Intelligent monitoring method for rotating machine equipment of thermal power plant based on multi-modal noise analysis
Through multimodal noise analysis and neural network models, the problems of high sensor requirements and difficulty in identifying aerodynamic faults in traditional monitoring methods are solved, and high-precision fault identification and early damage detection of thermal power plant transfer equipment are achieved.
Patent Information
- Application Number
- CN202510616113.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-05-14
- Publication Date
- 2025-10-03
AI Technical Summary
Traditional monitoring methods rely on vibration signal analysis, have high requirements for the installation of sensors on equipment, and are difficult to capture aerodynamic faults. Traditional vibration monitoring has low sensitivity to aerodynamic faults such as blade cavitation and duct surge, and cannot detect early microscopic damage. In addition, it is difficult to distinguish the root cause of the fault when mechanical vibration and aerodynamic noise are coupled.
Multimodal noise analysis is adopted to obtain the acoustic signals and operating parameters of the equipment, perform noise reduction processing in the spatial and frequency domains, extract time domain, frequency domain, nonlinear features and operating condition characteristics, combine the minimum redundancy and maximum correlation algorithm to screen key features, and use the neural network model to perform feature fusion and identification to output equipment warning information.
It improves the aerodynamic noise removal effect, realizes dynamic identification of equipment faults according to changes in the equipment operating environment, improves the accuracy of aerodynamic fault identification, and reduces sensor installation requirements and maintenance costs.
Smart Images
Figure CN120748436A_ABST
Abstract
Description
Technical Field
[0001] The present disclosure relates to the technical field of equipment monitoring, and in particular to an intelligent monitoring method for transfer equipment in a thermal power plant based on multimodal noise analysis. Background Art
[0002] In related technologies, traditional monitoring methods rely on vibration signal analysis, have high requirements for the installation of sensors on equipment, and are difficult to capture aerodynamic faults. Traditional vibration monitoring has low sensitivity to aerodynamic faults such as blade cavitation and duct surge, and cannot detect early microscopic damage. In addition, when mechanical vibration and aerodynamic noise are coupled, such as when bearing wear and blade cracks exist at the same time, the same monitoring method is difficult to distinguish the root cause of the equipment failure. Summary of the Invention
[0003] In order to overcome the problems existing in the related art, the present disclosure provides an intelligent monitoring method for thermal power plant transfer equipment based on multimodal noise analysis.
[0004] According to a first aspect of an embodiment of the present disclosure, a method for intelligently monitoring transfer equipment in a thermal power plant based on multimodal noise analysis is provided, comprising:
[0005] Obtain the acoustic signals and operating parameters of the equipment to be monitored;
[0006] Performing spatial domain noise reduction processing on the acoustic signal to obtain an acoustic signal after spatial domain noise reduction;
[0007] Performing frequency domain noise reduction processing on the acoustic signal after spatial domain noise reduction using the operating condition parameters to obtain an acoustic signal after noise separation;
[0008] Extracting time domain features, frequency domain features and nonlinear features from the acoustic signal after noise separation, and extracting operating condition features from the operating condition parameters;
[0009] Performing feature fusion on the time domain features, frequency domain features, nonlinear features and operating condition features to obtain a multi-dimensional feature matrix;
[0010] The multidimensional feature matrix is input into a trained neural network model to obtain equipment warning information output after the neural network model identifies the multidimensional feature matrix.
[0011] In some embodiments of the present disclosure, there are multiple groups of acoustic signals, and each group of acoustic signals is collected by a sensor;
[0012] The method of performing frequency domain noise reduction processing on the acoustic signal after spatial domain noise reduction by using the operating condition parameter to obtain the acoustic signal after noise separation includes:
[0013] Determine the signal-to-noise ratio of each sensor separately;
[0014] For each sensor, the acoustic signal after spatial domain noise reduction is weighted by using the signal-to-noise ratio of the sensor to obtain an intermediate acoustic signal.
[0015] The operating condition parameters are used to perform frequency domain noise reduction processing on the intermediate acoustic signal to obtain an acoustic signal after noise separation.
[0016] In some embodiments of the present disclosure, weighted processing is performed on the acoustic signal after spatial domain noise reduction using the signal-to-noise ratio of the sensor to obtain an intermediate acoustic signal, including:
[0017] Amplifying the signal-to-noise ratios of the sensors to obtain amplified signal-to-noise ratios;
[0018] Normalizing the signal-to-noise ratio after the amplification process to obtain a dynamic weight of the sensor;
[0019] The dynamic weight is used to perform weighted processing on the acoustic signal after the spatial domain noise reduction to obtain an intermediate acoustic signal.
[0020] In some embodiments of the present disclosure, the operating condition parameter includes a device speed; and performing frequency domain noise reduction processing on the intermediate acoustic signal using the operating condition parameter to obtain a noise-separated acoustic signal includes:
[0021] The intermediate acoustic signal is subjected to frequency domain noise reduction processing using the following filter transfer function H(z) to obtain the acoustic signal after noise separation:
[0022]
[0023] Among them, f s is the signal sampling frequency, n is the device speed, and D is the number of sampling points corresponding to one revolution of the device.
[0024] In some embodiments of the present disclosure, performing spatial domain noise reduction processing on the acoustic signal to obtain the acoustic signal after spatial domain noise reduction includes:
[0025] The acoustic signal is subjected to spatial domain noise reduction processing using the following formula:
[0026]
[0027] Among them, Y(f) is the frequency domain signal after noise reduction, W m (f) is the dynamic weight of the mth sensor, which is calculated based on the signal-to-noise ratio, X m (f) is the acoustic signal collected by the mth sensor, τ mis the time delay, which is calculated from the geometric distance from the acoustic signal source to the sensor and the speed of sound. is the phase compensation term.
[0028] In some embodiments of the present disclosure, the dynamic weight of the mth sensor is calculated using the following formula:
[0029]
[0030] Among them, SNR m (f) is the signal-to-noise ratio of the mth sensor.
[0031] In some embodiments of the present disclosure, before fusing the time domain features, frequency domain features, nonlinear features, and operating condition features to obtain a multidimensional feature matrix, the method further includes:
[0032] Using a minimum redundancy maximum correlation algorithm to screen key features from the time domain features, frequency domain features, nonlinear features, and operating condition features, to obtain screened time domain features, frequency domain features, nonlinear features, and operating condition features;
[0033] The feature fusion of the time domain features, frequency domain features, nonlinear features and operating condition features is performed to obtain a multidimensional feature matrix, including:
[0034] The filtered time domain features, frequency domain features, nonlinear features and operating condition features are fused to obtain a multi-dimensional feature matrix.
[0035] According to a second aspect of an embodiment of the present disclosure, there is provided an intelligent monitoring device for thermal power plant transfer equipment based on multimodal noise analysis, comprising:
[0036] An acquisition unit, used to acquire acoustic signals and operating parameters of the equipment to be monitored;
[0037] A spatial domain noise reduction unit, configured to perform spatial domain noise reduction processing on the acoustic signal to obtain an acoustic signal after spatial domain noise reduction;
[0038] a frequency domain noise reduction unit, configured to perform frequency domain noise reduction processing on the acoustic signal after the spatial domain noise reduction using the operating condition parameters to obtain an acoustic signal after noise separation;
[0039] an extraction unit, configured to extract time domain features, frequency domain features, and nonlinear features from the acoustic signal after noise separation, and to extract operating condition features from the operating condition parameters;
[0040] A fusion unit, configured to fuse the time domain features, frequency domain features, nonlinear features, and operating condition features to obtain a multidimensional feature matrix;
[0041] The early warning unit is used to input the multidimensional feature matrix into the trained neural network model to obtain the equipment early warning information output after the neural network model identifies the multidimensional feature matrix.
[0042] According to a third aspect of an embodiment of the present disclosure, an electronic device includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the method described in any one of the first aspects is implemented.
[0043] According to a fourth aspect of an embodiment of the present disclosure, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the method according to any one of the first aspects is implemented.
[0044] According to a fifth aspect of an embodiment of the present disclosure, a computer program product is provided, comprising a computer program, wherein the computer program implements the method as described in any one of the first aspects when executed by a processor.
[0045] The technical solution provided by the embodiments of the present disclosure may include the following beneficial effects: obtaining the acoustic signal and operating condition parameters of the equipment to be monitored; performing spatial domain noise reduction processing on the acoustic signal to obtain an acoustic signal after spatial domain noise reduction; performing frequency domain noise reduction processing on the acoustic signal after spatial domain noise reduction using the operating condition parameters to obtain an acoustic signal after noise separation; extracting time domain features, frequency domain features, and nonlinear features from the acoustic signal after noise separation, and extracting operating condition features from the operating condition parameters; performing feature fusion on the time domain features, frequency domain features, nonlinear features, and operating condition features to obtain a multidimensional feature matrix; inputting the multidimensional feature matrix into a trained neural network model to obtain equipment warning information output by the neural network model after identifying the multidimensional feature matrix. Thus, the weights of different acoustic signals are dynamically adjusted using the signal-to-noise ratio, thereby improving the effect of aerodynamic noise removal, thereby achieving dynamic identification of equipment faults based on changes in the actual equipment operating environment, and improving the accuracy of equipment aerodynamic fault identification.
[0046] It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the disclosure. BRIEF DESCRIPTION OF THE DRAWINGS
[0047] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the invention and, together with the description, serve to explain the principles of the invention.
[0048] Figure 1 The present invention is a flowchart showing a method for intelligent monitoring of transfer equipment in a thermal power plant based on multimodal noise analysis according to an exemplary embodiment.
[0049] Figure 2 The present invention is a block diagram showing an intelligent monitoring device for thermal power plant transfer equipment based on multimodal noise analysis according to an exemplary embodiment.
[0050] Figure 3 The present invention is a block diagram showing an apparatus for an intelligent monitoring method of thermal power plant transfer equipment based on multimodal noise analysis according to an exemplary embodiment. DETAILED DESCRIPTION
[0051] Exemplary embodiments will be described in detail herein, examples of which are illustrated in the accompanying drawings. In the following description, when referring to the drawings, like numbers in different figures represent like or similar elements unless otherwise indicated. The embodiments described in the following exemplary embodiments are not intended to represent all possible embodiments consistent with the present invention. Rather, they are merely examples of apparatus and methods consistent with certain aspects of the present invention, as detailed in the appended claims.
[0052] The terms used in the present disclosure are only for the purpose of describing specific embodiments and are not intended to limit the present disclosure. The singular forms "a", "an" and "the" used in the present disclosure and the appended claims are also intended to include the plural forms unless the context clearly indicates otherwise.
[0053] It should be understood that although the terms first, second, third, etc. may be used to describe various information in the embodiments of the present disclosure, such information should not be limited to these terms. These terms are only used to distinguish information of the same type from each other. For example, without departing from the scope of the embodiments of the present disclosure, the first information may also be referred to as the second information, and similarly, the second information may also be referred to as the first information. Depending on the context, the words "if" and "if" as used herein may be interpreted as "at the time of" or "when" or "in response to determining".
[0054] Furthermore, the various forms of processes shown in the embodiments of this disclosure may be used to reorder, add, or delete steps. For example, the steps described in this application may be performed in parallel, sequentially, or in a different order, as long as the desired results of the technical solutions disclosed in this disclosure can be achieved. This is not a limitation herein.
[0055] In related technologies, traditional monitoring methods rely on vibration signal analysis, have high requirements for the installation of sensors on equipment, and are difficult to capture aerodynamic faults. Traditional vibration monitoring has low sensitivity to aerodynamic faults such as blade cavitation and duct surge, and cannot detect early microscopic damage. In addition, when mechanical vibration and aerodynamic noise are coupled, such as when bearing wear and blade cracks exist at the same time, the same monitoring method is difficult to distinguish the root cause of the equipment failure.
[0056] To address the above-mentioned issues, the present disclosure provides an intelligent monitoring method for thermal power plant transfer equipment based on multimodal noise analysis. The method comprises the following steps: obtaining acoustic signals and operating parameters of the equipment to be monitored; performing spatial domain noise reduction processing on the acoustic signals to obtain a spatially noise-reduced acoustic signal; performing frequency domain noise reduction processing on the spatially noise-reduced acoustic signal using the operating parameters to obtain a noise-separated acoustic signal; extracting time domain features, frequency domain features, and nonlinear features from the noise-separated acoustic signal, and extracting operating condition features from the operating parameters; performing feature fusion on the time domain features, frequency domain features, nonlinear features, and operating condition features to obtain a multidimensional feature matrix; and inputting the multidimensional feature matrix into a trained neural network model to obtain equipment warning information output by the neural network model after identifying the multidimensional feature matrix. The method thus dynamically adjusts the weights of different acoustic signals using the signal-to-noise ratio, thereby improving the effect of aerodynamic noise removal. This method further enables dynamic identification of equipment faults based on changes in the actual equipment operating environment, thereby improving the accuracy of equipment aerodynamic fault identification.
[0057] Figure 1 FIG. 1 is a flow chart showing an intelligent monitoring method for thermal power plant transfer equipment based on multimodal noise analysis according to an exemplary embodiment. Figure 1 As shown, it should be noted that the intelligent monitoring method for thermal power plant transfer equipment based on multimodal noise analysis of the embodiment of the present disclosure is applied to the intelligent monitoring method device for thermal power plant transfer equipment based on multimodal noise analysis. Figure 1 As shown, the method may include the following steps:
[0058] Step 101: Acquire acoustic signals and operating parameters of the equipment to be monitored.
[0059] In one embodiment, the operating parameters may include physical quantities directly related to the operating state of the equipment, such as any one or more of rotation speed, temperature, pressure, and load.
[0060] As an example, acoustic signals at different positions may be acquired through multiple sensors, and the installation positions of the multiple sensors are optimized using a beamforming algorithm.
[0061] In one example, a multi-band composite sensor array can be pre-arranged: wideband microphones (20Hz-20kHz, 6 groups arranged in a ring with a spacing of λ / 2). Ultrasonic sensors (20-100kHz, 2 groups with diagonal focus). An embedded working condition synchronization module (real-time acquisition of speed, pressure, and temperature). The use of wideband microphones (20Hz-20kHz) and ultrasonic sensors (20-100kHz) can cover low-frequency mechanical vibrations and high-frequency aerodynamic noise, directly capturing aerodynamic fault signals (such as the high-frequency noise of blade cracks). The high-frequency characteristics of ultrasonic sensors (such as APC 200) can detect early microscopic damage (such as microcracks), solving the problem of traditional vibration monitoring missing early fault detection.
[0062] In addition, the above-mentioned sensors can be installed in a non-contact manner, and the sensors do not need to contact the surface of the equipment (such as through a ring arrangement and focusing technology), avoiding shutdowns for drilling and reducing unit maintenance costs.
[0063] Step 102: Perform spatial domain noise reduction processing on the acoustic signal to obtain an acoustic signal after spatial domain noise reduction.
[0064] In some embodiments of the present application, step 102 may specifically include the following steps:
[0065] The acoustic signal is subjected to spatial domain noise reduction processing using the following formula:
[0066]
[0067] Among them, Y(f) is the frequency domain signal after noise reduction, W m (f) is the dynamic weight of the mth sensor, which is calculated based on the signal-to-noise ratio, X m (f) is the acoustic signal collected by the mth sensor, τ m The time delay is calculated from the geometric distance and sound speed from the acoustic signal source to the sensor, ensuring that the target signal is in phase when superimposed and the noise is offset due to phase confusion. is the phase compensation term, and j is the imaginary unit.
[0068] The above-mentioned spatial domain noise reduction processing can enhance the target sound source signal and suppress background noise by adjusting the sensor weights and time delay.
[0069] In some embodiments of the present application, the dynamic weight W of the mth sensor m (f) is calculated by the following formula:
[0070]
[0071] Among them, SNR m (f) is the signal-to-noise ratio of the mth sensor.
[0072] Step 103 : performing frequency domain noise reduction processing on the acoustic signal after spatial domain noise reduction using the operating condition parameters to obtain an acoustic signal after noise separation.
[0073] In some embodiments of the present application, there are multiple groups of acoustic signals, each group of acoustic signals is collected by a sensor, and step 103 may specifically include the following steps:
[0074] Step a1: determine the signal-to-noise ratio of each sensor.
[0075] In one embodiment, the square of the signal amplitude of the mth sensor in the frequency domain can be averaged to obtain an average value. When the equipment is shut down or there is a known absence of a target signal, the background noise power is measured. The proportion of the average value in the background noise power is the signal-to-noise ratio.
[0076] Step a2: For the acoustic signal after spatial domain noise reduction corresponding to each sensor, weighted processing is performed on the acoustic signal after spatial domain noise reduction using the signal-to-noise ratio of the sensor to obtain an intermediate acoustic signal.
[0077] It can be understood that by performing weighted processing on acoustic signals based on the signal-to-noise ratio, the signals collected by sensors with high signal-to-noise ratios are amplified, and the signals collected by sensors with low signal-to-noise ratios are weakened, thereby enhancing high-quality signals, suppressing low-quality noise, and improving fault detection accuracy.
[0078] In some embodiments of the present application, step a2 may specifically include the following steps:
[0079] Amplifying the signal-to-noise ratio of the sensor respectively to obtain an amplified signal-to-noise ratio;
[0080] Normalize the signal-to-noise ratio after amplification to obtain the dynamic weight of the sensor;
[0081] The acoustic signal after spatial domain denoising is weighted using dynamic weights to obtain an intermediate acoustic signal.
[0082] In one embodiment, the signal-to-noise ratio of each sensor may be amplified by 1.5 to enhance the advantage of high-quality signals, and then normalized to obtain a dynamic weight.
[0083] Step a3: Perform frequency domain noise reduction processing on the intermediate acoustic signal using the operating condition parameters to obtain an acoustic signal after noise separation.
[0084] In some embodiments of the present application, the operating condition parameter includes the equipment speed, and step a3 may specifically include the following steps:
[0085] The intermediate acoustic signal is subjected to frequency domain noise reduction processing using the following adaptive comb filter transfer function H(z) to obtain an acoustic signal after noise separation:
[0086]
[0087] Among them, f s is the signal sampling frequency, n is the device speed, and D is the number of sampling points per device revolution. These factors are used to match the fundamental frequency and its harmonics. The numerator and denominator coefficients (0.8 and 0.2) can be determined through experimental optimization and are used to control the filter's frequency response, forming periodic notches and suppressing fundamental frequency interference.
[0088] It should be noted that dynamically adjusting the threshold or filter parameters of noise analysis through operating condition parameters avoids the false alarm problem of the traditional fixed threshold method during load fluctuations. Dynamically adjusting the filter parameter D according to the real-time speed eliminates baseband interference (such as noise drift caused by load changes).
[0089] In addition, in the present application, the dynamic weights are combined with the adaptive comb filter, which avoids the problem of insensitivity to speed changes when using the dynamic weights alone, and also avoids the problem of erroneous deletion of useful signals when using the adaptive comb filter alone.
[0090] Step 104 : extracting time domain features, frequency domain features, and nonlinear features from the acoustic signal after noise separation, and extracting operating condition features from the operating condition parameters.
[0091] In one embodiment, time domain features may include waveform complexity (such as kurtosis coefficient), reflecting impact faults (such as bearing spalling); frequency domain features may include dynamic Mel cepstral coefficients (DMCC), characterizing the spectral characteristics of aerodynamic noise; nonlinear features may include wavelet packet entropy, quantifying signal complexity (suitable for early crack detection); operating condition features may include real-time speed n, pressure P, and temperature T as additional feature inputs.
[0092] Step 105 , performing feature fusion on the time domain features, frequency domain features, nonlinear features and operating condition features to obtain a multi-dimensional feature matrix.
[0093] For example, the multidimensional feature matrix F = [time domain features, frequency domain features, nonlinear features, n, P, T].
[0094] In some embodiments of the present application, before step 105, the method may further include:
[0095] The minimum redundancy maximum correlation algorithm is used to screen key features from time domain features, frequency domain features, nonlinear features and operating condition features, and the screened time domain features, frequency domain features, nonlinear features and operating condition features are obtained;
[0096] Step 105 may include: performing feature fusion on the filtered time domain features, frequency domain features, nonlinear features, and operating condition features to obtain a multi-dimensional feature matrix.
[0097] In the embodiment of the present application, by extracting the time domain (waveform complexity), frequency domain (dynamic Mel cepstrum), and nonlinear features (wavelet packet entropy), and combining the minimum redundancy maximum correlation (mRMR) algorithm to screen key features, the feature dimension and fault sensitivity are significantly improved.
[0098] Step 106: input the multidimensional feature matrix into the trained neural network model to obtain the equipment warning information output by the neural network model after identifying the multidimensional feature matrix.
[0099] In one embodiment, the neural network model can be trained using a pre-training dataset (historical data of 300 units) and fine-tuning data (30 days of health data of the target unit), so that the model has the ability to recognize patterns of complex faults (such as bearing wear + blade cracks).
[0100] As an example, the neural network model can be a long short-term memory network LSTM. In the LSTM layer, the speed n is used as a time series input to help the model learn the dynamic impact of load changes on noise characteristics.
[0101] In one embodiment, the device warning information may include the fault type and the corresponding fault probability. The sound source position may also be calculated by combining a beamforming algorithm with the rotation speed n.
[0102] According to the embodiment of the present disclosure, the intelligent monitoring method for thermal power plant transfer equipment based on multimodal noise analysis is proposed. The method obtains the acoustic signal and operating parameters of the equipment to be monitored; performs spatial domain noise reduction processing on the acoustic signal to obtain a spatial domain noise-reduced acoustic signal; performs frequency domain noise reduction processing on the spatial domain noise-reduced acoustic signal using the operating parameters to obtain a noise-separated acoustic signal; extracts time domain features, frequency domain features, and nonlinear features from the noise-separated acoustic signal, and extracts operating condition features from the operating parameters; performs feature fusion on the time domain features, frequency domain features, nonlinear features, and operating condition features to obtain a multidimensional feature matrix; and inputs the multidimensional feature matrix into a trained neural network model to obtain equipment warning information output by the neural network model after identifying the multidimensional feature matrix. The method thus dynamically adjusts the weights of different acoustic signals using the signal-to-noise ratio, improves the effect of aerodynamic noise removal, and thus achieves dynamic identification of equipment faults based on changes in the actual equipment operating environment, thereby improving the accuracy of equipment aerodynamic fault identification.
[0103] Figure 2This is a block diagram of an intelligent monitoring device for thermal power plant transfer equipment based on multimodal noise analysis according to an exemplary embodiment. Figure 2 The device includes an acquisition unit 201, a spatial domain noise reduction unit 202, a frequency domain noise reduction unit 203, an extraction unit 204, a fusion unit 205 and an early warning unit 206.
[0104] The acquisition unit 201 is used to acquire the acoustic signal and operating parameters of the equipment to be monitored;
[0105] The spatial domain noise reduction unit 202 is configured to perform spatial domain noise reduction processing on the acoustic signal to obtain an acoustic signal after spatial domain noise reduction;
[0106] The frequency domain noise reduction unit 203 is used to perform frequency domain noise reduction processing on the acoustic signal after spatial domain noise reduction using the working condition parameters to obtain an acoustic signal after noise separation;
[0107] An extraction unit 204 is configured to extract time domain features, frequency domain features, and nonlinear features from the acoustic signal after noise separation, and to extract operating condition features from the operating condition parameters;
[0108] A fusion unit 205 is used to fuse the time domain features, frequency domain features, nonlinear features and operating condition features to obtain a multi-dimensional feature matrix;
[0109] The early warning unit 206 is used to input the multidimensional feature matrix into the trained neural network model to obtain the equipment early warning information output after the neural network model recognizes the multidimensional feature matrix.
[0110] In some embodiments of the present application, there are multiple groups of acoustic signals, each group of acoustic signals is collected by a sensor, and the frequency domain noise reduction unit 203 can be specifically used to:
[0111] Determine the signal-to-noise ratio of each sensor separately;
[0112] For the acoustic signal after spatial domain noise reduction corresponding to each sensor, the acoustic signal after spatial domain noise reduction is weighted using the signal-to-noise ratio of the sensor to obtain an intermediate acoustic signal;
[0113] The intermediate acoustic signal is subjected to frequency domain noise reduction processing using the operating condition parameters to obtain the acoustic signal after noise separation.
[0114] In some embodiments of the present application, the frequency domain noise reduction unit 203 may be specifically used to:
[0115] Amplifying the signal-to-noise ratio of the sensor respectively to obtain an amplified signal-to-noise ratio;
[0116] Normalize the signal-to-noise ratio after amplification to obtain the dynamic weight of the sensor;
[0117] The acoustic signal after spatial domain denoising is weighted using dynamic weights to obtain an intermediate acoustic signal.
[0118] In some embodiments of the present application, the operating condition parameter includes the equipment speed, and the frequency domain noise reduction unit 203 can be specifically used to:
[0119] The intermediate acoustic signal is subjected to frequency domain noise reduction processing using the following filter transfer function H(z) to obtain an acoustic signal after noise separation:
[0120]
[0121] Among them, f s is the signal sampling frequency, n is the device speed, and D is the number of sampling points corresponding to one revolution of the device.
[0122] In some embodiments of the present application, the spatial domain noise reduction unit 202 may be specifically configured to:
[0123] The acoustic signal is subjected to spatial domain noise reduction processing using the following formula:
[0124]
[0125] Among them, Y(f) is the frequency domain signal after noise reduction, W m (f) is the dynamic weight of the mth sensor, which is calculated based on the signal-to-noise ratio, X m (f) is the acoustic signal collected by the mth sensor, τ m is the time delay, which is calculated from the geometric distance from the acoustic signal source to the sensor and the speed of sound. is the phase compensation term.
[0126] In some embodiments of the present application, the dynamic weight of the mth sensor is calculated using the following formula:
[0127]
[0128] Among them, SNR m (f) is the signal-to-noise ratio of the mth sensor.
[0129] In some embodiments of the present application, the apparatus may further include:
[0130] A screening unit is used to screen key features from the time domain features, frequency domain features, nonlinear features and operating condition features using a minimum redundancy maximum correlation algorithm to obtain screened time domain features, frequency domain features, nonlinear features and operating condition features;
[0131] The fusion unit 205 can be specifically used for:
[0132] The filtered time domain features, frequency domain features, nonlinear features and operating condition features are fused to obtain a multi-dimensional feature matrix.
[0133] Regarding the apparatus in the above embodiment, the specific manner in which each module performs operations has been described in detail in the embodiment of the method, and will not be elaborated here.
[0134] According to the embodiment of the present disclosure, the intelligent monitoring device for thermal power plant transfer equipment based on multimodal noise analysis obtains the acoustic signal and operating parameters of the equipment to be monitored; performs spatial domain noise reduction processing on the acoustic signal to obtain a spatial domain noise-reduced acoustic signal; performs frequency domain noise reduction processing on the spatial domain noise-reduced acoustic signal using the operating parameters to obtain a noise-separated acoustic signal; extracts time domain features, frequency domain features, and nonlinear features from the noise-separated acoustic signal, and extracts operating condition features from the operating parameters; performs feature fusion on the time domain features, frequency domain features, nonlinear features, and operating condition features to obtain a multidimensional feature matrix; and inputs the multidimensional feature matrix into a trained neural network model to obtain equipment warning information output by the neural network model after identifying the multidimensional feature matrix. The device thus dynamically adjusts the weights of different acoustic signals using the signal-to-noise ratio, thereby improving the effect of aerodynamic noise removal, thereby achieving dynamic identification of equipment faults based on changes in the actual equipment operating environment and improving the accuracy of equipment aerodynamic fault identification.
[0135] Figure 3 This is a block diagram illustrating an apparatus for intelligently monitoring thermal power plant transfer equipment based on multimodal noise analysis, according to an exemplary embodiment. For example, apparatus 300 may be an electronic device, such as a mobile phone, computer, digital broadcast terminal, messaging device, game console, tablet device, medical device, fitness equipment, personal digital assistant, etc.
[0136] Reference Figure 3 , apparatus 300 may include one or more of the following components: a processing component 302 , a memory 304 , a power component 306 , a multimedia component 308 , an audio component 310 , an input / output (I / O) interface 312 , a sensor component 314 , and a communication component 316 .
[0137] The processing component 302 generally controls the overall operation of the device 300, such as operations associated with display, phone calls, data communications, camera operation, and recording operations. The processing component 302 may include one or more processors 320 to execute instructions to perform all or part of the steps of the above-described method. In addition, the processing component 302 may include one or more modules to facilitate interaction between the processing component 302 and other components. For example, the processing component 302 may include a multimedia module to facilitate interaction between the multimedia component 308 and the processing component 302.
[0138] The memory 304 is configured to store various types of data to support operations on the device 300. Examples of such data include instructions for any application or method operating on the device 300, contact data, phone book data, messages, pictures, videos, etc. The memory 304 can be implemented by any type of volatile or non-volatile storage device, or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk, or optical disk.
[0139] The power component 306 provides power to the various components of the device 300. The power component 306 may include a power management system, one or more power supplies, and other components associated with generating, managing, and distributing power to the device 300.
[0140] The multimedia component 308 includes a screen that provides an output interface between the device 300 and the user. In some embodiments, the screen may include a liquid crystal display (LCD) and a touch panel (TP). If the screen includes a touch panel, the screen may be implemented as a touch screen to receive input signals from the user. The touch panel includes one or more touch sensors to sense touches, slides, and gestures on the touch panel. The touch sensor can not only sense the boundaries of a touch or slide action, but also detect the duration and pressure associated with the touch or slide operation. In some embodiments, the multimedia component 308 includes a front camera and / or a rear camera. When the device 300 is in an operating mode, such as a shooting mode or a video mode, the front camera and / or the rear camera can receive external multimedia data. Each front camera and rear camera can be a fixed optical lens system or have focal length and optical zoom capabilities.
[0141] The audio component 310 is configured to output and / or input audio signals. For example, the audio component 310 includes a microphone (MIC) that is configured to receive external audio signals when the device 300 is in an operating mode, such as a call mode, a recording mode, and a voice recognition mode. The received audio signals may be further stored in the memory 304 or transmitted via the communication component 316. In some embodiments, the audio component 310 further includes a speaker for outputting audio signals.
[0142] I / O interface 312 provides an interface between processing component 302 and peripheral interface modules, such as a keyboard, click wheel, buttons, etc. These buttons may include but are not limited to: a home button, volume buttons, a start button, and a lock button.
[0143] The sensor assembly 314 includes one or more sensors for providing various aspects of the status assessment of the device 300. For example, the sensor assembly 314 can detect the open / closed state of the device 300, the relative positioning of components, such as the display and keypad of the device 300. The sensor assembly 314 can also detect changes in the position of the device 300 or a component of the device 300, the presence or absence of user contact with the device 300, the orientation or acceleration / deceleration of the device 300, and temperature changes of the device 300. The sensor assembly 314 may include a proximity sensor configured to detect the presence of nearby objects without any physical contact. The sensor assembly 314 may also include an optical sensor, such as a CMOS or CCD image sensor, for use in imaging applications. In some embodiments, the sensor assembly 314 may also include an accelerometer, a gyroscope sensor, a magnetic sensor, a pressure sensor, or a temperature sensor.
[0144] The communication component 316 is configured to facilitate wired or wireless communication between the device 300 and other devices. The device 300 can access a wireless network based on a communication standard, such as WiFi, 2G or 3G, or a combination thereof. In an exemplary embodiment, the communication component 316 receives a broadcast signal or broadcast-related information from an external broadcast management system via a broadcast channel. In an exemplary embodiment, the communication component 316 also includes a near field communication (NFC) module to facilitate short-range communication. For example, the NFC module can be implemented based on radio frequency identification (RFID) technology, infrared data association (IrDA) technology, ultra-wideband (UWB) technology, Bluetooth (BT) technology and other technologies.
[0145] In an exemplary embodiment, the apparatus 300 may be implemented by one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors, or other electronic components to perform the above-described method.
[0146] In an exemplary embodiment, a non-transitory computer-readable storage medium including instructions is also provided, such as a memory 304 including instructions, which can be executed by the processor 320 of the apparatus 300 to perform the above method. For example, the non-transitory computer-readable storage medium can be a ROM, a random access memory (RAM), a CD-ROM, a magnetic tape, a floppy disk, an optical data storage device, etc.
[0147] In an exemplary embodiment, a computer program product is also provided, comprising a computer program, which implements the above method when executed by the processor 320 of the apparatus 300 .
[0148] Other embodiments of the present invention will readily occur to those skilled in the art after considering the specification and practicing the invention disclosed herein. This disclosure is intended to cover any variations, uses, or adaptations of the invention that follow from the general principles of the invention and include common knowledge or customary techniques in the art not disclosed herein. The description and examples are to be considered as exemplary only, with the true scope and spirit of the invention being indicated by the following claims.
[0149] It should be understood that the present invention is not limited to the exact construction described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof. The scope of the present invention is limited only by the appended claims.
Claims
1. An intelligent monitoring method for thermal power plant transfer equipment based on multimodal noise analysis, characterized in that: include: Obtain the acoustic signals and operating parameters of the equipment to be monitored; Performing spatial domain noise reduction processing on the acoustic signal to obtain an acoustic signal after spatial domain noise reduction; Performing frequency domain noise reduction processing on the acoustic signal after spatial domain noise reduction using the operating condition parameters to obtain an acoustic signal after noise separation; Extracting time domain features, frequency domain features and nonlinear features from the acoustic signal after noise separation, and extracting operating condition features from the operating condition parameters; Performing feature fusion on the time domain features, frequency domain features, nonlinear features and operating condition features to obtain a multi-dimensional feature matrix; The multidimensional feature matrix is input into a trained neural network model to obtain equipment warning information output after the neural network model identifies the multidimensional feature matrix.
2. The intelligent monitoring method for thermal power plant transfer equipment based on multimodal noise analysis according to claim 1 is characterized in that: There are multiple groups of acoustic signals, and each group of acoustic signals is collected by a sensor; The method of performing frequency domain noise reduction processing on the acoustic signal after spatial domain noise reduction by using the operating condition parameter to obtain the acoustic signal after noise separation includes: Determine the signal-to-noise ratio of each sensor separately; For each sensor, the acoustic signal after spatial domain noise reduction is weighted by using the signal-to-noise ratio of the sensor to obtain an intermediate acoustic signal. The operating condition parameters are used to perform frequency domain noise reduction processing on the intermediate acoustic signal to obtain an acoustic signal after noise separation.
3. The intelligent monitoring method for thermal power plant transfer equipment based on multimodal noise analysis according to claim 2 is characterized in that: The acoustic signal after the spatial domain noise reduction is weighted by using the signal-to-noise ratio of the sensor to obtain an intermediate acoustic signal, including: Amplifying the signal-to-noise ratios of the sensors to obtain amplified signal-to-noise ratios; Normalizing the signal-to-noise ratio after the amplification process to obtain a dynamic weight of the sensor; The dynamic weight is used to perform weighted processing on the acoustic signal after the spatial domain noise reduction to obtain an intermediate acoustic signal.
4. The intelligent monitoring method for thermal power plant transfer equipment based on multimodal noise analysis according to claim 2 is characterized in that: The operating condition parameter includes a device speed; and performing frequency domain noise reduction processing on the intermediate acoustic signal using the operating condition parameter to obtain a noise-separated acoustic signal includes: The intermediate acoustic signal is subjected to frequency domain noise reduction processing using the following filter transfer function H(z) to obtain the acoustic signal after noise separation: Among them, f s is the signal sampling frequency, n is the device speed, and D is the number of sampling points corresponding to one revolution of the device.
5. The intelligent monitoring method for thermal power plant transfer equipment based on multimodal noise analysis according to claim 1 is characterized in that: The performing spatial domain noise reduction processing on the acoustic signal to obtain the acoustic signal after spatial domain noise reduction includes: The acoustic signal is subjected to spatial domain noise reduction processing using the following formula: Among them, Y(f) is the frequency domain signal after noise reduction, W m (f) is the dynamic weight of the mth sensor, which is calculated based on the signal-to-noise ratio, X m (f) is the acoustic signal collected by the mth sensor, τ m is the time delay, which is calculated from the geometric distance from the acoustic signal source to the sensor and the speed of sound. is the phase compensation term.
6. The intelligent monitoring method for thermal power plant transfer equipment based on multimodal noise analysis according to claim 5 is characterized in that: The dynamic weight of the mth sensor is calculated by the following formula: Among them, SNR m (f) is the signal-to-noise ratio of the mth sensor.
7. The intelligent monitoring method for thermal power plant transfer equipment based on multimodal noise analysis according to claim 1 is characterized in that: Before fusing the time domain features, frequency domain features, nonlinear features, and operating condition features to obtain a multi-dimensional feature matrix, the method further includes: Using a minimum redundancy maximum correlation algorithm to screen key features from the time domain features, frequency domain features, nonlinear features, and operating condition features, to obtain screened time domain features, frequency domain features, nonlinear features, and operating condition features; The feature fusion of the time domain features, frequency domain features, nonlinear features and operating condition features is performed to obtain a multidimensional feature matrix, including: The filtered time domain features, frequency domain features, nonlinear features and operating condition features are fused to obtain a multi-dimensional feature matrix.
8. An intelligent monitoring device for thermal power plant transfer equipment based on multimodal noise analysis, characterized in that: include: An acquisition unit, used to acquire acoustic signals and operating parameters of the equipment to be monitored; A spatial domain noise reduction unit, configured to perform spatial domain noise reduction processing on the acoustic signal to obtain an acoustic signal after spatial domain noise reduction; a frequency domain noise reduction unit, configured to perform frequency domain noise reduction processing on the acoustic signal after the spatial domain noise reduction using the operating condition parameters to obtain an acoustic signal after noise separation; an extraction unit, configured to extract time domain features, frequency domain features, and nonlinear features from the acoustic signal after noise separation, and to extract operating condition features from the operating condition parameters; A fusion unit, configured to fuse the time domain features, frequency domain features, nonlinear features, and operating condition features to obtain a multidimensional feature matrix; The early warning unit is used to input the multidimensional feature matrix into the trained neural network model to obtain the equipment early warning information output after the neural network model identifies the multidimensional feature matrix.
9. An electronic device, characterized in that: include: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the method according to any one of claims 1 to 7 is implemented.
10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method according to any one of claims 1 to 7 is implemented.