Method, device and storage medium for diagnosing health status of ultrasonic flowmeter
By obtaining the health status parameters of the ultrasonic flowmeter and analyzing its characteristic indicators, combining the K-means clustering algorithm and the maximum mutual information coefficient, the error problem of diagnosing the health status of the ultrasonic flowmeter in the existing technology is solved, and a more comprehensive and accurate health status monitoring is achieved.
Patent Information
- Application Number
- CN202210638407.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-06-07
- Publication Date
- 2025-05-30
- Estimated Expiration
- 2042-06-07
AI Technical Summary
In the prior art, when diagnosing the healthy state of ultrasonic flowmeters, relying on a single or a few indicators, errors are easily generated, and threshold settings are affected by subjective factors.
By obtaining the health status parameters of the ultrasonic flowmeter, a set of characteristic indicators is determined, including distortion degree, distribution characteristics, complexity and periodic indicators, and a K-means clustering algorithm is used to cluster the basic data set with the maximum mutual information coefficient to determine whether there are abnormal operating conditions.
It realizes a comprehensive and accurate diagnosis of the health status of the ultrasonic flowmeter, reduces the possibility of false alarms, and improves the comprehensiveness and accuracy of monitoring.
Smart Images
Figure CN115099312B_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of ultrasonic flow technology, and particularly to a method, device and storage medium for diagnosing the health status of an ultrasonic flowmeter. Background Art
[0002] Due to its working mechanism, an ultrasonic flowmeter has high requirements for the working health status of each component. In the process of calculating the sound path or sound delay, extremely small calculation deviations will lead to serious flow calculation error problems, seriously affecting the use of the ultrasonic flowmeter. Thus, achieving a complete, comprehensive and reliable diagnosis of the health status of the ultrasonic flowmeter is the key to ensuring its normal operation.
[0003] Currently, the common practice at home and abroad is as follows: Set a fixed sampling time to obtain the measured target time series. When the target time series are characteristics such as gas velocity and sound velocity, calculate the deviation between these indicators and the theoretical values, and take the maximum value or average value of the calculation results; when the target time series are characteristics such as vortex angle and asymmetry coefficient that do not have theoretical values, calculate their average value or maximum value. Further, compare these calculation results with the thresholds set artificially. When the calculated values of these evaluation indicators exceed the thresholds, it is determined as an abnormal working condition.
[0004] In summary, the above method for evaluating the health status of an ultrasonic flowmeter has the problem of one-sidedly relying on one or a few indicators. Moreover, setting the thresholds of the detection parameters of the ultrasonic flowmeter artificially is affected by subjective factors and is prone to errors. Summary of the Invention
[0005] The purpose of the embodiments of this application is to provide a method, device and storage medium for diagnosing the health status of an ultrasonic flowmeter, so as to solve the problem in the prior art of one-sidedly relying on one or a few indicators to evaluate the health status of the ultrasonic flowmeter.
[0006] To achieve the above purpose, in the first aspect of the embodiments of this application, a method for diagnosing the health status of an ultrasonic flowmeter is provided, and the method includes:
[0007] Obtain the health status parameters of the ultrasonic flowmeter;
[0008] Determine a set of characteristic indicators of the health status parameters according to the health status parameters;
[0009] Determine a basic data set according to the set of characteristic indicators;
[0010] Determine whether there is an abnormal working condition according to the basic data set.
[0011] In the embodiments of this application, the health status parameters may at least include one of the following:
[0012] Sampling rate, signal gain, signal-to-noise ratio, swirl angle, measured gas velocity of each channel, and measured flow velocity of each channel.
[0013] In an embodiment of the present application, the set of characteristic indicators may include:
[0014] Degree-of-distortion indicator, distribution characteristic indicator, complexity indicator, periodicity indicator.
[0015] In an embodiment of the present application, the distribution characteristic indicator includes kurtosis and skewness, and the kurtosis satisfies formula (1):
[0016]
[0017] where S 2 is the kurtosis, x i is the value of the i-th measurement point in the time series of health state parameters, is the mean value of all measurement points in the time series of health state parameters, std is the standard normal distribution, and n is the total number of measurement points;
[0018] The skewness satisfies formula (2):
[0019]
[0020] where S 3 is the skewness, is the mean value of all measurement points in the time series of health state parameters, and std is the standard normal distribution.
[0021] In an embodiment of the present application, the complexity indicator includes complexity and spectral statistical variance, and the complexity satisfies formula (3):
[0022]
[0023] where S 4 is the complexity, n is the total number of measurement points, x i is the value of the i-th measurement point in the time series of health state parameters, x i-1 is the value of the (i - 1)-th measurement point in the time series of health state parameters;
[0024] The spectral statistical variance satisfies formula (4):
[0025]
[0026] where S 5 is the spectral statistical variance, n is the total number of measurement points, k is the frequency of the sub-signal in the time series of health state parameters, X(k) is the amplitude corresponding to the frequency k after Fourier transform, is the mean value of the amplitudes of the frequencies of the time series of health state parameters after Fourier transform.
[0027] In an embodiment of the present application, the periodic index satisfies formula (5):
[0028]
[0029] Wherein, S 6 is the autoregressive coefficient, n is the total number of measurement points, t is the lag coefficient, std is the standard normal distribution, and x i is the value of the i-th measurement point in the time series of health state parameters, is the mean value of all measurement points in the time series of health state parameters, and x i+t is the value of the (i + t)-th measurement point in the time series of health state parameters.
[0030] In an embodiment of the present application, determining whether there is an abnormal working condition according to the basic data set includes:
[0031] Obtain the basic data set;
[0032] Cluster the basic data set according to the K-means clustering algorithm to determine the clustering result;
[0033] Determine the abnormal working condition according to the clustering result.
[0034] In an embodiment of the present application, clustering the basic data set according to the K-means clustering algorithm includes:
[0035] Cluster the basic data set in combination with the maximum mutual information coefficient.
[0036] In an embodiment of the present application, the maximum mutual information coefficient satisfies formula (6):
[0037]
[0038] Wherein, x is any point in the basic data set, y is another arbitrary point in the basic data set, a is the number of grids divided on the x-axis, b is the number of grids divided on the y-axis, B is a variable, the value of which is the 0.6th power of the data volume of the basic data set, and I(x, y) is the mutual information.
[0039] In an embodiment of the present application, it further includes:
[0040] Determine the alarm threshold according to the basic data set.
[0041] The second aspect of the present application provides a device for diagnosing the health state of an ultrasonic flowmeter, including:
[0042] A memory configured to store instructions; and
[0043] A processor, configured to call instructions from a memory and capable of implementing the method for diagnosing the health status of an ultrasonic flowmeter according to the above when executing the instructions.
[0044] A third aspect of the present application provides a machine-readable storage medium, on which instructions are stored, and the instructions are used to cause a machine to execute the method for diagnosing the health status of an ultrasonic flowmeter according to the above.
[0045] Through the above technical solutions, health status parameters of the ultrasonic flowmeter are obtained, and then a set of characteristic indicators of the health status parameters is determined according to the health status parameters. Thereby, a basic data set is determined according to the set of characteristic indicators, and finally, whether there is an abnormal working condition is determined according to the basic data set. On the one hand, this method promotes the basic research on the health status diagnosis method of the ultrasonic flowmeter. On the other hand, because it introduces intelligent and data-based algorithms and introduces historical data into real-time detection, it is beneficial to accurately evaluate the health status of the ultrasonic flowmeter. The present application reduces the possibility of false alarms about the health status of the ultrasonic flowmeter by comprehensively analyzing the health status parameters of the ultrasonic flowmeter, and improves the comprehensiveness and accuracy of the health status monitoring of the ultrasonic flowmeter.
[0046] Other features and advantages of the embodiments of the present application will be described in detail in the subsequent specific implementation part. Description of the Drawings
[0047] The drawings are used to provide a further understanding of the embodiments of the present application, and constitute a part of the specification. Together with the following specific implementation, they are used to explain the embodiments of the present application, but do not constitute a limitation to the embodiments of the present application. In the drawings:
[0048] Figure 1 Schematically shows a flowchart of a method for diagnosing the health status of an ultrasonic flowmeter according to an embodiment of the present application;
[0049] Figure 2 Schematically shows a schematic diagram of a distortion degree index and a distribution characteristic index according to an embodiment of the present application;
[0050] Figure 3 Schematically shows a schematic diagram of a complexity index according to an embodiment of the present application;
[0051] Figure 4 Schematically shows a schematic diagram of a periodicity index according to an embodiment of the present application;
[0052] Figure 5 Schematically shows a schematic diagram of clustering a basic data set according to an embodiment of the present application;
[0053] Figure 6Schematically shows a schematic diagram for determining an alarm threshold according to an embodiment of the present application;
[0054] Figure 7 Schematically shows a structural schematic diagram of a device for diagnosing the health state of an ultrasonic flowmeter according to an embodiment of the present application. Detailed implementation manners
[0055] To make the objectives, technical solutions, and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present application. It should be understood that the specific implementation manners described herein are only used to illustrate and explain the embodiments of the present application, and are not used to limit the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present application.
[0056] It should be noted that if there are directional indications (such as up, down, left, right, front, back...) involved in the embodiments of the present application, the directional indications are only used to explain the relative positional relationship and movement conditions between components in a specific posture (as shown in the accompanying drawings). If the specific posture changes, the directional indications will also change accordingly.
[0057] In addition, if there are descriptions such as "first" and "second" involved in the embodiments of the present application, the descriptions of "first" and "second" are only for descriptive purposes, and cannot be understood as indicating or implying their relative importance or implicitly indicating the quantity of the indicated technical features. Thus, the features defined with "first" and "second" may explicitly or implicitly include at least one such feature. In addition, the technical solutions between various embodiments can be combined with each other, but it must be based on the fact that those of ordinary skill in the art can implement them. When the combination of technical solutions is contradictory or cannot be implemented, it should be considered that such a combination of technical solutions does not exist and is not within the protection scope required by the present application.
[0058] Figure 1 Schematically shows a flowchart of a method for diagnosing the health state of an ultrasonic flowmeter according to an embodiment of the present application. As Figure 1 shown, the embodiments of the present application provide a method for diagnosing the health state of an ultrasonic flowmeter, and the method may include the following steps.
[0059] Step 101, obtain the health state parameters of the ultrasonic flowmeter.
[0060] In an embodiment of the present application, an ultrasonic flowmeter is an instrument that measures flow by detecting the effect of fluid flow on an ultrasonic beam or ultrasonic pulse, and mainly consists of a transducer and a converter. The processor extracts the time series of the health status parameters of the ultrasonic flowmeter from historical data. Among them, the health status parameters may include, but are not limited to, sampling rate, signal gain, signal-to-noise ratio, vortex angle, measured gas velocity of each channel, and measured flow velocity of each channel. A time series is a sequence formed by arranging the values of the same statistical indicator in the order of their occurrence time. By obtaining the health status parameters of the ultrasonic flowmeter, it is convenient to comprehensively evaluate the health status of the ultrasonic flowmeter in the subsequent process.
[0061] Step 102: Determine the set of characteristic indicators of the health status parameters according to the health status parameters.
[0062] In an embodiment of the present application, in order to comprehensively evaluate the health status of the ultrasonic flowmeter, the processor analyzes the health status parameters from four aspects. Among them, the health status parameters may include, but are not limited to, sampling rate, signal gain, signal-to-noise ratio, vortex angle, measured gas velocity of each channel, and measured flow velocity of each channel. The set of characteristic indicators is a set of distortion degree indicators, distribution characteristic indicators, complexity degree indicators, and periodicity indicators of a health status parameter of an ultrasonic flowmeter. By obtaining the health status parameters and determining the set of characteristic indicators of the health status parameters, the health status of the ultrasonic flowmeter can be comprehensively diagnosed.
[0063] Step 103: Determine the basic data set according to the set of characteristic indicators.
[0064] In an embodiment of the present application, ultrasonic flowmeters usually work collaboratively. In the specific application process, it is necessary to monitor the health status of multiple ultrasonic flowmeters. Therefore, the processor can determine the basic data set according to the set of characteristic indicators. Among them, the basic data set is a set of distortion degree indicators, distribution characteristic indicators, complexity degree indicators, and periodicity indicators of a health status parameter of multiple ultrasonic flowmeters. By determining the basic data set, the health status of multiple ultrasonic flowmeters can be diagnosed according to the characteristic indicators of multiple ultrasonic flowmeters, improving the efficiency of monitoring the health status of ultrasonic flowmeters.
[0065] Step 104: Determine whether there is an abnormal working condition according to the basic data set.
[0066] In the embodiments of the present application, the processor can determine whether there is an abnormal working condition based on the basic data set. The abnormal working conditions can include, but are not limited to, pipeline blockage and component damage. The processor can obtain the basic data set, cluster the basic data set according to the K-means clustering algorithm combined with the maximum mutual information coefficient, determine the clustering result, and then determine the abnormal working condition according to the clustering result, so as to more accurately judge whether an abnormal working condition occurs. By clustering the basic data set according to the K-means clustering algorithm combined with the maximum mutual information coefficient, the identification of characteristic indicators under abnormal working conditions is made more accurate.
[0067] Through the above technical solution, the health status parameters of the ultrasonic flowmeter are obtained, and then the set of characteristic indicators of the health status parameters is determined according to the health status parameters. Thus, the basic data set is determined according to the set of characteristic indicators, and finally, it is determined whether there is an abnormal working condition based on the basic data set. On the one hand, this method promotes the basic research on the health status diagnosis method of ultrasonic flowmeters. On the other hand, because it introduces intelligent and data-based algorithms and incorporates historical data into real-time detection, it is beneficial to accurately evaluate the health status of ultrasonic flowmeters. By comprehensively analyzing the health status parameters of the ultrasonic flowmeter, the present application reduces the possibility of false alarms about the health status of the ultrasonic flowmeter and improves the comprehensiveness and accuracy of the health status monitoring of the ultrasonic flowmeter.
[0068] In the embodiments of the present application, the health status parameters can at least include one of the following:
[0069] Sampling rate, signal gain, signal-to-noise ratio, vortex angle, measured gas velocity of each channel, and measured flow velocity of each channel.
[0070] Specifically, the processor can obtain the health status parameters of the ultrasonic flowmeter. The sampling rate, also known as the sampling speed, is the number of samples extracted from a continuous signal and composed into a discrete signal per second. The signal gain is the ratio of the signal output to the signal input of the ultrasonic flowmeter. The signal-to-noise ratio is the ratio between the intensity of the maximum undistorted sound signal generated by the sound source and the intensity of the simultaneously emitted noise, which can reflect whether the pipeline is blocked. The vortex angle is the angle measured according to the different flow states of the gas in the ultrasonic flowmeter. Under stable flow, the vortex angle is usually small. When the vortex angle is large, it indicates that the flow state control of the ultrasonic flowmeter is unstable and turbulent flow appears. A channel refers to an ultrasonic channel formed by installing a pair of transducers on the measured pipeline or channel. The measured gas velocity of each channel is the gas flow velocity of each channel of the ultrasonic flowmeter. The measured flow velocity of each channel is the fluid flow velocity of each channel of the ultrasonic flowmeter. By collecting multiple health status parameters of the ultrasonic flowmeter, it is convenient to comprehensively evaluate the health status of the ultrasonic flowmeter from multiple aspects in the follow-up.
[0071] In the embodiments of the present application, in the embodiments of the present application, the set of characteristic indicators can include:
[0072] The distortion degree index, the distribution characteristic index, the complexity index, and the periodicity index.
[0073] Specifically, the processor can analyze the health status parameters from four aspects, that is, determine the distortion degree index, the distribution characteristic index, the complexity index, and the periodicity index of the health status parameters. The distortion degree index can be used to evaluate the dispersion degree of the health status index, so as to judge the degree to which the time series of the health status parameters deviates from the true value. Figure 2 Schematically shows a schematic diagram of a distortion degree index and a distribution characteristic index according to an embodiment of the present application. As Figure 2 shown, line 1 is the measured gas velocity of each sound channel actually measured, and line 2 is the measured gas velocity of each sound channel under normal conditions. In one example, the processor can determine the distortion degree index according to the measured gas velocity of each sound channel of the ultrasonic flowmeter. The distortion degree index satisfies formula (7):
[0074]
[0075] where S 1 is the distortion degree index, that is, the mean value of the time series of the health status parameters, n is the total number of measurement points, VOG i is the measured gas velocity of each sound channel actually measured at the i-th measurement point, and VOG i ′ is the measured gas velocity of each sound channel theoretically at the i-th measurement point.
[0076] The distribution characteristic index can be used to describe the characteristics of the numerical distribution of a time series segment of the health status parameters. In one example, when the ultrasonic flowmeter is in a normal working state, the distribution characteristic index of the time series of its health status parameters remains relatively stable. The complexity index is used to describe the complexity of the time series of the health status parameters. Among them, the complexity index includes two values, namely complexity and spectral statistical variance. The complexity is used to describe the number of peaks and valleys in the time series of the health status parameters. The spectral statistical variance is used to describe the frequency distribution difference of the sub-signals in the time series of the health status parameters. The periodicity index refers to the index that describes the periodicity of the time series of the health status parameters. By analyzing the six characteristic indexes of the health status parameters of the ultrasonic flowmeter, the health status of the ultrasonic flowmeter can be accurately diagnosed from multiple aspects.
[0077] In the embodiment of the present application, the distribution characteristic index includes kurtosis and skewness, and the kurtosis satisfies formula (1):
[0078]
[0079] where S 2 is the kurtosis, and x iis the value of the i-th measurement point in the time series of health state parameters, is the mean value of the values of all measurement points in the time series of health state parameters, std is the standard normal distribution, and n is the total number of measurement points;
[0080] The skewness satisfies formula (2):
[0081]
[0082] where S 3 is the skewness, is the mean value of the values of all measurement points in the time series of health state parameters, and std is the standard normal distribution.
[0083] Specifically, the standard normal distribution satisfies formula (8):
[0084]
[0085] where std is the standard normal distribution, x i is the value of the i-th measurement point in the time series of health state parameters, is the mean value of the values of all measurement points in the time series of health state parameters, and n is the total number of measurement points.
[0086] Figure 2 Schematically shows a schematic diagram of a distortion degree index and a distribution characteristic index according to an embodiment of the present application. As Figure 2 shown, line 1 is the measured gas velocity of each sound channel actually measured, and line 2 is the measured gas velocity of each sound channel under normal conditions. In the embodiment of the present application, the processor can determine the distribution characteristic index of the time series of health state parameters according to the acquired time series of health state parameters. The distribution characteristic index can be used to describe the characteristics of the numerical distribution of a time series segment of health state parameters. In one example, when the ultrasonic flowmeter is in a normal working state, the distribution characteristic index of the time series of its health state parameters remains relatively stable. Therefore, when the ultrasonic flowmeter is in a healthy working state, the distribution mode of the measurement points of the time series of its health state parameters should remain relatively stable, that is, the kurtosis and skewness vary within a certain range. The kurtosis is a statistic that describes the degree of steepness of the distribution form of all values in the population. The skewness is a measure of the skewness direction and degree of statistical data distribution. When the values of the kurtosis and skewness deviate from the conventional values, it indicates that the ultrasonic flowmeter may have abnormal working conditions. By determining the distribution characteristic index, the operating state of the ultrasonic flowmeter can be judged from the characteristics of the numerical distribution of the time series of health state parameters.
[0087] In the embodiment of the present application, the complexity index includes complexity and spectral statistical variance, and the complexity satisfies formula (3):
[0088]
[0089] Among them, S 4 is the complexity, n is the total number of measurement points, and x i is the value of the i-th measurement point in the time series of health state parameters, and x i-1 is the value of the (i - 1)-th measurement point in the time series of health state parameters;
[0090] The spectral statistical variance satisfies formula (4):
[0091]
[0092] Among them, S 5 is the spectral statistical variance, n is the total number of measurement points, k is the frequency of the sub-signal in the time series of health state parameters, X(k) is the amplitude corresponding to the frequency k after Fourier transform, is the mean value of the amplitudes after Fourier transform of the frequencies of the time series of health state parameters.
[0093] Specifically, X(k) satisfies formula (9):
[0094]
[0095] Among them, X(k) is the amplitude corresponding to the frequency k after Fourier transform, n is the total number of measurement points, and x i is the value of the i-th measurement point in the time series of health state parameters.
[0096] In the embodiments of the present application, the processor can determine the complexity index of the time series of health state parameters according to the obtained time series of health state parameters. The complexity index is used to describe the complexity of the time series of health state parameters. Among them, the complexity index includes two values, namely, complexity and spectral statistical variance. Figure 3 Schematically shows a schematic diagram of a complexity index according to an embodiment of the present application. As Figure 3As shown, line 3 is the measured gas velocity of each sound channel under normal conditions, and line 4 is the measured gas velocity of each sound channel actually measured. Complexity is used to describe the number of peaks and valleys in the time series of health state parameters. Spectral statistical variance is used to describe the frequency distribution difference of sub-signals in the time series of health state parameters. In one example, the frequency k of sub-signals in the time series of health state parameters is processed by Fourier transform. Fourier transform can transform the signal from the time domain to the frequency domain, and then the health state parameters can be analyzed from the frequency domain and the frequency domain characteristics of the health state parameters can be studied. The time domain is the relationship of a mathematical function or a physical signal with respect to time. The frequency domain is a coordinate system used to describe the characteristics of a signal in terms of frequency. When the ultrasonic flowmeter is in a healthy working state, the complexity of the time series of its health state parameters should remain at a relatively low level, that is, there are few peaks and valleys. Therefore, when the complexity and spectral statistical variance take large values, it can be determined that the working state of the ultrasonic wave is abnormal. By analyzing the complexity index of the time series of health state parameters, the operating state of the ultrasonic flowmeter can be judged according to the values of the complexity and spectral statistical variance of the time series of health state parameters.
[0097] In the embodiment of the present application, the periodicity index satisfies formula (5):
[0098]
[0099] Where S 6 is the periodicity index, that is, the autoregressive coefficient, n is the total number of measurement points, t is the lag coefficient, std is the standard normal distribution, and x i is the value of the i-th measurement point in the time series of health state parameters, is the mean value of the values of all measurement points in the time series of health state parameters, and x i+t is the value of the (i + t)-th measurement point in the time series of health state parameters.
[0100] Specifically, Figure 4 schematically shows a schematic diagram of a periodicity index according to an embodiment of the present application. As Figure 4 shown, line 5 is the measured gas velocity of each sound channel actually measured, and line 6 is the measured gas velocity of each sound channel under normal conditions. In the embodiment of the present application, the processor can determine the periodicity index of the time series of health state parameters according to the obtained time series of health state parameters. When the lag coefficient is a fixed value, in the case of a large autoregressive coefficient, it indicates that the time series of health state parameters does not have an obvious periodic law. Therefore, for the health state index with a periodic law, in the case of a large autoregressive coefficient, it can be judged that the ultrasonic flowmeter has an abnormal working condition. By analyzing the periodicity index of the time series of health state parameters, the operating state of the ultrasonic flowmeter can be judged according to the periodic law of the time series of health state parameters.
[0101] In an embodiment of the present application, determining whether there is an abnormal working condition based on the basic data set includes:
[0102] Obtain the basic data set;
[0103] Cluster the basic data set according to the K-means clustering algorithm to determine the clustering result;
[0104] Determine the abnormal working condition according to the clustering result.
[0105] Specifically, the processor can cluster the basic data set according to the K-means clustering algorithm to determine whether there is an abnormal working condition. The K-means clustering algorithm determines its proximity relationship by calculating the distance between different samples and puts the similar samples into the same category. Usually, the traditional K-means clustering algorithm uses the Euclidean distance to determine its proximity relationship, where the Euclidean distance refers to the true distance between two points in a multi-dimensional space or the natural length of a vector. Based on the basic data set formed after analyzing the characteristic indexes of the time series of the health state parameters of a large number of ultrasonic flowmeters, clustering is performed by combining the K-means clustering algorithm with the maximum mutual information coefficient. The processor can obtain the dimension of the basic data set and the number N of categories, where the dimension is the number of independent parameters in mathematics. The number N of categories refers to the categories of the health state. For example, the health state of the ultrasonic flowmeter can be divided into healthy, good, and unhealthy, then the number of categories is 3. The processor can divide the health state parameters into K clusters, that is, randomly select K parameters as the initial clustering centers, then calculate the distance between each parameter and the initial clustering centers, and assign each parameter to the clustering center closest to the parameter. Each time a parameter is assigned, recalculate the clustering center of the cluster according to the existing parameters in the cluster until the clustering center no longer changes or no parameter can be reassigned to a different clustering center.
[0106] In the case where the maximum mutual information coefficient is calculated, recalculate the clustering center instead of using the Euclidean distance. Figure 5 Schematically shows a schematic diagram of clustering the basic data set according to an embodiment of the present application. As Figure 5As shown, in one example, the processor can perform dimensionality reduction on the clustering result through the Principal Components Analysis (PCA) technique, so as to display the clustering result. Let N be 2 and the dimension be 6, and cluster the data in the basic dataset. Compared with the clustering result when the ultrasonic flowmeter is in a healthy state, the characteristic indexes of the health state parameters under abnormal working conditions are small samples. Therefore, the classes with a smaller number displayed in the clustering result are regarded as abnormal. It should be noted that the number of clusters K increases with the increase of the number N of the health state categories of the ultrasonic flowmeter. By clustering through the K-means clustering algorithm combined with the maximum mutual information coefficient, the relationship between non-linear parameters can be introduced, so as to determine and accurately judge abnormal working conditions comprehensively.
[0107] In the embodiment of the present application, clustering the basic dataset according to the K-means clustering algorithm includes:
[0108] Clustering the basic dataset in combination with the maximum mutual information coefficient.
[0109] Specifically, the processor can cluster the basic dataset by combining the maximum mutual information coefficient. The traditional K-means clustering algorithm clusters data through the Euclidean distance. In the case of non-linear relationships in the data, the traditional K-means clustering algorithm cannot cluster the data. Therefore, it is necessary to introduce the maximum mutual information coefficient to cluster data with non-linear relationships, so as to meet the need for clustering data with different relationships. In one example, when the maximum mutual information coefficient is calculated, the Euclidean distance is replaced to recalculate the clustering center. By clustering the basic dataset in combination with the maximum mutual information coefficient, the characteristic indexes in the basic dataset can be comprehensively analyzed.
[0110] In the embodiment of the present application, the maximum mutual information coefficient satisfies formula (6):
[0111]
[0112] Where x is any point in the basic dataset, y is another arbitrary point in the basic dataset, a is the number of grids divided on the x-axis, b is the number of grids divided on the y-axis, B is a variable, and its value is the 0.6th power of the data volume of the basic dataset, and I(x, y) is the mutual information.
[0113] Specifically, the mutual information I(x, y) between two points x and y satisfies formula (10):
[0114]
[0115] Among them, I(x, y) is the mutual information, x is any point in the basic data set, y is another arbitrary point in the basic data set, and p(x, y) is the joint probability of points x and y.
[0116] The processor can cluster the basic data set by combining the maximum mutual information coefficient. The traditional K-means clustering algorithm clusters data through the Euclidean distance. In the case where the data has a non-linear relationship, the traditional K-means clustering algorithm cannot cluster the data. Therefore, it is necessary to introduce the maximum mutual information coefficient to cluster data with non-linear relationships, so as to meet the need for clustering data with different relationships. By clustering the basic data set by combining the maximum mutual information coefficient, the characteristic indicators in the basic data set can be comprehensively analyzed.
[0117] In the embodiment of the present application, it may further include:
[0118] Determine an alarm threshold according to the basic data set.
[0119] Specifically, Figure 6 Schematically shows a schematic diagram of determining an alarm threshold according to an embodiment of the present application. As Figure 6 shown, in the embodiment of the present application, the processor can perform statistics on the clustering result to determine the alarm threshold. In actual use, there are usually multiple ultrasonic flowmeters. Cluster the basic data set composed of the characteristic indicator sets of the health status parameters of multiple ultrasonic flowmeters, and determine the maximum and minimum values of the characteristic indicators according to the clustering result, so as to determine the alarm threshold of the characteristic indicators. In one example, when the characteristic indicator falls outside the threshold range, it can be determined as an abnormal working condition. When the characteristic indicator falls within the threshold range, it can be determined as a normal working condition. Further, in the case of a new characteristic indicator, when the new characteristic indicator falls outside the threshold range, it can be determined as an abnormal working condition. When the new characteristic indicator falls within the threshold range, it can be determined as a normal working condition. It should be noted that the alarm threshold can be changed according to the change of the value of the characteristic indicator of the basic data set. By continuously updating the alarm threshold according to the clustering result obtained in real time, the need for accurately diagnosing the health status of ultrasonic flowmeters under different working conditions can be met.
[0120] Figure 7 Schematically shows a structural block diagram of a device for diagnosing the health status of an ultrasonic flowmeter according to an embodiment of the present application. As Figure 7 shown, the embodiment of the present application provides a device for diagnosing the health status of an ultrasonic flowmeter, which may include:
[0121] A memory 710 configured to store instructions; and
[0122] A processor 720, configured to call instructions from a memory 710 and capable of implementing the above method for diagnosing the health status of an ultrasonic flowmeter when executing the instructions.
[0123] Specifically, in an embodiment of the present application, the processor 720 may be configured to:
[0124] Obtain the health status parameters of the ultrasonic flowmeter;
[0125] Determine a set of characteristic indicators of the health status parameters according to the health status parameters;
[0126] Determine a basic data set according to the set of characteristic indicators;
[0127] Determine whether there is an abnormal working condition according to the basic data set.
[0128] In an embodiment of the present application, the health status parameters may at least include one of the following:
[0129] Sampling rate, signal gain, signal-to-noise ratio, vortex angle, measured gas velocity of each channel, and measured flow velocity of each channel.
[0130] In an embodiment of the present application, the set of characteristic indicators may include:
[0131] Distortion degree index, distribution characteristic index, complexity index, periodicity index.
[0132] In an embodiment of the present application, the distribution characteristic index includes kurtosis and skewness, and the kurtosis satisfies formula (1):
[0133]
[0134] Wherein, S 2 is the kurtosis, x i is the value of the i-th measurement point in the time series of health status parameters, is the mean value of all measurement points in the time series of health status parameters, std is the standard normal distribution, and n is the total number of measurement points;
[0135] The skewness satisfies formula (2):
[0136]
[0137] Wherein, S 3 is the skewness, is the mean value of all measurement points in the time series of health status parameters, and std is the standard normal distribution.
[0138] In an embodiment of the present application, the complexity index includes complexity and spectral statistical variance, and the complexity satisfies formula (3):
[0139]
[0140] Among them, S 4 is the complexity, n is the total number of measurement points, and x i is the value of the i-th measurement point in the time series of health state parameters, and x i-1 is the value of the (i - 1)-th measurement point in the time series of health state parameters;
[0141] The spectral statistical variance satisfies formula (4):
[0142]
[0143] Among them, S 5 is the spectral statistical variance, n is the total number of measurement points, k is the frequency of the sub-signal in the time series of health state parameters, X(k) is the amplitude corresponding to the frequency k after Fourier transform, is the mean value of the amplitudes after Fourier transform of the frequencies of the time series of health state parameters.
[0144] In the embodiments of the present application, the periodicity index satisfies formula (5):
[0145]
[0146] Among them, S 6 is the autoregressive coefficient, n is the total number of measurement points, t is the lag coefficient, std is the standard normal distribution, and x i is the value of the i-th measurement point in the time series of health state parameters, is the mean value of the values of all measurement points in the time series of health state parameters, and x i+t is the value of the (i + t)-th measurement point in the time series of health state parameters.
[0147] Furthermore, the processor 720 can also be configured to:
[0148] Obtain the basic data set;
[0149] Cluster the basic data set according to the K-means clustering algorithm to determine the clustering result;
[0150] Determine the abnormal working conditions according to the clustering result.
[0151] Furthermore, the processor 720 can also be configured to:
[0152] Cluster the basic data set in combination with the maximum mutual information coefficient.
[0153] In the embodiments of the present application, the maximum mutual information coefficient satisfies formula (6):
[0154]
[0155] Wherein, x is any point in the basic dataset, y is another arbitrary point in the basic dataset, a is the number of grids divided on the x-axis, b is the number of grids divided on the y-axis, B is a variable, and its value is the 0.6th power of the data volume of the basic dataset, and I(x, y) is the mutual information.
[0156] Furthermore, the processor 720 can also be configured to:
[0157] Determine an alarm threshold according to the basic dataset.
[0158] Through the above technical solution, the health status parameters of the ultrasonic flowmeter are obtained, and then the characteristic index set of the health status parameters is determined according to the health status parameters, so as to determine the basic dataset according to the characteristic index set, and finally determine whether there is an abnormal working condition according to the basic dataset. On the one hand, this method promotes the basic research on the health status diagnosis method of ultrasonic flowmeters. On the other hand, because it introduces intelligent and data-based algorithms and introduces historical data into real-time detection, it is beneficial to accurately evaluate the health status of ultrasonic flowmeters. This application comprehensively analyzes the health status parameters of the ultrasonic flowmeter, reduces the possibility of false alarms for the health status of the ultrasonic flowmeter, and improves the comprehensiveness and accuracy of the health status monitoring of the ultrasonic flowmeter.
[0159] The embodiment of the present application also provides a machine-readable storage medium, on which instructions are stored, and the instructions are used to make the machine execute the above method for diagnosing the health status of the ultrasonic flowmeter.
[0160] Those skilled in the art should understand that the embodiments of the present application can be provided as methods, systems, or computer program products. Therefore, the present application can adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.
[0161] The present application is described with reference to the flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or block in the flowchart and / or block diagram, and the combination of processes and / or blocks in the flowchart and / or block diagram can be realized by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing devices generate for realizing in the process Figure 1 one process or multiple processes and / or blocks Figure 1means for the functions specified in one or more blocks.
[0162] These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to operate in a particular manner, such that the instructions stored in the computer-readable memory produce a manufacture including an instruction means that implements the functions specified in one Figure 1 one or more processes and / or blocks Figure 1 means for the functions specified in one or more blocks.
[0163] These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus, such that a series of operational steps are performed on the computer or other programmable apparatus to produce a computer-implemented process, whereby the instructions executed on the computer or other programmable apparatus provide steps for implementing the functions specified in one Figure 1 one or more processes and / or blocks Figure 1 means for the functions specified in one or more blocks.
[0164] In a typical configuration, a computing device includes one or more processors (CPUs), an input / output interface, a network interface, and memory.
[0165] The memory may include non-permanent memory in the form of computer-readable media, random access memory (RAM) and / or non-volatile memory such as read-only memory (ROM) or flash memory (flash RAM). Memory is an example of computer-readable media.
[0166] Computer-readable media includes both permanent and non-permanent, removable and non-removable media implemented by any method or technology for storing information. The information can be computer-readable instructions, data structures, program modules, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, compact disc read-only memory (CD-ROM), digital versatile discs (DVD) or other optical storage, magnetic cassettes, magnetic tape disk storage or other magnetic storage devices, or any other non-transmission media that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transitory media such as modulated data signals and carrier waves.
[0167] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.
[0168] The above are only embodiments of the present application and are not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included within the scope of the claims of the present application.
Claims
1. A method for diagnosing the health status of an ultrasonic flowmeter, characterized in that, the method includes: obtaining the health status parameters of the ultrasonic flowmeter; determining a set of characteristic indicators of the health status parameters according to the health status parameters; determining a basic data set according to the set of characteristic indicators; determining whether there is an abnormal working condition according to the basic data set; wherein, the health status parameters at least include one of the following: sampling rate, signal gain, signal-to-noise ratio, vortex angle, measured gas velocity of each channel, and measured flow velocity of each channel; the set of characteristic indicators includes a distortion degree indicator, a distribution characteristic indicator, a complexity indicator, and a periodicity indicator corresponding to the health status parameters of the ultrasonic flowmeter; the basic data set includes distortion degree indicators, distribution characteristic indicators, complexity indicators, and periodicity indicators corresponding to the health status parameters of multiple ultrasonic flowmeters; the determining whether there is an abnormal working condition according to the basic data set includes: obtaining the basic data set; clustering the basic data set according to the K-means clustering algorithm to determine the clustering result; determining the abnormal working condition according to the clustering result; the clustering the basic data set according to the K-means clustering algorithm includes: clustering the basic data set in combination with the maximum mutual information coefficient.
2. The method according to claim 1, characterized in that, the distribution characteristic indicator includes kurtosis and skewness, and the kurtosis satisfies formula (1): Among them, S 2 is the kurtosis, x i is the value of the i-th measurement point in the time series of health state parameters, is the mean value of the values of all measurement points in the time series of health state parameters, std is the standard normal distribution, and n is the total number of measurement points; the skewness satisfies formula (2): Among them, S 3 is the skewness, is the mean value of all measured points in the time series of the health status parameter, and std is the standard normal distribution.
3. The method according to claim 1, characterized in that, the complexity indicator includes complexity and spectral statistical variance, and the complexity satisfies formula (3): Among them, S 4 is the complexity, n is the total number of measurement points, x i is the value of the i-th measurement point in the time series of health state parameters, x i-1 is the value of the (i - 1)-th measurement point in the time series of health state parameters; the spectral statistical variance satisfies formula (4): Among them, S 5 is the spectral statistical variance, n is the total number of measurement points, k is the frequency of the sub-signal in the time series of the health state parameter, and X(k) is the amplitude corresponding to the frequency k after Fourier transform. is the mean value of the amplitudes of the frequencies of the time series of the health state parameters after Fourier transform.
4. The method according to claim 1, characterized in that, the periodicity indicator satisfies formula (5): Among them, S 6 is the autoregressive coefficient, n is the total number of measurement points, t is the lag coefficient, std is the standard normal distribution, and x i is the value of the i-th measurement point in the time series of health state parameters, is the mean value of the values of all measurement points in the time series of health state parameters, and x i+t is the value of the (i + t)-th measurement point in the time series of health state parameters.
5. The method according to claim 1, characterized in that, the maximum mutual information coefficient satisfies formula (6): wherein, x is any point in the basic data set, y is another arbitrary point in the basic data set, a is the number of grids divided on the x-axis, b is the number of grids divided on the y-axis, B is a variable, and its value is the 0.6th power of the data volume of the basic data set, and I(x, y) is the mutual information.
6. The method according to claim 1, characterized in that, the method further includes: determining an alarm threshold according to the basic data set.
7. A device for diagnosing the health status of an ultrasonic flowmeter, characterized in that, including: a memory configured to store instructions; and a processor configured to call the instructions from the memory and be able to implement the method for diagnosing the health status of an ultrasonic flowmeter according to any one of claims 1 to 6 when executing the instructions.
8. A machine-readable storage medium, characterized in that, instructions are stored on the machine-readable storage medium, and the instructions are used to cause the machine to execute the method for diagnosing the health status of an ultrasonic flowmeter according to any one of claims 1 to 6.