A natural gas leak quantitative detection method and system based on voiceprint

Abstract

The present invention relates to the technical field of leakage detection, and discloses a natural gas leakage quantitative detection method and system based on voiceprint, the method comprising: collecting vibration signals of an acoustic wave sensor and determining the position of a leakage point according to a voiceprint array, obtaining the fluid leakage velocity of each fluid detection area, pre-processing the fluid leakage velocity to obtain the leakage velocity, constructing a leakage velocity sample for the corresponding fluid detection area according to the leakage velocity, determining the current leakage environment according to the increase number, decrease number and negative number; obtaining the historical leakage velocity of the leakage environment that is the same as the current leakage environment, comparing the leakage velocity with the historical leakage velocity, judging whether to correct the leakage velocity according to the comparison result, determining the leakage correction factor according to the similarity result between the leakage velocity and the historical leakage velocity data set, and determining the leakage interval according to the leakage type and the final leakage velocity. The present invention has good characteristics of quantitatively detecting leakage.

Description

Technical Field

[0001] The present invention relates to the technical field of leakage detection, and in particular to a method and system for quantitatively detecting natural gas leakage based on voiceprint. Background Art

[0002] With the widespread use of natural gas, the safety of natural gas leaks has become a major concern for public safety and environmental protection. While natural gas, as a clean energy source, offers high safety during use, it is highly susceptible to leakage in its gaseous state. Leakage accidents can lead to serious consequences such as explosions and fires, potentially endangering life and property.

[0003] Currently, there are some technical difficulties in the quantitative detection of natural gas leaks. For example, traditional leak detection methods can only identify the presence or absence of a leak, making it difficult to accurately classify the leak type, and are unable to distinguish reasonable levels based on the measured leakage flow rate, resulting in false alarms and missed opportunities for emergency response.

[0004] Therefore, how to provide a natural gas leakage quantitative detection method and system based on voiceprint is a technical problem that those skilled in the art urgently need to solve. Summary of the Invention

[0005] In view of this, the present invention proposes a natural gas leak quantitative detection method and system based on voiceprint, aiming to solve the problems in current quantitative detection technology that are unable to quantitatively detect the leakage flow rate according to the environment, reasonably define the leakage range, and classify the leakage type.

[0006] In one aspect, the present invention provides a method for quantitatively detecting natural gas leaks based on voiceprints, comprising:

[0007] Collecting vibration signals from the acoustic wave sensor and determining the location of the leakage point based on the soundprint array, dividing the leakage point into multiple fluid detection areas based on the soundprint array, obtaining the fluid leakage flow rate in each fluid detection area, and pre-processing the fluid leakage flow rate to obtain the leakage flow rate;

[0008] Constructing a leakage flow rate sample for the corresponding fluid detection area according to the leakage flow rate, wherein the leakage flow rate sample includes an increasing sample, a decreasing sample and a negative sample;

[0009] Counting the number of the increasing samples and recording it as an increasing number, counting the number of the decreasing samples and recording it as a decreasing number, counting the number of the negative samples and recording it as a negative number, and determining a current leakage environment according to the increasing number, the decreasing number, and the negative number;

[0010] Obtaining a historical leakage flow rate for the same leakage environment as the current leakage environment, comparing the leakage flow rate with the historical leakage flow rate, and determining whether to correct the leakage flow rate based on the comparison result;

[0011] When it is determined that the leakage flow rate needs to be corrected, a clustering algorithm is used to determine a historical leakage flow rate data set based on the historical leakage flow rate, a leakage correction factor is determined based on similarity between the leakage flow rate and the historical leakage flow rate data set, and the leakage flow rate is corrected based on the leakage correction factor to obtain a final leakage flow rate;

[0012] The leakage type is determined according to the final leakage flow rate, and the leakage interval is determined according to the leakage type and the final leakage flow rate.

[0013] Furthermore, when constructing a leakage flow rate sample for the corresponding fluid detection area according to the leakage flow rate, it includes:

[0014] Counting all the leakage flow rates and determining a standard leakage flow rate corresponding to each fluid detection area;

[0015] Comparing the leakage flow rate with the standard leakage flow rate, and if all leakage flow rates are greater than the standard leakage flow rate, establishing the growth sample for the corresponding fluid detection area;

[0016] If all leakage flow rates are less than or equal to the standard leakage flow rate, then establishing the reduced sample for the corresponding fluid detection area;

[0017] If one or more of the leakage flow rates are greater than the standard leakage flow rate, and one or more of the leakage flow rates are less than or equal to the standard leakage flow rate, then the negative sample is established for the corresponding fluid detection area.

[0018] Furthermore, when determining the current leakage environment according to the increase number, the decrease number, and the negative number, the method includes:

[0019] The leakage environment value is calculated based on the increase quantity, the decrease quantity and the negative quantity, and the leakage environment value is obtained by the following formula:

[0020]

[0021]

[0022] in, represents the total number of samples, To increase the number, To reduce the number, is a negative quantity, 、 、 is the weight coefficient and , is the leakage environment value, and the current leakage environment is determined according to the leakage environment value.

[0023] Furthermore, when comparing the leakage flow rate with the historical leakage flow rate and determining whether to correct the leakage flow rate according to the comparison result, the method includes:

[0024] When there is a leakage flow rate that is smaller than the minimum value of the historical leakage flow rate, determining to correct the leakage flow rate that is smaller than the historical leakage flow rate;

[0025] When there is no leakage flow rate less than the minimum value of the historical leakage flow rate, it is determined that the leakage flow rate is not corrected, and the average of all leakage flow rates is determined as the final leakage flow rate.

[0026] Furthermore, when determining to correct the leakage flow rate, the historical leakage flow rate is determined using a clustering algorithm to determine a historical leakage flow rate data set, including:

[0027] Count all historical leakage flow rates and establish a historical flow rate collection;

[0028] Set the number of clusters k=2, randomly select two historical leakage flow rates in the historical flow rate set as the first cluster center and the second cluster center, calculate the distance from each historical leakage flow rate in the historical flow rate set to the first cluster center and the second cluster center, and assign it to the cluster with the closest distance. If the distances are the same, assign it to the cluster where the first cluster center is located;

[0029] According to the allocation result, the cluster where the center of the first cluster is located is determined as a first flow rate cluster, and the cluster where the center of the second cluster is located is determined as a second flow rate cluster;

[0030] The first flow rate cluster and the second flow rate cluster are determined as the historical leakage flow rate dataset.

[0031] Furthermore, when determining the leakage correction factor based on the similarity between the leakage flow rate and the historical leakage flow rate data set, the method includes:

[0032] Obtaining a leakage flow rate that is less than a minimum value of the historical leakage flow rate and recording it as a flow rate to be corrected;

[0033] Calculating flow velocity similarity based on the mean of the first flow velocity cluster, the mean of the second flow velocity cluster, and the flow velocity to be corrected;

[0034] The flow rate similarity is obtained by the following formula:

[0035] ;

[0036] in represents the flow velocity similarity, Indicates the flow rate to be corrected. represents the mean of the first flow rate cluster, represents the mean of the second flow rate cluster, and is the weight coefficient and , determine the leakage correction factor based on the flow rate similarity.

[0037] Furthermore, when correcting the leakage flow rate according to the leakage correction factor to obtain a final leakage flow rate, the method includes:

[0038] When the flow rate to be corrected exists in all the leakage flow rates, a corrected flow rate is obtained according to the product of the correction factor and the flow rate to be corrected;

[0039] A validation set is established for all corrected flow rates, and a data set is established for the uncorrected leakage flow rate, and a normal distribution model is fitted based on the mean and standard deviation of the data set;

[0040] When the probability of the presence of a corrected flow rate in the normal distribution model in the validation set is less than 0.3, the corrected flow rate is determined to be an erroneous correction and is removed;

[0041] When the probability of all the corrected flow rates in the validation set in the normal distribution model is greater than or equal to 0.3, the validation set is determined to be correct;

[0042] All corrected flow rates in the validation set that are determined to be correct and all uncorrected leakage flow rates in the data set are summed and averaged, and the result based on the average is determined as the final leakage flow rate.

[0043] Furthermore, when determining the leakage type according to the final leakage flow rate, the method includes:

[0044] Comparing the final leakage flow rate with a first preset leakage flow rate and a second preset leakage flow rate, respectively, the first preset leakage flow rate being greater than the second preset leakage flow rate, and determining the leakage type of the leakage point according to the comparison results;

[0045] When the final leakage flow rate is greater than or equal to the first preset leakage flow rate, determining that the leakage type of the leakage point is a serious leakage;

[0046] When the final leakage flow rate is less than the first preset leakage flow rate and greater than or equal to the second preset leakage flow rate, determining that the leakage type of the leakage point is a moderate leakage;

[0047] When the final leakage flow rate is less than the second preset leakage flow rate, it is determined that the leakage type of the leakage point is a slight leakage.

[0048] Furthermore, when determining the leakage interval according to the leakage type and the final leakage flow rate, the method includes:

[0049] Obtaining a historical leakage flow rate of the severe leakage and establishing a first interval data set, obtaining a mean value in the first interval data set to determine a first-level leakage interval;

[0050] Obtaining the historical leakage flow rate of the moderate leakage and establishing a second interval data set, and obtaining the mean value in the second interval data set to determine a secondary leakage interval;

[0051] Obtaining the historical leakage flow rate of the mild leakage and establishing a third interval data set, obtaining the mean value in the third interval data set to determine a third-level leakage interval;

[0052] The leakage interval includes the primary leakage interval, the secondary leakage interval and the tertiary leakage interval;

[0053] The final leakage flow rate is substituted into the leakage interval to determine the result of the quantitative detection.

[0054] Compared with the existing technology, the beneficial effect of the present invention is that by collecting vibration signals through acoustic wave sensors and combining them with voiceprint array technology, the position of the leakage point can be accurately located, avoiding the inefficiency and errors of traditional methods that rely on manual detection or other sensors to conduct large-scale searches. By detecting the leakage flow rate in real time and classifying and counting the leakage flow rate samples, the current environment can be accurately judged. By comparing with the historical leakage flow rate and combining the clustering algorithm to correct the leakage flow rate, the adaptive ability of the quantitative detection process is improved, and it can automatically adjust according to different environmental changes, thereby improving the accuracy and reliability of quantitative detection. According to the correction of the final leakage flow rate and the determination of the leakage type, natural gas leaks can be finely classified, thereby providing a detailed leakage interval division. It helps to carry out effective leakage prevention and control and emergency response, ensuring that appropriate measures can be taken when a leak occurs to reduce the damage caused by natural gas leaks.

[0055] On the other hand, the present application also provides a natural gas leak quantitative detection system based on voiceprint, which is used to apply the above-mentioned natural gas leak quantitative detection method based on voiceprint, including:

[0056] an acquisition unit, configured to acquire vibration signals from the acoustic wave sensor and determine the location of a leakage point based on a soundprint array, divide the leakage point into a plurality of fluid detection areas based on the soundprint array, obtain the fluid leakage flow rate in each fluid detection area, and pre-process the fluid leakage flow rate to obtain the leakage flow rate;

[0057] a leakage environment unit, configured to construct a leakage flow rate sample for a corresponding fluid detection area according to the leakage flow rate, wherein the leakage flow rate sample includes an increasing sample, a decreasing sample, and a negative sample;

[0058] Counting the number of the increasing samples and recording it as an increasing number, counting the number of the decreasing samples and recording it as a decreasing number, counting the number of the negative samples and recording it as a negative number, and determining a current leakage environment according to the increasing number, the decreasing number, and the negative number;

[0059] A judgment unit, configured to obtain a historical leakage flow rate of the same leakage environment as the current leakage environment, compare the leakage flow rate with the historical leakage flow rate, and determine whether to correct the leakage flow rate according to the comparison result;

[0060] a correction unit, when determining to correct the leakage flow rate, applying a clustering algorithm to the historical leakage flow rate to determine a historical leakage flow rate data set, determining a leakage correction factor based on a similarity between the leakage flow rate and the historical leakage flow rate data set, and correcting the leakage flow rate based on the leakage correction factor to obtain a final leakage flow rate;

[0061] The type interval determining unit is configured to determine the leakage type according to the final leakage flow rate, and determine the leakage interval according to the leakage type and the final leakage flow rate.

[0062] It is understandable that the above-mentioned natural gas leakage quantitative detection method and system based on voiceprint have the same beneficial effects, which will not be described in detail here. BRIEF DESCRIPTION OF THE DRAWINGS

[0063] Various other advantages and benefits will become apparent to those skilled in the art upon reading the detailed description of the preferred embodiment below. The accompanying drawings are for illustration purposes only and are not to be considered as limiting the present invention. The same reference symbols are used throughout the drawings to represent the same components. In the drawings:

[0064] Figure 1 A flow chart of a method for quantitatively detecting natural gas leaks based on voiceprints provided in an embodiment of the present invention;

[0065] Figure 2 This is a structural block diagram of a natural gas leak quantitative detection system based on voiceprint provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0066] Exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although exemplary embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the present disclosure to those skilled in the art. It should be noted that, unless there is a conflict, the embodiments of the present disclosure and the features in the embodiments can be combined with each other. The present invention will be described in detail below with reference to the accompanying drawings and in conjunction with the embodiments.

[0067] Existing quantitative natural gas leak detection technologies mostly rely on visual inspection, gas sensors, and infrared detection. While these technologies can detect leaks, they often fail to provide accurate data on the amount of leakage. This is especially true in complex environments, where quantitative assessment of leakage based on actual conditions is difficult. This makes existing detection methods unable to provide accurate data support for leak assessment and emergency response, and it makes it difficult to provide a scientific and reliable basis for decision makers.

[0068] On the one hand, although existing natural gas leak detection technology can detect the existence of leaks, it cannot provide accurate leakage volume data. Especially in some specific environments, the leakage volume data obtained will deviate according to the actual environment. This makes it impossible for existing detection methods to quantitatively detect the leakage volume flow rate and it is difficult to provide a scientific and reliable basis.

[0069] On the other hand, traditional leak monitoring methods struggle to accurately classify leak types and are unable to distinguish appropriate levels based on leak volume and flow rate, resulting in low accuracy and reliability of leak measurement results. For example, traditional methods cannot accurately determine the type of leak with a low flow rate, resulting in false alarms. For more severe leaks, however, timely response is delayed, missing the optimal opportunity for emergency response.

[0070] Therefore, it is necessary to design a natural gas leak quantitative detection method based on voiceprint to solve the problems existing in current technology.

[0071] In some embodiments of the present application, see Figure 1 As shown, a natural gas leak quantitative detection method based on voiceprint includes:

[0072] S100: collecting vibration signals from the acoustic wave sensor and determining the location of the leakage point based on the soundprint array, dividing the leakage point into multiple fluid detection areas based on the soundprint array, obtaining the fluid leakage flow rate in each fluid detection area, and pre-processing the fluid leakage flow rate to obtain the leakage flow rate;

[0073] S200: constructing a leakage flow rate sample for the corresponding fluid detection area according to the leakage flow rate, wherein the leakage flow rate sample includes an increasing sample, a decreasing sample and a negative sample;

[0074] S300: Counting the number of increasing samples and recording it as the increasing number, counting the number of decreasing samples and recording it as the decreasing number, and counting the number of negative samples and recording it as the negative number, and determining the current leakage environment based on the increasing number, decreasing number, and negative number;

[0075] S400: Obtaining a historical leakage flow rate for the same leakage environment as the current leakage environment, comparing the leakage flow rate with the historical leakage flow rate, and determining whether to correct the leakage flow rate based on the comparison result;

[0076] S500: When it is determined that the leakage flow rate needs to be corrected, a clustering algorithm is used to determine a historical leakage flow rate data set based on the historical leakage flow rate, a leakage correction factor is determined based on similarity between the leakage flow rate and the historical leakage flow rate data set, and the leakage flow rate is corrected based on the leakage correction factor to obtain a final leakage flow rate;

[0077] S600: Determine the leakage type according to the final leakage flow rate, and determine the leakage range according to the leakage type and the final leakage flow rate.

[0078] Specifically, acoustic wave sensors sense the vibrations generated by natural gas pipeline leaks, thereby acquiring location information related to the leak point, improving detection accuracy. Furthermore, voiceprint array analysis can pinpoint the exact location of small leaks and divide the location into multiple fluid detection zones. Using the soundprint propagation time difference, each fluid detection zone can be monitored and the leakage flow rate can be obtained. The leakage flow rate is then pre-processed to remove erroneous flow rates and environmental noise that may contribute to the leakage flow rate, improving its accuracy and providing an accurate basis for subsequent leak analysis and assessment. Leakage flow rate samples are categorized as increasing, decreasing, and negative based on their changing trends. Increasing samples represent a high leakage flow rate, decreasing samples represent a low leakage flow rate, and negative samples indicate a potential impact on the leakage flow rate samples being processed. By counting these samples, the current leak status can be determined. The number of increasing, decreasing, and negative samples, recorded as increasing, decreasing, and negative, respectively, will serve as baseline data for leak analysis. After the leakage environment is determined, historical leakage flow rates that are identical to the current leakage environment are collected. The historical leakage flow rates provide a reference benchmark for the leakage flow rate. The comparison process ensures that the current leakage flow rate accurately reflects the actual leakage situation, avoiding quantitative detection errors caused by environmental changes or sensor errors. When it is determined that the leakage flow rate needs to be corrected, the historical leakage flow rates are aggregated through a clustering algorithm to avoid large differences between the historical leakage flow rates obtained. The leakage correction factor is determined based on the historical leakage flow rate data set. The leakage correction factor is a factor used to adjust the leakage flow rate to ensure the accuracy of the final leakage flow rate and provide data support for subsequent leakage type analysis. The leak type is determined and combined with the final leakage flow rate to determine the leakage range, ensuring the scientific nature and reliability of quantitative leakage detection.

[0079] It is understandable that by comprehensively using voiceprint arrays, fluid detection technology, sample analysis and historical data comparison for correction, accurate detection of natural gas leakage flow rate is achieved. It can not only detect leakage, but also dynamically correct the leakage flow rate according to specific circumstances, thereby providing reliable technical support for the safety monitoring and risk assessment of natural gas leakage.

[0080] In some embodiments of the present application, when constructing a leakage flow rate sample for the corresponding fluid detection area based on the leakage flow rate, it includes: counting all leakage flow rates and determining the standard leakage flow rate corresponding to each fluid detection area, comparing the leakage flow rate with the standard leakage flow rate, if all leakage flow rates are greater than the standard leakage flow rate, then establishing an increase sample for the corresponding fluid detection area; if all leakage flow rates are less than or equal to the standard leakage flow rate, then establishing a decrease sample for the corresponding fluid detection area; if there is one or more leakage flow rates greater than the standard leakage flow rate, and there is one or more leakage flow rates less than or equal to the standard leakage flow rate, then establishing a negative sample for the corresponding fluid detection area.

[0081] It's understood that the standard leakage flow rate is the industry standard. Based on the comparison results, corresponding samples are created for each fluid detection area. Increasing samples indicate that the leakage flow rate in that fluid detection area is greater than the standard flow rate, indicating a relatively high leakage level. Decreasing samples indicate that the leakage flow rate in that fluid detection area is less than or equal to the standard flow rate, indicating a relatively low leakage level. Negative samples indicate that the leakage level is relatively low and may not be expanding or is decreasing. Real-time statistics and comparisons of leakage flow rates enhance the scientific nature of identifying changing trends in natural gas leaks, avoid subjective bias in human judgment, and improve the accuracy of accurately identifying the current leakage environment.

[0082] In some embodiments of the present application, when determining the current leakage environment based on the increase number, the decrease number, and the negative number, the method includes: calculating a leakage environment value based on the increase number, the decrease number, and the negative number, and the leakage environment value is obtained by the following formula:

[0083]

[0084]

[0085] in, represents the total number of samples, To increase the number, To reduce the number, is a negative quantity, 、 、 is the weight coefficient and , The leakage environment value is used to determine the current leakage environment.

[0086] It is understandable that the first preset leakage environment, the second preset leakage environment and the third preset leakage environment are preset. ≤0.8, the first preset leakage environment is used as the current leakage environment. When ≤1, the second preset leakage environment is used as the current leakage environment. When 1<n, the third preset leakage environment is used as the current leakage environment. The leakage environment value can be accurately determined according to the increase number, decrease number and negative number, and according to the corresponding preset leakage environment, the current leakage environment can be accurately judged, laying the foundation for subsequent corrections and adjustments.

[0087] In some embodiments of the present application, when comparing the leakage flow rate with the historical leakage flow rate and determining whether to correct the leakage flow rate based on the comparison result, it includes: when there is a leakage flow rate that is less than the minimum value of the historical leakage flow rate, determining to correct the leakage flow rate that is less than the historical leakage flow rate; when there is no leakage flow rate that is less than the minimum value of the historical leakage flow rate, determining not to correct the leakage flow rate, and taking the average of all the leakage flow rates to determine the final leakage flow rate.

[0088] It is understandable that the historical leakage flow rate provides a reference benchmark, representing a reasonable range of all leakage flow rates that occur. By comparing with the historical leakage flow rate, abnormal situations can be effectively identified. For example: the leakage flow rate is abnormally high or low. This situation may be caused by human detection operation errors and fluctuations in environmental conditions. The historical leakage flow rate is the result of long-term accumulation in actual use, and can reflect the actual performance of the leakage flow rate. Comparison with the historical leakage flow rate avoids the subsequent accumulation of errors and causes errors in quantitative detection results. When there is a leakage flow rate that is less than the minimum value of the historical leakage flow rate, it means that the measured leakage flow rate has a large deviation and needs to be corrected to the correct flow rate range. Taking the average of all leakage flow rates avoids accidental errors in the obtained leakage flow rate and improves the accuracy of quantitative detection.

[0089] In some embodiments of the present application, when it is determined that the leakage flow rate should be corrected, the historical leakage flow rate is determined using a clustering algorithm to determine a historical leakage flow rate data set, including: counting all historical leakage flow rates and establishing a historical flow rate set, setting the cluster number k=2, randomly selecting two historical leakage flow rates in the historical flow rate set as the first cluster center and the second cluster center, calculating the distance to the first cluster center and the second cluster center for each historical leakage flow rate in the historical flow rate set, and assigning them to the cluster with the closest distance. When the distances are the same, they are assigned to the cluster where the first cluster center is located. According to the allocation result, the cluster where the first cluster center is located is determined as the first flow rate cluster, and the cluster where the second cluster center is located is determined as the second flow rate cluster. The first flow rate cluster and the second flow rate cluster are determined as the historical leakage flow rate data set.

[0090] It can be understood that the clustering algorithm can automatically divide the historical leakage flow rate and accurately obtain the data distribution in the historical leakage flow rate, so as to accurately judge the difference between the leakage flow rate and the historical leakage flow rate. The clustering algorithm simplifies the historical leakage flow rate and processes the data of each cluster separately, avoiding the errors caused by relying solely on judgment.

[0091] In some embodiments of the present application, when determining a leakage correction factor based on similarity between the leakage flow rate and a historical leakage flow rate dataset, the method includes: obtaining a leakage flow rate that is less than the minimum value of the historical leakage flow rate and recording it as the flow rate to be corrected; and calculating the flow rate similarity based on the mean of the first flow rate cluster, the mean of the second flow rate cluster, and the flow rate to be corrected. The flow rate similarity is calculated using the following formula:

[0092] ;

[0093] in represents the flow velocity similarity, Indicates the flow rate to be corrected. represents the mean of the first flow rate cluster, represents the mean of the second flow rate cluster, and is the weight coefficient and , determine the leakage correction factor based on the flow rate similarity.

[0094] It is understandable that when When it is greater than or equal to 0.6, the first preset leakage correction factor is used as the correction factor of the flow rate to be corrected. When it is less than 0.6, the second preset leakage correction factor is used as the correction factor for the flow rate to be corrected. The first preset leakage correction factor is greater than the second preset leakage correction factor. The flow rate similarity shows the similarity with the historical leakage flow rate. The different historical leakage flow rates are combined, and the corresponding preset leakage correction factor is selected according to the flow rate similarity, thereby realizing dynamic optimization of the flow rate to be corrected and ensuring the accuracy of quantitative detection.

[0095] In some embodiments of the present application, when the leakage flow rate is corrected according to the leakage correction factor to obtain the final leakage flow rate, it includes: when there is a flow rate to be corrected among all the leakage flow rates, the corrected flow rate is obtained according to the product of the correction factor and the flow rate to be corrected, a verification set is established for all the corrected flow rates, and a data set is established for the leakage flow rates that have not been corrected, a normal distribution model is fitted according to the mean and standard deviation of the data set, when the probability of the existence of a corrected flow rate in the normal distribution model in the verification set is less than 0.3, the corrected flow rate is determined to be an erroneous correction and is removed, when the probability of all the corrected flow rates in the verification set in the normal distribution model is greater than or equal to 0.3, the verification set is determined to be correct, all the corrected flow rates in the correct verification set and all the uncorrected leakage flow rates in the data set are summed and averaged, and the result based on the mean is determined as the final leakage flow rate.

[0096] It is understandable that by introducing a correction factor to correct the flow rate to be corrected, the accuracy of the final leakage flow rate can be effectively improved. By verifying the corrected flow rate based on the normal distribution model, the probability of the normal distribution model can be appropriately adjusted to ensure that the corrected flow rate conforms to the normal leakage flow rate distribution law, ensuring the rationality and accuracy of the corrected flow rate, thereby improving the accuracy of the verification set and reducing the interference of abnormal conditions on the final leakage flow rate. All corrected flow rates in the correct verification set are added to all uncorrected leakage flow rates and the average is taken, which improves the comprehensiveness of quantitative detection and prevents the loss of a certain data from interfering with the accuracy of obtaining the final leakage flow rate.

[0097] In some embodiments of the present application, when determining the leakage type based on the final leakage flow rate, it includes: comparing the final leakage flow rate with a first preset leakage flow rate and a second preset leakage flow rate, respectively, the first preset leakage flow rate is greater than the second preset leakage flow rate, and determining the leakage type of the leakage point according to the comparison result; when the final leakage flow rate is greater than or equal to the first preset leakage flow rate, the leakage type of the leakage point is determined to be a severe leakage; when the final leakage flow rate is less than the first preset leakage flow rate and greater than or equal to the second preset leakage, the leakage type of the leakage point is determined to be a moderate leakage; when the final leakage flow rate is less than the second preset leakage flow rate, the leakage type of the leakage point is determined to be a mild leakage.

[0098] It can be understood that by comparing the final leakage flow rate with the two preset leakage flow rates, the leakage type of the leakage point can be determined, which reduces the deviation of human judgment and clearly shows the situation of natural gas leakage, improves the efficiency of quantitative leakage detection, and provides an accurate scientific basis for emergency treatment of natural gas leaks.

[0099] In some embodiments of the present application, when determining the leakage interval based on the leakage type and the final leakage flow rate, it includes: obtaining the historical leakage flow rate of severe leakage and establishing a first interval data set, obtaining the mean value in the first interval data set to determine the first-level leakage interval; obtaining the historical leakage flow rate of moderate leakage and establishing a second interval data set, obtaining the mean value in the second interval data set to determine the second-level leakage interval; obtaining the historical leakage flow rate of mild leakage and establishing a third interval data set, obtaining the mean value in the third interval data set to determine the third-level leakage interval, the leakage interval includes a first-level leakage interval, a second-level leakage interval and a third-level leakage interval, and the result of quantitative detection is determined by substituting the final leakage flow rate into the leakage interval.

[0100] It can be understood that the minimum value of the first-level leakage interval is the mean of the first interval data set, and the maximum value is four times its mean; the minimum value of the second-level leakage interval is the mean of the second interval data set, and the maximum value is the mean of the first interval data set; the minimum value of the third-level leakage interval is the mean of the third interval data set, and the maximum value is the mean of the second interval data set. Substituting the final leakage flow rate into the leakage interval can clearly determine the level of leakage, providing an accurate and scientific basis for the quantitative detection of natural gas leaks. The standardized processing method of quantitative detection improves the degree of automation of natural gas leak detection, reduces manual intervention and subjective judgment, and provides scientific, accurate and efficient support for natural gas leak monitoring and emergency response.

[0101] In summary, the beneficial effects of the present invention are as follows: by collecting vibration signals through acoustic wave sensors and combining them with voiceprint array technology, the position of the leakage point can be accurately located, avoiding the inefficiency and errors of traditional methods that rely on manual detection or other sensors for large-scale searches. By detecting the leakage flow rate in real time and classifying and counting the leakage flow rate samples, the current environment can be accurately judged. By comparing with the historical leakage flow rate and combining the clustering algorithm to correct the leakage flow rate, the adaptive ability of the quantitative detection process is improved, and it can be automatically adjusted according to different environmental changes, thereby improving the accuracy and reliability of quantitative detection. According to the correction of the final leakage flow rate and the determination of the leakage type, natural gas leaks can be finely classified, which is conducive to effective leakage prevention and control and emergency response, ensuring that appropriate measures can be taken when a leak occurs to reduce the damage caused by natural gas leaks.

[0102] In another preferred embodiment based on the above embodiment, refer to Figure 2 As shown, this embodiment provides a natural gas leak quantitative detection system based on voiceprint, which is used to apply the above-mentioned natural gas leak quantitative detection method based on voiceprint, including:

[0103] An acquisition unit, configured to acquire the vibration signal of the acoustic wave sensor and determine the location of the leakage point based on the soundprint array, divide the leakage point into multiple fluid detection areas based on the soundprint array, obtain the fluid leakage flow rate in each fluid detection area, and pre-process the fluid leakage flow rate to obtain the leakage flow rate;

[0104] A leakage environment unit is used to construct a leakage flow rate sample for the corresponding fluid detection area according to the leakage flow rate, and the leakage flow rate sample includes an increasing sample, a decreasing sample and a negative sample;

[0105] Count the number of increasing samples and record it as the increasing number, count the number of decreasing samples and record it as the decreasing number, count the number of negative samples and record it as the negative number, and determine the current leakage environment based on the increasing number, decreasing number, and negative number;

[0106] A judgment unit, configured to obtain a historical leakage flow rate of the same leakage environment as the current leakage environment, compare the leakage flow rate with the historical leakage flow rate, and determine whether to correct the leakage flow rate based on the comparison result;

[0107] a correction unit, when determining to correct the leakage flow rate, applying a clustering algorithm to the historical leakage flow rate to determine a historical leakage flow rate data set, determining a leakage correction factor based on similarity between the leakage flow rate and the historical leakage flow rate data set, and correcting the leakage flow rate based on the leakage correction factor to obtain a final leakage flow rate;

[0108] The type interval determination unit is used to determine the leakage type according to the final leakage flow rate, and determine the leakage interval according to the leakage type and the final leakage flow rate.

[0109] It is understandable that through the coordinated work of multiple units, efficient and accurate leak detection and correction of pipeline natural gas leaks are achieved. The acquisition unit uses acoustic wave sensors and voiceprint arrays to locate the leak point, and obtains the leakage flow rate through preprocessing, ensuring the accuracy of data collection. The leakage environment unit generates different types of samples based on the leakage flow rate, and analyzes the current leakage environment through statistical quantities to provide a basis for subsequent judgment. The judgment unit determines whether the leakage flow rate needs to be corrected by comparing it with the historical leakage flow rate, further enhancing the reliability of quantitative detection. The correction unit uses a clustering algorithm to aggregate historical leakage flow rates and adjust the leakage flow rate to ensure the accuracy of the final leakage flow rate. The type interval determination unit divides the leakage type and interval according to the final leakage flow rate, improving the accuracy and scientificity of quantitative detection of natural gas leaks.

[0110] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware. Furthermore, the present application may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0111] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems) and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0112] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0113] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0114] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents. Any modification or equivalent replacement that does not depart from the spirit and scope of the present invention should be covered by the scope of protection of the claims of the present invention.

Claims

1. A natural gas leakage quantitative detection method based on voiceprint, characterized in that: include: Collecting the vibration signal of the acoustic sensor and determining the location of the leakage point according to the soundprint array, dividing the leakage point into a plurality of fluid detection areas based on the soundprint array, obtaining the fluid leakage flow rate of each fluid detection area, and pre-processing the fluid leakage flow rate to obtain the leakage flow rate; Constructing a leakage flow rate sample for the corresponding fluid detection area according to the leakage flow rate, wherein the leakage flow rate sample includes an increase sample, a decrease sample and a negative sample; Counting the number of the increasing samples and recording it as the increasing number, counting the number of the decreasing samples and recording it as the decreasing number, counting the number of the negative samples and recording it as the negative number, and determining the current leakage environment according to the increasing number, the decreasing number and the negative number; Obtaining a historical leakage flow rate of the same leakage environment as the current leakage environment, comparing the leakage flow rate with the historical leakage flow rate, and determining whether to correct the leakage flow rate according to the comparison result; When it is determined that the leakage flow rate is to be corrected, a historical leakage flow rate data set is determined by using a clustering algorithm for the historical leakage flow rate, a leakage correction factor is determined according to similarity between the leakage flow rate and the historical leakage flow rate data set, and the leakage flow rate is corrected according to the leakage correction factor to obtain a final leakage flow rate; The leakage type is determined according to the final leakage flow rate, and the leakage interval is determined according to the leakage type and the final leakage flow rate.

2. The method for quantitative detection of natural gas leakage based on voiceprint according to claim 1 is characterized in that: When constructing a leakage flow rate sample for a corresponding fluid detection area according to the leakage flow rate, it includes: Counting all the leakage flow rates and determining a standard leakage flow rate corresponding to each fluid detection area; Comparing the leakage flow rate with the standard leakage flow rate, if all the leakage flow rates are greater than the standard leakage flow rate, establishing the growth sample for the corresponding fluid detection area; If all leakage flow rates are less than or equal to the standard leakage flow rate, then the reduced sample is established for the corresponding fluid detection area; If there is one or more leakage flow rates greater than the standard leakage flow rate, and there is one or more leakage flow rates less than or equal to the standard leakage flow rate, then the negative sample is established for the corresponding fluid detection area.

3. The method for quantitative detection of natural gas leakage based on voiceprint according to claim 1 is characterized in that: When determining the current leakage environment according to the increase number, the decrease number and the negative number, it includes: The leakage environment value is calculated according to the increase quantity, the decrease quantity and the negative quantity, and the leakage environment value is obtained by the following formula: ； ； in, represents the total number of samples, To increase the number, To reduce the number, is a negative quantity, , , is the weight coefficient and , is a leakage environment value, and the current leakage environment is determined according to the leakage environment value.

4. The method for quantitative detection of natural gas leakage based on voiceprint according to claim 1 is characterized in that: When comparing the leakage flow rate with the historical leakage flow rate and determining whether to correct the leakage flow rate according to the comparison result, the method includes: When there is a leakage flow rate that is less than the minimum value of the historical leakage flow rate, it is determined to correct the leakage flow rate that is less than the historical leakage flow rate; When there is no leakage flow rate less than the minimum value of the historical leakage flow rate, it is determined that the leakage flow rate is not to be corrected, and the average of all leakage flow rates is determined as the final leakage flow rate.

5. The method for quantitative detection of natural gas leakage based on voiceprint according to claim 4 is characterized in that: When it is determined that the leakage flow rate is to be corrected, the historical leakage flow rate is determined by using a clustering algorithm to determine a historical leakage flow rate data set, including: Count all historical leakage flow rates and establish a historical flow rate collection; Set the cluster number k=2, randomly select two historical leakage flow rates in the historical flow rate set as the first cluster center and the second cluster center, calculate the distance from each historical leakage flow rate in the historical flow rate set to the first cluster center and the second cluster center, and assign it to the cluster with the closest distance. If the distances are the same, assign it to the cluster where the first cluster center is located. According to the allocation result, the cluster where the center of the first cluster is located is determined as a first flow rate cluster, and the cluster where the center of the second cluster is located is determined as a second flow rate cluster; The first flow rate cluster and the second flow rate cluster are determined as the historical leakage flow rate dataset.

6. The method for quantitative detection of natural gas leakage based on voiceprint according to claim 5 is characterized in that: When determining the leakage correction factor according to the similarity results between the leakage flow rate and the historical leakage flow rate data set, it includes: Obtaining a leakage flow rate that is less than the minimum value of the historical leakage flow rate and recording it as the flow rate to be corrected; Calculating the flow velocity similarity according to the mean of the first flow velocity cluster, the mean of the second flow velocity cluster and the flow velocity to be corrected; The flow rate similarity is obtained by the following formula: ； in represents the flow velocity similarity, Indicates the flow rate to be corrected. represents the mean of the first flow rate cluster, represents the mean of the second flow rate cluster, and is the weight coefficient and , determine the leakage correction factor based on the flow velocity similarity.

7. The method for quantitative detection of natural gas leakage based on voiceprint according to claim 6 is characterized in that: When the leakage flow rate is corrected according to the leakage correction factor to obtain a final leakage flow rate, the method includes: When the flow rate to be corrected exists in all the leakage flow rates, a corrected flow rate is obtained according to the product value of the correction factor and the flow rate to be corrected; A validation set is established for all corrected flow rates, a data set is established for the uncorrected leakage flow rate, and a normal distribution model is fitted according to the mean and standard deviation of the data set; When the probability that the corrected flow rate exists in the normal distribution model in the validation set is less than 0.3, the corrected flow rate is determined to be an erroneous correction and is removed; When the probability of all the corrected flow velocities in the verification set in the normal distribution model is greater than or equal to 0.3, the verification set is determined to be correct; All corrected flow rates in the validation set that are determined to be correct and all uncorrected leakage flow rates in the data set are summed and averaged, and the result based on the average is determined as the final leakage flow rate.

8. The method for quantitative detection of natural gas leakage based on voiceprint according to claim 1, characterized in that: When determining the leakage type according to the final leakage flow rate, it includes: Comparing the final leakage flow rate with a first preset leakage flow rate and a second preset leakage flow rate respectively, the first preset leakage flow rate being greater than the second preset leakage flow rate, and determining the leakage type of the leakage point according to the comparison result; When the final leakage flow rate is greater than or equal to the first preset leakage flow rate, determining that the leakage type of the leakage point is a serious leakage; When the final leakage flow rate is less than the first preset leakage flow rate and greater than or equal to the second preset leakage flow rate, determining that the leakage type of the leakage point is a moderate leakage; When the final leakage flow rate is less than the second preset leakage flow rate, it is determined that the leakage type of the leakage point is a slight leakage.

9. The method for quantitative detection of natural gas leakage based on voiceprint according to claim 8, characterized in that: When determining the leakage interval according to the leakage type and the final leakage flow rate, it includes: Obtaining the historical leakage flow rate of the serious leakage and establishing a first interval data set, and obtaining the mean value in the first interval data set to determine a primary leakage interval; Obtaining the historical leakage flow rate of the moderate leakage and establishing a second interval data set, and obtaining the mean value in the second interval data set to determine the secondary leakage interval; Obtain the historical leakage flow rate of the slight leakage and establish a third interval data set, and obtain the mean value in the third interval data set to determine the third-level leakage interval; The leakage interval includes the primary leakage interval, the secondary leakage interval and the tertiary leakage interval; The result of the quantitative detection is determined by substituting the final leakage flow rate into the leakage interval.

10. A natural gas leakage quantitative detection system based on voiceprint, used for applying the natural gas leakage quantitative detection method based on voiceprint according to any one of claims 1 to 9, characterized in that: include: A collection unit, used to collect the vibration signal of the acoustic wave sensor and determine the location of the leakage point according to the soundprint array, divide the leakage point into multiple fluid detection areas based on the soundprint array, obtain the fluid leakage flow rate of each fluid detection area, and pre-process the fluid leakage flow rate to obtain the leakage flow rate; A leakage environment unit, used for constructing a leakage flow rate sample for a corresponding fluid detection area according to the leakage flow rate, wherein the leakage flow rate sample includes an increase sample, a decrease sample and a negative sample; Counting the number of the increasing samples and recording it as the increasing number, counting the number of the decreasing samples and recording it as the decreasing number, counting the number of the negative samples and recording it as the negative number, and determining the current leakage environment according to the increasing number, the decreasing number and the negative number; A judgment unit, used for obtaining a historical leakage flow rate of the same leakage environment as the current leakage environment, comparing the leakage flow rate with the historical leakage flow rate, and judging whether to correct the leakage flow rate according to the comparison result; a correction unit, when determining to correct the leakage flow rate, using a clustering algorithm to determine a historical leakage flow rate data set for the historical leakage flow rate, determining a leakage correction factor according to a similarity result between the leakage flow rate and the historical leakage flow rate data set, and correcting the leakage flow rate according to the leakage correction factor to obtain a final leakage flow rate; The type interval determination unit is used to determine the leakage type according to the final leakage flow rate, and determine the leakage interval according to the leakage type and the final leakage flow rate.

Images