of poor flexibility and poor separation effect existing in the existing algorithms when separating signals depending on the preset threshold.

[0036] Step 103: Separate the power transformer sound signal from the mixed signal according to the blind source separation adaptive threshold corresponding to each mixed signal unit.

[0037] Specifically, after determining the blind source separation adaptive threshold corresponding to each mixed signal unit, the power transformer sound signal can be separated from the mixed signal according to the blind source separation adaptive threshold corresponding to each mixed signal unit, realizing the effective separation of the transformer operation sound signal. Compared with the traditional fixed threshold method, it can better cope with different intensities and types of noise, and improve the accuracy and robustness of blind source separation.

[0038] The method of the above embodiment effectively separates the sound signals of transformer operation by utilizing the periodic repetition patterns in the mixed signals and determining the dynamic blind source separation threshold, significantly improving the signal separation accuracy in a noisy environment. Compared with the traditional fixed threshold method, it can better cope with noises of different intensities and types, and improve the accuracy and robustness of blind source separation.

[0039] In one embodiment, determining the blind source separation adaptive threshold corresponding to each mixed signal unit includes: Determining the energy information of each mixed signal unit; According to the energy information of each mixed signal unit, determining the blind source separation adaptive threshold corresponding to each mixed signal unit.

[0040] Specifically, in the embodiment of the present application, the blind source separation adaptive threshold corresponding to each mixed signal unit is determined based on the energy information of each mixed signal unit, that is, the optimal threshold is dynamically determined based on the energy statistical characteristics of the mixed signals within each repetition period, realizing the dynamic adjustment of the blind source separation threshold based on the real-time changes of the mixed signals, and can effectively improve the accuracy and robustness of blind source separation. Optionally, the energy information of the mixed signal unit includes the mean, variance, etc. of the amplitude spectrum of the mixed signal within each repetition period.

[0041] For example, after dividing the mixed signal into multiple mixed signal units according to the repetition period of the mixed signal, that is, after dividing the amplitude spectrum diagram of the mixed signal into multiple segments based on the identified repetition period of the mixed signal, statistical feature analysis can be performed on each segment, such as the mean and variance of the amplitude spectrum of the mixed signal within each repetition period, constructing a model representing the statistical characteristics of the repetitive structure, and obtaining the energy information of each mixed signal unit.

[0042] The method of the above embodiment dynamically determines the optimal threshold according to the energy statistical characteristics of the mixed signals within each repetition period, achieving the effect of dynamically determining and adjusting the blind source separation threshold based on the real-time changes of the mixed signals, and can effectively improve the accuracy and robustness of blind source separation.

[0043] In one embodiment, determining the energy information of each mixed signal unit; according to the energy information of each mixed signal unit, determining the blind source separation adaptive threshold corresponding to each mixed signal unit includes: Determining the energy mean and energy standard deviation of each mixed signal unit; Taking the arithmetic value between the energy mean and the energy standard deviation of each mixed signal unit as the blind source separation adaptive threshold corresponding to each mixed signal unit.

[0044] Specifically, in the embodiments of the present application, the arithmetic value between the energy mean and the energy standard deviation of each mixed signal unit is used as the blind source separation adaptive threshold corresponding to each mixed signal unit. That is, in the process of dynamically determining the optimal threshold according to the energy statistical characteristics of the mixed signal in each repetition period, the data statistical characteristics of each local region in the mixed signal are fully considered, so that the finally determined blind source separation adaptive threshold corresponding to each mixed signal unit is more accurate and reasonable, effectively improving the accuracy and robustness of blind source separation.

[0045] In one embodiment, the arithmetic value between the energy mean and the energy standard deviation of each mixed signal unit includes: A + 2×B; where A represents the energy mean of each mixed signal unit; B represents the energy standard deviation of each mixed signal unit.

[0046] Specifically, in the embodiments of the present application, by calculating the average energy and its standard deviation corresponding to each mixed signal unit, and then determining the adaptive dynamic threshold of blind source separation corresponding to each mixed signal unit, the effect of accurately determining the blind source separation adaptive threshold according to the statistical characteristics of local energy is achieved. Optionally, the threshold can be based on "average energy + 2×energy standard deviation" of the mixed signal unit to ensure that the threshold can be dynamically adjusted with the change of the signal, so as to finally accurately obtain the sound signal of the power transformer separated from the mixed signal.

[0047] For the method of the above embodiment, when there are many types of interfering sounds in the environment and they change randomly, based on "average energy + 2×energy standard deviation" of the mixed signal unit as the threshold, it is possible to effectively separate and reduce the noise of the transformer operation sound signal, improve the accuracy of voiceprint recognition, and provide a more reliable data basis for transformer fault diagnosis based on sound signals.

[0048] In one embodiment, separating the sound signal of the power transformer from the mixed signal according to the blind source separation adaptive threshold corresponding to each mixed signal unit includes: Based on the soft time-frequency mask algorithm and the blind source separation adaptive threshold corresponding to each mixed signal unit, separating the sound signal of the power transformer from the mixed signal.

[0049] Specifically, after determining the blind source separation adaptive threshold corresponding to each mixed signal unit, in the embodiments of the present application, blind source separation is performed based on the soft time-frequency mask algorithm and the blind source separation adaptive threshold corresponding to each mixed signal unit, which can smooth the transition region, avoid the hard boundary effect, and suppress non-repetitive foreground content. Even under low signal-to-noise ratio conditions, it can effectively separate the sound signal of the transformer and achieve accurate and efficient blind source separation.

[0050] The method of the above embodiment combines multiple steps such as the recognition of the periodic structure of the mixed signal, the setting of the adaptive threshold, and the generation of the soft time-frequency mask, forming a complete blind source separation scheme for the mixed signal, which can accurately and efficiently implement the voiceprint recognition task in complex environments such as power systems.

[0051] In one embodiment, after separating the sound signal of the power transformer from the mixed signal according to the blind source separation adaptive threshold corresponding to each mixed signal unit, the method further includes: Determine the difference between the mixed signal and the sound signal of the power transformer as the environmental noise signal in the mixed signal.

[0052] Specifically, after separating the sound signal of the power transformer from the mixed signal according to the blind source separation adaptive threshold corresponding to each mixed signal unit, in the embodiment of the present application, the difference between the mixed signal and the sound signal of the power transformer is determined as the environmental noise signal in the mixed signal, thus accurately realizing the separation of the foreground signal and the background signal.

[0053] Exemplarily, as Figure 2 shown, the embodiment of the present application also provides a blind source separation method for the sound signal of a power transformer based on an adaptive threshold repeated extraction mode, which is mainly divided into four steps: repeated period recognition, repeated segment modeling, and repeated pattern extraction.

[0054] (1) Repeated period recognition.

[0055] First, perform a short-time Fourier transform (STFT) on the input mixed signal, identify the repeated period in the signal through the autocorrelation matrix, and determine the repeated period of the signal by analyzing the peaks in the autocorrelation matrix.

[0056] (2) Repeated segment modeling.

[0057] Based on the identified repeated period, divide the amplitude spectrogram of the mixed signal into multiple segments, each segment corresponding to a repeated period. Perform statistical feature analysis on each segment, such as mean, variance, etc., to construct a model representing the statistical features of the repeated structure.

[0058] (3) Repeated pattern extraction.

[0059] Calculate the local energy: For each time-frequency unit, calculate the energy of its local area.

[0060] Calculate the average energy and standard deviation: Calculate the average energy and its standard deviation of the entire signal or local area.

[0061] Calculate the adaptive dynamic threshold: Set the adaptive threshold according to the statistical characteristics of the local energy.

[0062] For example, "average energy + two standard deviations" can be used as the threshold to ensure that the threshold can be dynamically adjusted according to the changes in the signal.

[0063] Generate a soft time-frequency mask for distinguishing foreground and background signals, and apply adaptive thresholding to finally obtain a clear target signal. That is, use the time-frequency mask technology to extract the background part from the mixed signal and suppress the non-repetitive foreground content.

[0064] In the embodiments of the present application, an adaptive threshold calculation mechanism is introduced to dynamically determine the optimal threshold according to the statistical characteristics of the data in the local area, effectively enhancing the adaptability and robustness to different signal statistical characteristics, reducing the false alarm rate, and improving the reliability and accuracy of the fault diagnosis system. In addition, this method is not only applicable to the monitoring of power transformers in the subway system, but can also be widely applied to the health condition monitoring of other electrical equipment. For example, in the fields of industrial production, power transmission and distribution, etc. The present application provides a new way to improve the health condition monitoring technology of electrical equipment based on acoustic fingerprints, provides a more accurate and reliable means for real-time monitoring and fault warning, helps to detect potential faults in advance and take corresponding measures, thereby avoiding the occurrence of major accidents and ensuring the safe and stable operation of the power system. In short, the present application not only achieves a breakthrough in technology, but also shows broad prospects and important practical value in practical applications.

[0065] The following describes the power transformer sound signal blind source separation device based on the adaptive threshold repeated extraction mode provided by the present invention. The power transformer sound signal blind source separation device based on the adaptive threshold repeated extraction mode described below can be correspondingly referred to the power transformer sound signal blind source separation method based on the adaptive threshold repeated extraction mode described above. The power transformer sound signal blind source separation device based on the adaptive threshold repeated extraction mode in the embodiments of the present application is as Figure 3 shown and includes: A first determination module 310, configured to determine the repetition period of the mixed signal; the mixed signal includes a power transformer sound signal and an environmental noise signal; A second determination module 320, configured to divide the mixed signal into a plurality of mixed signal units according to the repetition period of the mixed signal, and determine the blind source separation adaptive threshold corresponding to each mixed signal unit; A separation module 330, configured to separate the power transformer sound signal from the mixed signal according to the blind source separation adaptive threshold corresponding to each mixed signal unit.

[0066] Optionally, the second determination module 320 is specifically configured to: Determine the energy information of each mixed signal unit; Determine the blind source separation adaptive threshold corresponding to each mixed signal unit according to the energy information of each mixed signal unit.

[0067] Optionally, the second determination module 320 is specifically configured to: Determine the energy mean value and energy standard deviation of each mixed signal unit; Determine the arithmetic value between the energy mean value and the energy standard deviation of each mixed signal unit as the blind source separation adaptive threshold corresponding to each mixed signal unit.

[0068] Optionally, the arithmetic value between the energy mean value and the energy standard deviation of each mixed signal unit includes: Wherein, A represents the energy mean value of each mixed signal unit; B represents the energy standard deviation of each mixed signal unit.

[0069] Optionally, the separation module 330 is specifically configured to: Based on the soft time-frequency mask algorithm and the blind source separation adaptive threshold corresponding to each mixed signal unit, separate the power transformer sound signal from the mixed signal.

[0070] Optionally, the separation module 330 is further configured to: Determine the difference between the mixed signal and the power transformer sound signal as the environmental noise signal in the mixed signal.

[0071] Figure 4 The schematic diagram of the physical structure of an electronic device is exemplified. The electronic device may include: a processor 410, a communication interface 420, a memory 430, and a communication bus 440. Among them, the processor 410, the communication interface 420, and the memory 430 complete mutual communication through the communication bus 440. The processor 410 can call the logical instructions in the memory 430 to execute the blind source separation method of the power transformer sound signal based on the adaptive threshold repeated extraction mode. The method includes: determining the repetition period of the mixed signal; the mixed signal includes a power transformer sound signal and an environmental noise signal; according to the repetition period of the mixed signal, dividing the mixed signal into multiple mixed signal units, determining the blind source separation adaptive threshold corresponding to each mixed signal unit; separating the power transformer sound signal from the mixed signal according to the blind source separation adaptive threshold corresponding to each mixed signal unit.

[0072] In addition, when the logical instructions in the above-mentioned memory 430 are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of the present invention. The aforementioned storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memories (ROM, Read-Only Memory), random access memories (RAM, Random Access Memory), magnetic disks, or optical discs that can store program codes.

[0073] On the other hand, the present invention also provides a computer program product. The computer program product includes a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the blind source separation method for the sound signal of a power transformer based on the adaptive threshold repeated extraction mode provided by the above-mentioned various methods. The method includes: determining the repetition period of the mixed signal; the mixed signal includes the sound signal of the power transformer and the environmental noise signal; according to the repetition period of the mixed signal, dividing the mixed signal into multiple mixed signal units, and determining the blind source separation adaptive threshold corresponding to each mixed signal unit; separating the sound signal of the power transformer from the mixed signal according to the blind source separation adaptive threshold corresponding to each mixed signal unit.

[0074] On another aspect, the present invention also provides a non-transitory computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, it is implemented to execute the blind source separation method for the sound signal of a power transformer based on the adaptive threshold repeated extraction mode provided by the above-mentioned various methods. The method includes: determining the repetition period of the mixed signal; the mixed signal includes the sound signal of the power transformer and the environmental noise signal; according to the repetition period of the mixed signal, dividing the mixed signal into multiple mixed signal units, and determining the blind source separation adaptive threshold corresponding to each mixed signal unit; separating the sound signal of the power transformer from the mixed signal according to the blind source separation adaptive threshold corresponding to each mixed signal unit.

[0075] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment. Those of ordinary skill in the art can understand and implement it without creative efforts.

[0076] Through the description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on such an understanding, the essence of the above technical solution, or the part that contributes to the prior art, can be embodied in the form of a software product. The computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions for causing a computer device (which can be a personal computer, server, or network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0077] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements on some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A blind source separation method for the sound signal of a power transformer based on an adaptive threshold repeated extraction pattern, characterized in that, Including: Determine the repetition period of the mixed signal; the mixed signal includes a power transformer sound signal and an environmental noise signal; According to the repetition period of the mixed signal, divide the mixed signal into multiple mixed signal units, and determine the blind source separation adaptive threshold corresponding to each mixed signal unit; Separate the power transformer sound signal from the mixed signal according to the blind source separation adaptive threshold corresponding to each mixed signal unit.

2. The blind source separation method of the power transformer sound signal based on the adaptive threshold repeated extraction pattern according to claim 1, characterized in that The determining the blind source separation adaptive threshold corresponding to each mixed signal unit includes: Determine the energy information of each mixed signal unit; Determine the blind source separation adaptive threshold corresponding to each mixed signal unit according to the energy information of each mixed signal unit.

3. The blind source separation method of the power transformer sound signal based on the adaptive threshold repeated extraction pattern according to claim 1, wherein The determining the energy information of each mixed signal unit; The determining the blind source separation adaptive threshold corresponding to each mixed signal unit according to the energy information of each mixed signal unit includes: Determine the energy mean value and energy standard deviation of each mixed signal unit; Determine the arithmetic value between the energy mean value and the energy standard deviation of each mixed signal unit as the blind source separation adaptive threshold corresponding to each mixed signal unit.

4. The blind source separation method of the power transformer sound signal based on the adaptive threshold repeated extraction pattern according to claim 1, characterized in that The arithmetic value between the energy mean value and the energy standard deviation of each mixed signal unit includes: Where A represents the energy mean value of each mixed signal unit; B represents the energy standard deviation of each mixed signal unit.

5. The blind source separation method of the power transformer sound signal based on the adaptive threshold repeated extraction pattern according to claim 1, wherein The separating the power transformer sound signal from the mixed signal according to the blind source separation adaptive threshold corresponding to each mixed signal unit includes: Based on the soft time-frequency mask algorithm and the blind source separation adaptive threshold corresponding to each mixed signal unit, separate the power transformer sound signal from the mixed signal.

6. The blind source separation method for power transformer sound signals based on the repeated extraction pattern of adaptive threshold according to claim 1, characterized in that After separating the power transformer sound signal from the mixed signal according to the blind source separation adaptive threshold corresponding to each mixed signal unit, the method further includes: Determine the difference between the mixed signal and the power transformer sound signal as the environmental noise signal in the mixed signal.

7. A blind source separation device for power transformer sound signals based on an adaptive threshold repeated extraction pattern, characterized in that, Including: A first determination module for determining the repetition period of the mixed signal; the mixed signal includes a power transformer sound signal and an environmental noise signal; A second determination module for dividing the mixed signal into multiple mixed signal units according to the repetition period of the mixed signal and determining the blind source separation adaptive threshold corresponding to each mixed signal unit; A separation module for separating the power transformer sound signal from the mixed signal according to the blind source separation adaptive threshold corresponding to each mixed signal unit.

8. An electronic device, comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the blind source separation method of the power transformer sound signal in the adaptive threshold repeated extraction mode as described in any one of claims 1 to 6.

9. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the blind source separation method of the power transformer sound signal in the adaptive threshold repeated extraction mode as described in any one of claims 1 to 6.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by the processor, it implements the blind source separation method of the power transformer sound signal in the adaptive threshold repeated extraction mode as described in any one of claims 1 to 6.

Images