Hydroelectric generating set flat pressing pipe working condition monitoring method based on MFCC and fault model library

Abstract

The invention relates to a hydroelectric generating set flat pressing pipe working condition monitoring method based on MFCC and a fault model library, and the method comprises the steps: collecting a voiceprint signal of a flat pressing pipe through a voiceprint sensor, and carrying out the monitoring of the working condition of the flat pressing pipe according to the known historical data of a voiceprint signal, a liquid level signal, a vibration signal and a pressure pulsation signal of a flat pressing pipe fault; a voiceprint model library of common faults including typical voiceprint features is established and is used for carrying out rapid comparison according to the voiceprint signals of the current flat pressing pipe and identifying and judging whether the common faults occur in the current working condition state or not; and a fault model library containing voiceprint, liquid level, vibration and pressure pulsation characteristic data is constructed and is used for identifying and judging whether a known fault occurs in the current working condition state according to comparison of voiceprint, liquid level, vibration and pressure pulsation signals of the current unit. According to the invention, real-time on-line high-precision automatic identification of the fault of the flat pressing pipe is realized, the efficiency and the precision of fault diagnosis are obviously improved, and the downtime and the maintenance cost caused by misdiagnosis or missed diagnosis are reduced.

Description

Technical Field

[0001] The invention belongs to the field of safety monitoring of hydropower units, and in particular relates to a method for monitoring the working condition of a surge pipe of a hydropower unit based on MFCC and a fault model library. Background Art

[0002] The surge tank of the top cover of a hydropower station is an important part of the hydropower station unit. Its main function is to balance the accumulated water pressure between the top cover and the upper cavity of the runner, reduce the axial water thrust transmitted to the thrust bearing, and ensure the safe and stable operation of the unit. The vibration of the top cover often drives the surge tank to vibrate. Under the long-term action of this vibration, the surge tank bolts may loosen, the seal may be damaged, and the weld may crack. What is more complicated is that the water flow in the surge tank often carries sediment, which will cause wear on the inner wall of the surge tank during the long-term flow process, making the pipe wall gradually thinner. At the same time, the existence of cavitation further aggravates the damage to the pipe wall, especially at the welding point between the surge tank and the top cover or the inner wall of the surge tank, where sand holes are easily formed. The above reasons will cause the surge tank to leak. These leakage phenomena will not only affect the normal operation of the system, but also, in severe cases, may pose a threat to the safe and stable operation of the unit. At present, the treatment of surge tank leakage usually requires shutdown and drainage for maintenance, which not only increases the complexity of maintenance, but also prolongs the maintenance period, causing a great impact on production operation.

[0003] Traditional methods of monitoring water leakage in surge tanks mainly rely on manual inspections and regular inspections. Manual inspections are conducted by visual inspections, touch perception, and listening judgments of surge tanks and their ancillary equipment to detect possible water leakage problems. Inspectors need to have certain professional knowledge and experience, and be able to make a preliminary judgment on the operating status of surge tanks based on changes in the appearance, sound, temperature, etc. of the equipment. However, this method is affected by the inspectors' skill level, experience accumulation, and inspection cycle, and its accuracy and timeliness are often difficult to guarantee. Regular inspections are systematic and comprehensive inspections of surge tanks according to preset cycles. This inspection usually includes appearance inspections, performance tests, pressure tests, etc. of the equipment to assess the health of the surge tanks.

[0004] Regular inspections can improve the accuracy of monitoring to a certain extent, but there are also some problems. First, the setting of regular inspection cycles needs to be comprehensively considered based on various factors such as the operating status of the equipment and the use environment. If the cycle is too long, water leakage problems may not be discovered in time; if the cycle is too short, maintenance costs and workload will increase. Secondly, regular inspections still cannot monitor the real-time working conditions of the pressure-surge pipes and cannot detect sudden water leakage incidents in time.

[0005] Whether it is manual inspection or regular inspection, the inspection cycle is relatively long. The frequency of manual inspection is limited by the number, skills and work arrangements of the inspectors, while regular inspection is limited by the preset cycle. The long inspection cycle makes it impossible for traditional methods to detect and deal with water leakage problems in a timely manner, especially in the early stage of water leakage, which may cause the problem to worsen due to failure to detect it in time.

[0006] Traditional monitoring methods rely heavily on the experience, skills, and attention of inspectors. The judgement of inspectors may be affected by many factors, such as fatigue and lack of concentration, which may lead to missed detection or misjudgment. In addition, due to differences in everyone's perception and judgment, there may be inconsistent detection results between different inspectors.

[0007] In summary, the prior art lacks a real-time online monitoring method for abnormal operating conditions of a surge tank. Summary of the invention

[0008] The purpose of the present invention is to address the above problems and provide a method for monitoring the working condition of a surge pipe of a hydropower unit based on MFCC and a fault model library, establish a voiceprint model library of common faults containing typical voiceprint features and a fault model library containing multimodal data such as the voiceprint, liquid level, vibration and pressure pulsation signals of the unit, extract feature vectors from the voiceprint signal of the surge pipe under the current working condition, and compare them with the common faults in the voiceprint model library to achieve rapid identification of common faults; compare the feature vectors of the voiceprint, liquid level, vibration and pressure pulsation signals under the current working condition with the feature vectors of known faults in the fault model library to achieve automatic identification of known faults that have occurred in the surge pipe of the unit top cover; and use multimodal data fusion analysis to improve the accuracy of fault identification and real-time online monitoring of surge pipe abnormalities.

[0009] The technical solution of the present invention is a method for monitoring the working condition of a surge pipe of a hydropower unit based on MFCC and a fault model library, comprising the following steps: Step 1: Collect the soundprint signal of the equalizing pipe, the liquid level signal at the top cover, the vibration signal of the unit, and the pressure pulsation signal data when the unit is in working state; Step 2: Use the Mel-frequency cepstral coefficient method MFCC to extract the feature vector i-Vector from the equalizer tube voiceprint signal; Step 2.1: Pre-emphasize the voiceprint signal; Step 2.2: Frame and window the pre-emphasized voiceprint signal; Step 2.3: Use Fourier transform method to deconstruct the framed time domain signal into frequency domain signal; Step 2.4: Filter the obtained spectrum through a Mel filter and divide it into multiple Mel filter groups; Step 2.5: Endpoint detection, removing invalid data from the signal data; Step 2.6: Extract feature vector i-Vector; Step 3: Construct a probabilistic linear discriminant analysis model PLDA and train it to obtain the PLDA model parameters; Step 4: Using the PLDA model, calculate the similarity between the feature vector i-Vector in step 2 and the feature vector of a typical fault in the voiceprint model library, and score the current working condition of the surge tank; Step 5: Preliminarily judge whether the current working state of the surge tank is normal according to the score of the working state of the surge tank in step 4. If it is normal, execute step 1 to continue to collect the sound print, vibration, liquid level and pressure pulsation signal data of the surge tank in the working state; if the working state of the surge tank is abnormal, execute step 6; Step 6: Compare the characteristic data of the sound print, liquid level, vibration and pressure pulsation signals under the current working condition with the characteristic data of various faults in the fault model library, calculate the degree of correlation, and determine whether it is a known fault in the fault model library according to the degree of correlation. If it is a known fault, issue a fault alarm signal, otherwise execute step 7; Step 7: According to the liquid level signal data in the top cover obtained in step 1, determine whether the liquid level is normal. If it is normal, execute step 8. If it is abnormal, determine whether there is a water leakage point fault in the top cover and the equalizing pipe according to the image recognition of the top cover; Step 8: According to the vibration signal and pressure pulsation signal data of the unit in the current working state, determine whether the unit has abnormal vibration. If the unit has abnormal vibration, further determine whether the unit is in a special vibration zone. If so, adjust the working condition of the unit. If it is not in a special vibration zone, determine whether there are loose bolts, damaged seals, cracked welds or sand holes and water leakage faults based on the image recognition of the top cover and the equalizing pipe. If there is no abnormal vibration, execute step 9; Step 9: Conduct on-site inspection of the unit top cover and equalizing pipe to determine the fault type, and store the characteristic data of the sound print signal, liquid level signal, vibration signal and pressure pulsation signal of the fault type into the fault model library.

[0010] Preferably, in step 5, before executing step 6, it is first determined and identified whether it is a cable damage fault based on the voiceprint signal, and whether it is a phase A, phase B or phase C cable fault is determined based on the position of the voiceprint sensor.

[0011] Preferably, the step 2.6 specifically includes: 1) Build and train the Gaussian mixture model-universal background model GMM-UBM; 2) Inputting the MFCC feature vector of the voiceprint signal of the equalizer into the Gaussian mixture model-general background model, and obtaining the Gaussian mixture model parameters of the equalizer by the maximum a posteriori probability method; 3) Using the universal background model UBM and Baum-Welch statistics, the total variation model TVM is trained. TVM is a low-dimensional subspace model that is used to capture all possible variation factors in the speech signal. 4) Project the Baum-Welch statistic of the equalizer onto the low-dimensional subspace of TVM, and the resulting projection vector is the i-Vector of the equalizer.

[0012] Preferably, in step 4, the similarity between the characteristic vector i-Vector in step 2 and the characteristic vector of a typical fault in the voiceprint model library is calculated by the following formula: ; In the formula represents the similarity score, The characteristic vector i-Vector representing the voiceprint signal of the equalizing tube, The feature vector representing a typical fault in the voiceprint model library, Expressing the same assumption, that is, assuming and For the same fault type, Indicates different assumptions, that is, and For different fault types; In the same assumption Lower eigenvector and The joint probability density function value of In different assumptions Lower eigenvector The probability density function value of In different assumptions Lower eigenvector The probability density function value of .

[0013] Compared with the prior art, the beneficial effects of the present invention include: 1) The present invention constructs a voiceprint model library of common faults containing typical voiceprint features, using Mel frequency The cepstral coefficient method extracts feature vectors from the equalizer pipe voiceprint signal under the current working condition, and uses the probabilistic linear discriminant analysis model to calculate the similarity between the feature vector of the voiceprint signal under the current working condition and the common faults in the voiceprint model library, so as to identify and judge whether common faults with typical voiceprint characteristics occur under the current working condition, thus realizing the online rapid and high-precision automatic recognition of common equalizer pipe faults with typical voiceprint characteristics.

[0014] 2) The present invention constructs a fault model library of multimodal data such as the unit's soundprint, liquid level, vibration and pressure pulsation signals, calculates the degree of correlation between the characteristic vectors of the unit's soundprint, liquid level, vibration and pressure pulsation signals under the current operating condition and the characteristic vectors of known faults in the fault model library, determines whether a known fault occurs under the current operating condition, and realizes online real-time and high-precision identification and judgment of the unit's top cover equalizing pipe faults and abnormalities.

[0015] 3) After the method of the present invention determines the unknown fault through image recognition, on-site inspection and confirmation, the characteristic data of the unit sound print, liquid level, vibration and pressure pulsation signal corresponding to the fault are incorporated into the fault model library, and the content of the fault database is continuously enriched and improved, so that the monitoring system based on the present invention can automatically match and identify more quickly and accurately when facing similar faults in the future, which significantly improves the efficiency and accuracy of fault diagnosis and reduces the downtime and maintenance costs caused by misdiagnosis or missed diagnosis.

[0016] 4) Based on the voiceprint signal, the present invention realizes the rapid and efficient diagnosis of the damage fault of the A, B, and C phase transmission cables of the unit, and at the same time sends a cable damage alarm signal to the unit operation and maintenance personnel, so as to facilitate timely treatment measures to avoid affecting the normal power generation of the unit.

[0017] 5) The present invention is based on voiceprint recognition technology, and through comprehensive judgment and analysis of multimodal data such as images, vibrations, and liquid levels, it can greatly improve the accuracy and efficiency of identifying abnormal working conditions of the surge tank, and realize online real-time monitoring and identification of faults and abnormalities. It saves human resources and reduces the time required for fault identification, providing an opportunity for emergency handling of unit accidents and greatly improving the stability of the power system. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0019] Figure 1 It is a schematic structural diagram of the unit top cover and the equalizing pipe according to an embodiment of the present invention.

[0020] Figure 2 Schematic diagram of a data acquisition system according to an embodiment of the present invention.

[0021] Figure 3 The figure is a schematic diagram of a process for calculating a similarity score of a voiceprint signal in an embodiment of the present invention.

[0022] Figure 4 The present invention is a flowchart of a method for monitoring the working condition of a surge pipe of a hydropower unit according to an embodiment of the present invention.

[0023] Figure 5 It is a schematic diagram of a system for monitoring the working condition of a surge pipe of a hydropower unit according to an embodiment of the present invention. DETAILED DESCRIPTION

[0024] In the embodiment, eight bone conduction pickups 1 are arranged on the wall of the equalizing tube 7, two cameras 2 with a thermal infrared temperature measurement function are installed above the equalizing tube 7 and on the inner wall of the top cover 8, two horizontal vibration sensors 3 are arranged on the inner rib plate of the top cover 8, two vertical vibration sensors 4 are arranged on the inner rib plate of the top cover, two water level sensors 5 and two pressure pulsation sensors 6 are installed inside the top cover, one of the pressure pulsation sensors collects pressure pulsation signal data between the impeller 9 and the movable guide vane 10, and the other pressure pulsation sensor collects pressure pulsation signal data between the movable guide vane 10 and the fixed guide vane 11, as shown in FIG. Figure 1 shown.

[0025] like Figure 2 As shown, the data of the corresponding sensors are transmitted to the cloud server platform through the optical fiber channel for further analysis and judgment. The network transmission system includes optoelectronic switches, convergence switches, servers, forward isolation devices, cloud platforms, etc. Among them, the voiceprint feature data is set with a local control host, which has the function of feature data extraction and model downlink, which is convenient for voiceprint feature data extraction and known fault model judgment in the local control host. The thermal infrared camera has temperature measurement and image monitoring functions, and is also equipped with a local control host to facilitate uploading temperature feature values ​​and video files, and has the function of downlink camera motion control.

[0026] like Figure 3 and Figure 4 As shown in the figure, the hydropower unit surge pipe condition monitoring method based on MFCC and fault model library includes: Step 1: Collect the soundprint signal of the equalizing pipe, the liquid level signal at the top cover, the vibration signal of the unit, and the pressure pulsation signal data when the unit is in working state; Step 2: Use the Mel-frequency cepstral coefficient method MFCC to extract the feature vector i-Vector from the equalizer tube voiceprint signal; Step 2.1: Pre-emphasize the voiceprint signal; Step 2.2: Frame and window the pre-emphasized voiceprint signal; Step 2.3: Use Fourier transform method to deconstruct the framed time domain signal into frequency domain signal; Step 2.4: Filter the obtained spectrum through a Mel filter and divide it into multiple Mel filter groups; Step 2.5: Endpoint detection, removing invalid data from the signal data; Step 2.6: Extract feature vector i-Vector; Step 3: Construct a probabilistic linear discriminant analysis model PLDA and train it to obtain the PLDA model parameters; The probabilistic linear discriminant analysis model PLDA is as follows:

[0027] In the formula Indicates i The target object j Voiceprint sample data; represents the mean of all sample data, reflecting the central tendency of all sample data; F represents the identity space, h i represents the fixed effect part of the voiceprint feature sample data of the i-th target object, that is, the individual characteristics that do not change over time; G represents the error space, w ij It represents the random effect part of the jth sample data of the ith target object, reflecting the random changes or differences of the target object over time; It represents the noise part of the jth sample data of the ith target object, representing the random error or unpredictable factors in the observed data.

[0028] Step 4: Using the PLDA model, calculate the similarity between the feature vector i-Vector in step 2 and the feature vector of the typical fault in the voiceprint model library, and score the current working state of the surge tank. The score calculation formula is: ; In the formula represents the similarity score, The characteristic vector i-Vector representing the voiceprint signal of the equalizing tube, The feature vector representing a typical fault in the voiceprint model library, Expressing the same assumption, that is, assuming and For the same fault type, Indicates different assumptions, that is, and For different fault types; In the same assumption Lower eigenvector and The joint probability density function value of In different assumptions Lower eigenvector The probability density function value of In different assumptions Lower eigenvector The probability density function value of .

[0029] Step 5: Preliminarily judge whether the current working state of the surge tank is normal according to the score of the working state of the surge tank in step 4. If it is normal, execute step 1 to continue to collect the sound print, vibration, liquid level and pressure pulsation signal data of the surge tank in the working state; if the working state of the surge tank is abnormal, execute step 6; Step 6: Compare the characteristic data of the sound print, liquid level, vibration and pressure pulsation signals under the current working condition with the characteristic data of various faults in the fault model library, calculate the degree of correlation, and determine whether it is a known fault in the fault model library according to the degree of correlation. If it is a known fault, issue a fault alarm signal, otherwise execute step 7; The calculation formula for the degree of association is:

[0030] In the formula The characteristic vector representing the current working state of the surge tank The characteristic vector of the known fault The degree of correlation, is the resolution coefficient; The characteristic vector of the known fault is: ; In the formula Indicates the characteristic data of the flat-pressure tube voiceprint signal. Indicates the characteristic data of the liquid level signal at the top cover, Indicates the characteristic data of the unit's horizontal vibration signal. Indicates the characteristic data of the vertical vibration signal of the unit, It indicates the characteristic data of the pressure pulsation signal between the unit runner and the movable guide vane. Represents characteristic data of pressure pulsation signal between movable guide vanes and fixed guide vanes; The characteristic vector under the current working condition is:

[0031] and Correspondingly, They respectively represent the characteristic data of the sound pattern, liquid level, horizontal vibration, vertical vibration, and pressure pulsation signal under the current working condition.

[0032] Step 7: According to the liquid level signal data in the top cover obtained in step 1, determine whether the liquid level is normal. If it is normal, execute step 8. If it is abnormal, determine whether there is a water leakage point fault in the top cover and the equalizing pipe according to the image recognition of the top cover; Step 8: According to the vibration signal and pressure pulsation signal data of the unit in the current working state, determine whether the unit has abnormal vibration. If the unit has abnormal vibration, further determine whether the unit is in a special vibration zone. If so, adjust the working condition of the unit. If it is not in a special vibration zone, determine whether there are loose bolts, damaged seals, cracked welds or sand holes and water leakage faults based on the image recognition of the top cover and the equalizing pipe. If there is no abnormal vibration, execute step 9; Step 9: Conduct on-site inspection of the unit top cover and equalizing pipe to determine the fault type, and store the characteristic data of the sound print signal, liquid level signal, vibration signal and pressure pulsation signal of the fault type into the fault model library.

[0033] In the embodiment, before executing step 6, a vector quantization algorithm is used to determine and identify whether it is a cable damage fault based on the voiceprint feature signal, and determine whether it is a phase A, phase B, or phase C cable fault based on the position of the voiceprint sensor, and send a cable damage alarm signal to the unit operation and maintenance personnel.

[0034] The hydropower unit surge pipe condition monitoring system using the above method is as follows: Figure 5 As shown, it includes an acquisition module, a data network transmission module, a voiceprint training and scoring module, a comprehensive fault judgment module, and a result output execution module.

[0035] The acquisition module is used to collect multi-modal data of the unit top cover surge pipe under the current operating condition.

[0036] The data network transmission module is used to transmit the collected multimodal data to the cloud server platform.

[0037] The voiceprint training and scoring module is used to calculate the similarity score between the voiceprint signal feature data under the current working condition and the feature data of the voiceprint model library, and to identify and judge common faults with typical voiceprint features.

[0038] The comprehensive fault judgment module is used to compare the characteristic data of the unit's sound print, liquid level, vibration and pressure pulsation signals under the current operating condition with the fault characteristic data in the fault model library to determine the specific type of fault.

[0039] The result output execution module sends out corresponding alarm information or performs corresponding equipment operations according to the faults identified by the voiceprint training and scoring module and the comprehensive fault judgment module.

[0040] The present invention realizes real-time online monitoring of abnormal working conditions of the surge tank through voiceprint recognition technology, filling the industry gap in real-time online monitoring methods for abnormal working conditions of the surge tank; by installing bone conduction microphones, cameras with thermal infrared temperature measurement functions, vibration sensors, water level sensors, pressure pulsation sensors, etc. in suitable areas, and integrating multimodal data into the same system, comprehensive diagnosis of multimodal abnormal working conditions of the surge tank is realized, greatly improving the timeliness of abnormal working condition or fault judgment of the surge tank.

Claims

1. A method for monitoring the working condition of a surge pipe of a hydropower unit based on MFCC and a fault model library, characterized in that: The operating condition monitoring method uses a soundprint sensor to collect the soundprint signal of the surge tank, and establishes a soundprint model library of common faults containing typical soundprint features based on the historical data of the soundprint signal, liquid level signal, vibration signal and pressure pulsation signal of the known surge tank fault, so as to quickly compare, identify and judge whether common faults occur in the current operating condition according to the soundprint signal of the current surge tank; and constructs a fault model library containing soundprint, liquid level, vibration and pressure pulsation feature data, so as to compare, identify and judge whether known faults occur in the current operating condition according to the soundprint, liquid level, vibration and pressure pulsation signal of the current unit; The working condition monitoring method comprises the following steps: Step 1: Collect the soundprint signal of the equalizing pipe, the liquid level signal at the top cover, the vibration signal of the unit, and the pressure pulsation signal data when the unit is in working state; Step 2: Use the Mel-frequency cepstral coefficient method MFCC to extract the feature vector i-Vector from the equalizer tube voiceprint signal; Step 3: Construct a probabilistic linear discriminant analysis model PLDA and train it to obtain the PLDA model parameters; Step 4: Using the PLDA model, calculate the similarity between the feature vector i-Vector in step 2 and the feature vector of a typical fault in the voiceprint model library, and score the current working condition of the surge tank; Step 5: Preliminarily judge whether the current working state of the surge tank is normal according to the score of the working state of the surge tank in step 4. If it is normal, execute step 1 to continue to collect the sound print, vibration, liquid level and pressure pulsation signal data of the surge tank in the working state; if the working state of the surge tank is abnormal, execute step 6; Step 6: Compare the characteristic data of the sound print, liquid level, vibration and pressure pulsation signals under the current working condition with the characteristic data of various faults in the fault model library, calculate the degree of correlation, and determine whether it is a known fault in the fault model library according to the degree of correlation. If it is a known fault, issue a fault alarm signal, otherwise execute step 7; Step 7: According to the liquid level signal data in the top cover obtained in step 1, determine whether the liquid level is normal. If it is normal, execute step 8. If it is abnormal, determine whether there is a water leakage point fault in the top cover and the equalizing pipe according to the image recognition of the top cover; Step 8: According to the vibration signal and pressure pulsation signal data of the unit in the current working state, determine whether the unit has abnormal vibration. If the unit has abnormal vibration, further determine whether the unit is in a special vibration zone. If so, adjust the working condition of the unit. If it is not in a special vibration zone, determine whether there are loose bolts, damaged seals, cracked welds or sand holes and water leakage faults based on the image recognition of the top cover and the equalizing pipe. If there is no abnormal vibration, execute step 9; Step 9: Conduct on-site inspection of the unit top cover and equalizing pipe to determine the fault type, and store the characteristic data of the sound print signal, liquid level signal, vibration signal and pressure pulsation signal of the fault type into the fault model library.

2. The method for monitoring the working condition of the surge pipe of a hydropower unit according to claim 1 is characterized in that: In step 5, before executing step 6, first determine and identify whether it is a cable damage fault based on the voiceprint signal, and determine whether it is a phase A, phase B, or phase C cable fault based on the position of the voiceprint sensor, and issue a cable damage alarm signal.

3. The method for monitoring the working condition of the surge pipe of a hydropower unit according to claim 1 or 2, characterized in that: In the step 2, the Mel frequency cepstral coefficient method MFCC is used to extract the feature vector i-Vector, which specifically includes the following sub-steps: Step 2.1: Pre-emphasize the voiceprint signal; Step 2.2: Frame the pre-emphasized voiceprint signal; Step 2.3: Use Fourier transform method to deconstruct the framed time domain signal into frequency domain signal; Step 2.4: Filter the obtained spectrum through a Mel filter and divide it into multiple Mel filter groups; Step 2.5: Endpoint detection, removing invalid data from the signal data; Step 2.6: Extract feature vector i-Vector.

4. The method for monitoring the working condition of the surge pipe of a hydropower unit according to claim 3 is characterized in that: In step 2.4, the filter calculation formula is: ； Where k is the frequency index; m is the serial number of the Mel filter; f is the actual frequency; Represents the filter function.

5. The method for monitoring the working condition of the surge pipe of a hydropower unit according to claim 3 is characterized in that: The step 2.6 specifically includes: 1) Build and train the Gaussian mixture model-universal background model GMM-UBM; 2) Inputting the MFCC feature vector of the voiceprint signal of the equalizer into the Gaussian mixture model-general background model, and obtaining the Gaussian mixture model parameters of the equalizer by the maximum a posteriori probability method; 3) Using the universal background model UBM and Baum-Welch statistics, the total variation model TVM is trained. TVM is a low-dimensional subspace model that is used to capture all possible variation factors in the speech signal. 4) Project the Baum-Welch statistic of the equalizer onto the low-dimensional subspace of TVM, and the resulting projection vector is the i-Vector of the equalizer.

6. The method for monitoring the working condition of the surge pipe of a hydropower unit according to claim 5 is characterized in that: In step 3, the probabilistic linear discriminant analysis model PLDA is as follows: ； In the formula Indicates i The target object j Voiceprint sample data; represents the mean of all sample data; F represents the identity space, h i represents the fixed effect part of the voiceprint feature sample data of the i-th target object, that is, the individual characteristics that do not change over time; G represents the error space, w ij It represents the random effect part of the jth sample data of the ith target object, reflecting the random changes or differences of the target object over time; It represents the noise part of the jth sample data of the ith target object, representing the random error or unpredictable factors in the observed data.

7. The method for monitoring the working condition of the surge pipe of a hydropower unit according to claim 6 is characterized in that: In step 4, the similarity between the feature vector i-Vector in step 2 and the feature vector of the typical fault in the voiceprint model library is calculated by the following formula: ； In the formula represents the similarity score, The characteristic vector i-Vector representing the voiceprint signal of the equalizing tube, The feature vector representing a typical fault in the voiceprint model library, Expressing the same assumption, that is, assuming and For the same fault type, Indicates different assumptions, that is, and For different fault types; In the same assumption Lower eigenvector and The joint probability density function value of In different assumptions Lower eigenvector The probability density function value of In different assumptions Lower eigenvector The probability density function value of .

8. The method for monitoring the working condition of the surge pipe of a hydropower unit according to claim 7 is characterized in that: In step 6, the correlation degree is used to determine whether it is a known fault in the fault model library. The calculation formula for the correlation degree is: In the formula The characteristic vector representing the current working state of the surge tank The characteristic vector of the known fault The degree of correlation, is the resolution coefficient; The characteristic vector of the known fault is: ； In the formula Indicates the characteristic data of the flat-pressure tube voiceprint signal. Indicates the characteristic data of the liquid level signal at the top cover, Indicates the characteristic data of the unit's horizontal vibration signal. Indicates the characteristic data of the vertical vibration signal of the unit, Indicates the characteristic data of the pressure pulsation signal between the unit runner and the movable guide vane. Represents the characteristic data of the pressure pulsation signal between the movable guide vane and the fixed guide vane.

Images