Full waveform diagnostic and adaptive gas flow metering system

Abstract

The present application relates to the technical field of flow measurement and industrial monitoring, in particular to a full waveform diagnosis and adaptive gas flow measurement system; comprising full waveform acquisition, feature extraction, diagnostic analysis, dynamic weight calculation and health degree evaluation modules; the system obtains acoustic channel characteristics through digital sampling and multi-dimensional analysis of ultrasonic echoes; the core is to dynamically adjust the weight of each acoustic channel according to the waveform similarity and sound velocity deviation index to obtain the standard flow and generate the measurement health index; the present application realizes the transformation from single feature analysis to full waveform physical state deconstruction, can accurately distinguish scale, noise and turbulence anomalies, and overcomes the defect that the traditional technology cannot identify the root cause of the fault.

Description

Technical Field

[0001] This invention relates to the field of flow metering and industrial monitoring technology, specifically to a full-waveform diagnostic and adaptive gas flow metering system. Background Technology

[0002] As the requirements for measurement accuracy in natural gas trade settlement continue to increase, the application scenarios of gas ultrasonic flow meters are becoming increasingly complex.

[0003] Currently, flow monitoring mainly relies on traditional ultrasonic flow meters, which typically calculate flow based on preset fixed logic or single signal characteristics. However, traditional methods lack effective self-diagnostic capabilities when faced with harsh operating conditions such as dirt, liquid accumulation, or valve noise. Single detection indicators are difficult to accurately distinguish the root cause of the fault, and when some channels fail or sensors age, the system cannot adaptively adjust the calculation weights, resulting in decreased measurement robustness and easy generation of imperceptible long-term measurement deviations, which in turn affect the fairness of trade settlement.

[0004] Therefore, how to achieve multidimensional holographic diagnosis and adaptive precision measurement under complex working conditions has become an urgent problem to be solved in this field. Summary of the Invention

[0005] To address the aforementioned technical problems, this invention provides a full-waveform diagnostic and adaptive gas flow metering system. Specifically, the technical solution of this invention includes: a full-waveform acquisition module for digitally sampling received ultrasonic echoes to obtain a full-waveform digital signal; an acoustic feature extraction module for performing multi-dimensional feature analysis on the full-waveform digital signal to obtain an acoustic channel feature set; a diagnostic analysis module for performing matching analysis on the acoustic channel feature set based on pre-stored reference waveform data to obtain a waveform similarity index, and verifying it in conjunction with theoretical sound velocity to obtain a sound velocity deviation index; a dynamic weighted flow calculation module for dynamically adjusting the calculation weights of each channel according to the waveform similarity index and the sound velocity deviation index, and calculating standard flow data based on the adjusted calculation weights and the full-waveform digital signal; and a metering health assessment module for generating a metering health index based on the acoustic channel feature set and the waveform similarity index and outputting diagnostic status data.

[0006] Preferably, the acoustic feature extraction module is used to: extract gain value features reflecting the signal attenuation level from the full waveform digital signal; calculate signal-to-noise ratio features reflecting the background noise level; analyze the waveform distortion level and peak value ratio features; calculate the turbulence intensity features reflecting the flow field stability; and combine the gain value features, signal-to-noise ratio features, waveform width features, peak value ratio features, and turbulence intensity features to obtain the acoustic channel feature set.

[0007] Preferably, the diagnostic analysis module includes: a waveform matching unit, used to call a cross-correlation algorithm to compare the current full-waveform digital signal with the reference waveform data, and calculate the waveform similarity index; and a sound velocity verification unit, used to calculate the difference between the measured sound velocity based on the full-waveform digital signal and the theoretical sound velocity calculated based on the current operating parameters, so as to obtain the sound velocity deviation index.

[0008] Preferably, the sound velocity verification unit is configured to: acquire current pressure parameters, temperature parameters, and gas composition parameters; calculate the theoretical sound velocity based on the pressure parameters, temperature parameters, and gas composition parameters; generate warning data indicating a risk of measurement inaccuracy in response to the sound velocity deviation index being greater than a preset safety threshold; and generate confirmation data indicating normal measurement in response to the sound velocity deviation index being less than or equal to the safety threshold.

[0009] Preferably, the dynamic weighted flow calculation module is used to: in a multi-channel system, acquire the waveform similarity index and signal-to-noise ratio feature of each channel in real time; in response to the waveform similarity index of a certain channel being lower than a preset quality threshold or the signal-to-noise ratio feature being lower than a preset noise threshold, set the calculation weight of that channel to a preset downgrade weight value that is lower than the standard weight; and perform a weighted average calculation on the flow rate data of each channel based on the adjusted calculation weight of each channel to obtain the standard flow data.

[0010] Preferably, the calculation of the turbulence intensity feature reflecting the stability of the flow field in the acoustic feature extraction module is specifically used for: collecting a flow velocity data sequence within a preset time window; calculating the normalized variance of the flow velocity data sequence; and using the normalized variance of the flow velocity as the turbulence intensity feature to characterize the degree of disturbance in the flow field.

[0011] Preferably, the metrological health assessment module is configured to: invoke a preset multi-dimensional scoring logic, the logic including a normalization algorithm for gain value, signal-to-noise ratio, waveform distortion degree, and turbulence degree, and corresponding weighting coefficients; input the acoustic channel feature set into the multi-dimensional scoring logic for normalization and weighted summation calculation to generate the metrological health index; and generate alarm signal data in response to the metrological health index being lower than a preset maintenance threshold.

[0012] Preferably, the metering health assessment module is further configured to: generate a specific diagnostic code characterizing a transducer fouling warning in response to the gain value characteristic exceeding a preset normal gain range; generate a specific diagnostic code characterizing flow field vortex interference in response to the turbulence intensity characteristic being greater than a preset turbulence threshold; and generate a specific diagnostic code characterizing valve noise exceeding limits in response to the signal-to-noise ratio characteristic being less than a preset signal-to-noise ratio threshold.

[0013] Compared with the prior art, the present invention has the following beneficial effects:

[0014] 1. By combining the full waveform acquisition module and the acoustic feature extraction module, this system is not limited to a single signal feature, but deconstructs multiple independent physical indicators such as gain, signal-to-noise ratio, waveform distortion, and turbulence intensity from the full waveform digital signal. This multi-dimensional feature extraction mechanism can transform abstract waveform data into specific physical states, thereby accurately distinguishing signal attenuation caused by transducer fouling, high-frequency interference caused by pressure regulating valve noise, and turbulence anomalies caused by flow field instability, overcoming the shortcomings of traditional technologies that cannot identify the root cause of faults with a single indicator.

[0015] 2. Utilizing a dynamic weighted flow calculation module, this system can monitor the waveform similarity and signal-to-noise ratio of each channel in a multi-channel system in real time. When a channel is detected to have degraded signal quality due to environmental interference or sensor aging, the system will automatically reduce the calculation weight of the faulty channel and reasonably allocate the remaining weight to other healthy channels. This mechanism ensures that even when some channels fail or are in a sub-healthy state, the flow meter can still maintain a high-confidence standard flow output, avoiding overall measurement deviation caused by local faults.

[0016] 3. Through the sound velocity verification unit in the diagnostic analysis module, this system compares the actual sound velocity measured based on the full waveform with the theoretical sound velocity calculated based on pressure, temperature, and gas composition. By using the gas law as the benchmark anchor point for measurement, it can promptly detect parameter drift or undetected measurement inaccuracies. This real-time verification based on physical laws ensures that every measurement data is supported by thermodynamic theory, effectively preventing long-term measurement losses and enhancing the reliability of the data.

[0017] 4. The metering health assessment module of this system uses multi-dimensional scoring logic to integrate underlying data such as gain, signal-to-noise ratio, and waveform distortion to generate a unified metering health index. This index can intuitively reflect the overall health of the instrument and issue an alarm in a timely manner when the index is lower than the maintenance threshold. This allows maintenance personnel to intuitively grasp the equipment status and carry out targeted maintenance before a fault occurs, thereby significantly improving fault handling efficiency and reducing unplanned gas outage maintenance time. Attached Figure Description

[0018] The present invention will be further explained below with reference to the accompanying drawings and embodiments:

[0019] Figure 1 This is a structural diagram of the system of the present invention. Detailed Implementation

[0020] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to specific embodiments.

[0021] Example 1:

[0022] Please see Figure 1 The full-waveform diagnostic and adaptive gas flow metering system includes: a full-waveform acquisition module for digitally sampling received ultrasonic echoes to obtain full-waveform digital signals; an acoustic feature extraction module for performing multi-dimensional feature analysis on the full-waveform digital signals to obtain acoustic channel feature sets; a diagnostic analysis module for performing matching analysis on the acoustic channel feature sets based on pre-stored benchmark waveform data to obtain waveform similarity indices, and verifying them with theoretical sound speed to obtain sound speed deviation indices; a dynamic weighted flow calculation module for dynamically adjusting the calculation weights of each channel according to the waveform similarity index and sound speed deviation index, and calculating standard flow data based on the adjusted calculation weights and the full-waveform digital signals; and a metering health assessment module for generating a metering health index based on the acoustic channel feature sets and waveform similarity indices and outputting diagnostic status data.

[0023] This embodiment details the physical architecture and signal flow logic of the flowmeter. The full waveform acquisition module is equipped with a high-speed analog-to-digital converter (ADC), and the sampling frequency is set to more than 10 times the probe center frequency, such as 20MHz, to acquire raw data containing complete acoustic physical information. This process not only records the main peak but also covers the leading wave and subsequent reverberation, thereby generating a full waveform digital signal. The acoustic feature extraction module uses digital signal processing (DSP) algorithms to deconstruct waveforms from both the time and frequency domains, transforming abstract waveform data into physical state indicators that can be identified by diagnostic logic, namely, the acoustic channel feature set.

[0024] The diagnostic analysis module, as the core computing unit of the system, calculates the waveform similarity index between the current waveform and the baseline waveform under factory clean conditions. Simultaneously, the sound velocity deviation index is calculated based on the current operating parameters. This establishes dual verification at both the signal and physical layers. Based on this, the dynamic weighted flow calculation module executes an adaptive metering strategy, dynamically adjusting the calculated weights of each channel in real time. When a channel is contaminated or interfered with, its weight is automatically reduced, thereby calculating high-confidence standard flow data based on the adjusted weight. The quantitative health assessment module generates a quantitative health index through a multi-dimensional scoring model. It also outputs diagnostic status data to enable predictive maintenance.

[0025] This embodiment constructs a metering mode based on acoustic feature holographic diagnosis through a full waveform acquisition and multi-dimensional feature analysis architecture. It effectively solves the problem that traditional flow meters lack self-diagnostic capabilities under harsh working conditions such as dirt, liquid accumulation, or valve noise. Through a dynamic weight adjustment mechanism, it maintains the robustness of overall metering even when some channels fail, avoiding unfair trade settlements caused by sensor aging or environmental interference.

[0026] Example 2:

[0027] The acoustic feature extraction module is used to: extract gain features reflecting the degree of signal attenuation from the full waveform digital signal; calculate signal-to-noise ratio features reflecting the background noise level; analyze the waveform distortion features and peak value ratio features reflecting the degree of waveform distortion; calculate the turbulence intensity features reflecting the stability of the flow field; and combine the gain features, signal-to-noise ratio features, waveform width features, peak value ratio features, and turbulence intensity features to obtain the acoustic channel feature set.

[0028] This embodiment details the steps of the acoustic feature extraction module in constructing a feature set; the module extracts gain value features that reflect the degree of signal attenuation. The calculation formula is as follows: ;in, The peak amplitude of the signal acquired by the ADC comes from real-time acquisition. The transducer excitation voltage amplitude is derived from a hardware preset constant.

[0029] The module calculates the signal-to-noise ratio characteristics that reflect the background noise level. The aim is to quantify the high-frequency noise interference generated by the pressure regulating valve, and the calculation formula is as follows: ;in, The signal power within the signal time window is derived from the calculation of the full waveform digital signal. The noise power within the noise window before the signal arrives is derived from the calculation of the full waveform digital signal.

[0030] The module analyzes waveform distortion features, including waveform width characteristics. Peak ratio characteristics Specifically, wavewidth characteristics The calculation formula is as follows: ;in, The waveform envelope amplitude increases and reaches a specific threshold for the first time. At that moment, The waveform envelope amplitude decreases and falls below this specific threshold for the first time. The moment; a specific threshold Set to the current signal's maximum peak amplitude of ,Right now .

[0031] Peak ratio characteristics The calculation formula is as follows: ;in, This represents the amplitude of the main peak, which has the largest amplitude in the entire waveform digital signal. The amplitude value of the secondary peak immediately following the primary peak; this ratio is used to detect wave skipping risk.

[0032] Turbulence intensity characteristics that reflect flow field stability The calculation method is specified in this embodiment as follows: ;in, To prevent extremely small positive numbers with a denominator of zero, for example ; For the first time within the preset time window A sample of instantaneous flow velocity. This represents the absolute value of the average flow velocity within the window. The total number of samples; the module will feature the above gain values. Signal-to-noise ratio characteristics Wavelength characteristics Peak-to-peak ratio characteristics and turbulence characteristics Combined into vector form, we obtain the acoustic channel feature set. .

[0033] This embodiment decouples a single ultrasonic signal into independent physical indicators that reflect the degree of fouling, noise interference, waveform integrity, and flow field stability, providing rich data dimensions for subsequent accurate diagnosis. This multi-dimensional feature extraction mechanism overcomes the deficiency that a single indicator cannot distinguish the root cause of the fault, enabling the system to accurately identify whether the signal amplitude drop is caused by transducer fouling or fluid attenuation.

[0034] Example 3:

[0035] The diagnostic analysis module includes: a waveform matching unit, which calls a cross-correlation algorithm to compare the current full-waveform digital signal with the reference waveform data and calculates the waveform similarity index; and a sound velocity verification unit, which calculates the difference between the measured sound velocity based on the full-waveform digital signal and the theoretical sound velocity calculated based on the current operating parameters to obtain the sound velocity deviation index.

[0036] This embodiment details the operational logic of the core unit in the diagnostic analysis module, particularly how to perform matching analysis on the acoustic channel feature set based on reference waveform data; the waveform matching unit performs multi-dimensional matching calculations: it not only calls the Normalized Cross-Correlation (NCC) algorithm to compare the current full-waveform digital signal, but also... With pre-stored reference waveform data To obtain a direct correlation with waveform morphology, it further compares with the current set of acoustic channel features. This includes gain, signal-to-noise ratio, etc., and a set of reference features derived from the reference waveform. Waveform similarity index The calculation incorporates waveform cross-correlation coefficients. Similarity with feature vectors The calculation logic is as follows: In the specific implementation of the project: ;in, The cross-correlation coefficient of the entire waveform; while This is derived based on the following standardized weighted Euclidean distance algorithm.

[0037] This formula is a general vector space model for similarity calculation. In specific engineering implementations, the following standardized weighted Euclidean distance algorithm is used. To eliminate the significant differences in numerical dimensions between different physical features such as gain and turbulence intensity, and to avoid large numerical features masking the changes in small numerical features, this embodiment introduces Z-Score standardization processing before calculating the distance: This standardization is applied to the current acoustic channel feature set. and benchmark feature set Each feature component in Standardization transformation is performed to obtain : ;in, It is a very small positive number; and The first The mean and standard deviation of each feature in the historical benchmark dataset collected during the equipment factory calibration phase.

[0038] Calculate the standardized weighted Euclidean distance : In the formula, This is the feature vector obtained after standardization of the current acoustic channel feature set. The feature vectors are the standardized versions of the baseline feature set. For the first in the benchmark feature set The baseline value of each feature component after standardization transformation.

[0039] Based on the normalized distance calculation of the feature similarity term, the feature vector distance in the conceptual formula is concretized into a weighted Euclidean distance form, at which point the waveform similarity index... The complete calculation formula is: .

[0040] in, The effective decision radius of the feature space is typically set to a value of [value to be filled in]. ,correspond Deviation; The weighting coefficients are set to balance the contributions of the overall waveform shape and physical feature details in the similarity evaluation. Considering that the overall matching degree of waveform shape is the primary basis for judging signal quality, it is usually set to... Correspondingly In the preferred configuration of this embodiment, the following is taken: , .

[0041] This process ensures that the diagnostic analysis module strictly adheres to the requirements of the embodiment for matching and analyzing the acoustic channel feature set; simultaneously, the sound velocity verification unit uses the physical properties of the gas for logical verification and calculates the sound velocity deviation index. The calculation formula is as follows: ;in, To actually measure the speed of sound, This is the theoretical speed of sound calculated based on operating parameters.

[0042] This embodiment constructs a dual verification mechanism of the signal layer and the physical layer. By introducing matching analysis for the acoustic channel feature set, the waveform similarity index not only reflects the difference in waveform geometry, but also covers the deviation of signal physical characteristics, such as signal-to-noise ratio and gain, thus more comprehensively characterizing the health status of the vocal tract. Combined with the theoretical verification of the sound velocity verification unit, the reliability of the diagnostic system is effectively improved.

[0043] Example 4:

[0044] The sound velocity verification unit is configured to: acquire the current pressure parameters, temperature parameters, and gas composition parameters; calculate the theoretical sound velocity based on the pressure parameters, temperature parameters, and gas composition parameters; generate warning data indicating a risk of measurement inaccuracy in response to the sound velocity deviation index being greater than a preset safety threshold; and generate confirmation data indicating normal measurement in response to the sound velocity deviation index being less than or equal to the safety threshold.

[0045] This embodiment further refines the data interaction and judgment logic of the sound velocity verification unit; this unit obtains the pressure parameters measured by the field transmitter through a digital communication interface. Temperature parameters In addition to the gas component parameters provided by the gas chromatograph; the built-in calculation engine calculates the theoretical velocity of sound based on the above parameters and in accordance with relevant international standards. The coefficients and calculation logic of the gas state equations involved in this embodiment shall be performed in accordance with the international standards ISO20765-1 or AGA Report No.8.

[0046] Specifically, theoretical speed of sound The calculation uses the following thermodynamic formula based on the Helmholtz equation of state: ;in, It is a universal gas constant; The molar mass of the gas mixture, in units of It is calculated by weighting the mole fraction and molecular weight of each component.

[0047] The following dimensionless quantities are introduced in the formula: For comparison of density, where, molar density, The critical molar density; For temperature comparison, among which, The critical temperature. The fluid thermodynamic temperature; The remainder of the Helmholtz free energy is dimensionless. , , They are respectively about First-order partial derivatives, second-order partial derivatives, and related and The mixed partial derivatives; It is a dimensionless constant-volume isochoric heat capacity.

[0048] Define contrast density and comparison temperature Remaining Helmholtz free energy function The fitting calculation is performed using the following polynomial expansion: In the above formula, These are the preset coefficients for the corresponding gas law equation.

[0049] Based on the expansion defined above, the dimensionless partial derivatives required for calculating the speed of sound are expanded as follows: ; ;

[0050] In the formula, and These are the dimensionless first and second partial derivatives of the remaining Helmholtz free energy with respect to the relative density, respectively.

[0051] In addition, dimensionless constant-volume heat capacity The calculation is as follows: ;in Let be the dimensionless isochoric heat capacity of an ideal gas.

[0052] To Second-order partial derivatives: .

[0053] Cross partial derivatives for: .

[0054] Regarding the core variable molar density in the formula For problems that cannot be directly measured, we use methods based on known information. arrive Iterative solution algorithm: due to the state equation It is a nonlinear equation about density. Characterizing the nonlinear thermodynamic relationship between pressure, molar density, and temperature derived from the Helmholtz equation of state, the system first generates initial values ​​using the ideal gas equation of state. The Newton-Raphson method is used for iterative solution: In the formula, For the first The theoretical pressure calculated based on the current molar density and temperature in the next iteration. The actual measured operating pressure. Let be the partial derivative of pressure with respect to molar density under isothermal conditions; to achieve the above Newton-Raphson iteration, the pressure must be clearly defined. Regarding molar density The partial derivative analytical expression.

[0055] This embodiment constructs a pressure system based on the aforementioned Helmholtz free energy equation. The explicit physical definition is as follows: Based on this equation of state, the molar density... Take the partial derivative, noting the contrast density. The derivative terms required for the iteration are derived. : The system will calculate the dimensionless partial derivatives in real time. and Substituting into the above equation yields the exact Jacobian derivative term.

[0056] in, This is the derivative term mentioned above; the iterative process continues until the difference between the calculated pressure and the measured pressure is reached. Less than the preset convergence accuracy, such as MPa, thus obtaining accurate Used for subsequent sound speed calculations; the system will calculate the sound speed deviation index. With preset safety threshold Comparison; response Greater than If the system determines that the reliability of the current measurement result is questionable, it generates warning data indicating a risk of measurement inaccuracy. This data includes specific deviation values, suggesting to the user that there may be an error in the input of gas components or a slight change in the length of the flow meter channel; conversely, in response to... Less than or equal to The system generates confirmation data indicating that the measurement is normal.

[0057] This embodiment uses the gas law to establish a benchmark anchor point for measurement. In trade transactions, this real-time verification based on physical laws provides strong non-repudiable evidence, ensuring that every measurement data is supported by thermodynamic theory and effectively preventing long-term measurement losses caused by undetected parameter drift.

[0058] Example 5:

[0059] The dynamic weighted flow calculation module is used to: acquire the waveform similarity index and signal-to-noise ratio characteristics of each channel in real time in a multi-channel system; in response to the waveform similarity index of a certain channel being lower than a preset quality threshold or the signal-to-noise ratio characteristics being lower than a preset noise threshold, set the calculation weight of that channel to a preset downgrade weight value that is lower than the standard weight; and perform a weighted average calculation on the flow rate data of each channel based on the adjusted calculation weight of each channel to obtain the standard flow rate data.

[0060] This embodiment details the adaptive weight adjustment mechanism in a multi-channel system, particularly how to dynamically adjust it in conjunction with the sound velocity deviation index of Embodiment 1; within each calculation cycle, the module obtains the first... Waveform similarity index of vocal tract Signal-to-noise ratio characteristics and sound speed deviation index The system compares these indicators with preset quality thresholds. Noise threshold A comparison is performed; to ensure the consistency of the technical solutions, the system executes the following comprehensive criteria: response to a certain channel. Meet any of the following conditions: This means that the waveform distortion is severe. This means that the noise interference is too great.

[0061] The system determines that the channel is in a sub-healthy or interfered state and adjusts the calculated weight of that channel. Forced to be set as a downgrade weight value ; The value is determined based on the severity level of the fault: when the vocal tract is completely lost, the waveform or waveform similarity is determined. At that time, set That is, completely exclude that channel from the calculation; when And sound speed deviation When not exceeding the limit, set ,in, The standard weights are used to retain a portion of the contribution to maintain the integrity of the flow field information.

[0062] Simultaneously, the weights of other healthy channels are increased accordingly to maintain weight normalization; the specific weight redistribution mathematical operator is as follows: To prevent simple normalization from causing the weights of downgraded faulty channels to unexpectedly rebound, this embodiment adopts a residual weight redistribution strategy, with the specific steps as follows: Step 1: Lock faulty weights; identify all faulty or sub-healthy channels, i.e., those not belonging to The channel's weight is forcibly fixed to a preset degradation value. ,like or And calculate the total weight occupied in this part. : .

[0063] Step 2: Calculate the remaining quota; calculate the weight margin available for allocation to the healthy audio channel. : .

[0064] Step 3: Proportional Allocation; for the healthy vocal tract set any of the first Each channel, its updated weight The remaining amount will be divided according to the proportion of their original standard weights: In the formula, and The first The and the first The factory standard calibrated weights of each healthy channel when the system is fault-free.

[0065] This algorithm can strictly guarantee the sum of the weights of all channels. Furthermore, the weight of the faulty channel is substantially suppressed and will not drift due to normalization calculation.

[0066] To prevent situations where severe contamination of the entire pipe section leads to all channels being flagged as faulty, therefore... The program crashes due to an empty set causing the denominator to be zero. This embodiment includes a logical fallback mechanism: the denominator value is checked before performing division; if... If this occurs, the entire system's failure handling procedure is triggered, forcibly maintaining the previous weight configuration or outputting a specific fault code, skipping the current weight update calculation; this logic ensures... The closed loop; standard flow data is calculated based on the adjusted weights. To strictly comply with the definition of standard flow rate for gas trade metering, the system first calculates the volumetric flow rate under operating conditions. The formula is as follows: In the formula, After the above adaptive adjustment, the first The final weight value used by each channel in the current calculation cycle, which includes the updated weight if it is in a healthy state or the degraded weight if it is in a faulty state. For the first The average flow velocity of the sound channel; combined with real-time acquired operating pressure. Operating temperature and the compressibility factor of the gas mixture Convert the operating flow rate to standard reference conditions, i.e., pressure. ,temperature Standard traffic data The calculation formula is as follows: ;in, The compression factor under standard reference conditions. The compressibility factor under operating conditions is calculated by the sound velocity verification unit based on the AGA8 or ISO20765 standard. The cross-sectional area of ​​the pipe. The total number of channels configured for the flow meter. For the first The average velocity of the vocal tract.

[0067] This embodiment fully implements the technical requirements for dynamic weight adjustment in the previous embodiment. By using the sound velocity deviation index as the basic verification condition and superimposing the signal-to-noise ratio feature as an enhanced criterion, it can effectively identify and isolate faulty channels that, although the waveforms appear normal, actually have physical errors in the sound velocity measurement values, such as erroneous waveforms or clock drift, thereby further enhancing the robustness and accuracy of the metering system.

[0068] Example 6:

[0069] The acoustic feature extraction module calculates the turbulence intensity feature, which reflects the stability of the flow field. Specifically, it is used to: collect the velocity data sequence within a preset time window; calculate the normalized variance of the velocity data sequence; and use the normalized variance of the velocity as the turbulence intensity feature to characterize the degree of disturbance in the flow field.

[0070] This embodiment defines the turbulence intensity characteristics. The specific calculation method; the module within the preset time window Inside, collect a set containing Instantaneous flow velocity data sequence of a sample To clarify the instantaneous flow velocity The source of the signal, as specified in this embodiment, is obtained from the full waveform digital signal using the time difference method. Mid-flow velocity calculation: The propagation time of ultrasonic waves in the downstream and upstream directions is determined by cross-correlation algorithm or zero-crossing detection method. and And according to the formula Convert the time quantity into a flow rate quantity, where... For the length of the vocal tract, The vocal tract angle; where the time window is... The value range is set to 2 to 5 seconds, preferably 3 seconds, to cover the period of low-frequency pulsations in the flow field; the number of samples... Based on sampling frequency Decision, satisfying the relation For example, at a sampling rate of 20Hz, a 3-second window corresponds to The module performs statistical calculations to solve for the normalized variance of the velocity sequence, i.e., the turbulence intensity. To eliminate the dimension of velocity and prevent overflow caused by the denominator being zero under zero or very low velocity conditions, this embodiment introduces a numerical stabilization term into the denominator. Corrected turbulence The calculation formula is as follows: ;in, For the first One instantaneous flow velocity sample; The average flow velocity within the time window; To prevent extremely small positive numbers with a denominator of zero, this embodiment sets... ; The total number of samples; the calculated dimensionless value is used as the turbulence intensity characteristic. Output.

[0071] This embodiment captures the unsteady characteristics of the flow field by quantifying the statistical fluctuations of the flow velocity. This feature can effectively identify the flow velocity profile distortion caused by upstream rectifier failure or insufficient straight pipe section, providing users with a quantitative basis for optimizing pipeline installation conditions and ensuring that the flow meter operates in an environment that meets the requirements of fluid mechanics.

[0072] Example 7:

[0073] The metrological health assessment module is used to: invoke a preset multi-dimensional scoring logic, which includes normalization algorithms and corresponding weighting coefficients for gain, signal-to-noise ratio, waveform distortion, and turbulence; input the acoustic channel feature set into the multi-dimensional scoring logic for normalization and weighted summation to generate a metrological health index; and generate alarm signal data in response to the metrological health index falling below a preset maintenance threshold.

[0074] This embodiment illustrates the specific logic of the measurement health assessment module in generating a comprehensive score, and clarifies the application of the waveform similarity index in it; the module calls the preset multi-dimensional scoring logic, which pre-sets the normalization algorithm and weight coefficients of each feature and waveform similarity index; the acoustic channel feature set, including gain, signal-to-noise ratio, waveform distortion, turbulence, and waveform similarity index, is input into the model.

[0075] Implement the health index measurement The calculation model is as follows: ;in, In order to evaluate the total number of parameters, this embodiment ; :No. The evaluation parameters include various features in the acoustic channel feature set and waveform similarity index. Regarding the wavewidth characteristics defined in Example 2 Peak ratio characteristics .

[0076] This embodiment explicitly defines the input variables. , i.e., waveform distortion, is a weighted combination of the two deviations: ;in, and Based on reference waveform data Pre-parsed and stored reference feature values; Regarding the first A normalized penalty function for each parameter; for waveform similarity index Its penalty function is designed as follows: when hour, ;when hour, Among them, the coefficient In this embodiment, it is defined as a sensitivity constant with a value of 2000; the basis for setting this constant is to aim at reducing waveform similarity. exist The key change range is mapped to a significant decrease in the health score; specifically, when the similarity decreases from 0.95 to 0.90, the penalty item... This results in a score of 0 for that item, triggering an immediate alarm from the system.

[0077] Normalization functions for other key features to The specific calculation model disclosed in this embodiment is as follows: gain penalty ,in, Defined as the reference gain constant when the flow meter is calibrated at the factory.

[0078] Signal-to-noise ratio penalty Using sigmoid function mapping: ;in, Signal-to-noise ratio eigenvalue, in dB, threshold. The coefficient represents the inflection point of signal-to-noise ratio degradation. This is an empirical constant used to control the slope of the curve.

[0079] Waveform distortion penalty .

[0080] Turbulence penalty .

[0081] For the first The weighting coefficients of each parameter are redistributed as follows: Gain value Signal-to-noise ratio Waveform distortion turbulence Waveform similarity ,satisfy .

[0082] The system monitors this index in real time and responds accordingly. If the score falls below a preset maintenance threshold, such as 60 points, an alarm signal is generated.

[0083] This embodiment incorporates waveform similarity index into the health assessment system, making up for the inadequacy of relying solely on a single physical feature to comprehensively reflect the overall quality of the waveform. This multi-dimensional scoring logic can comprehensively capture various anomalies from signal strength and noise level to waveform morphology, providing users with a quantitative indicator that can truly reflect the overall health of the instrument, thus achieving accurate predictive maintenance.

[0084] Example 8:

[0085] The metering health assessment module is also used to: generate specific diagnostic codes to indicate transducer fouling warnings when the gain value characteristics exceed the preset normal gain range; generate specific diagnostic codes to indicate flow field vortex interference when the turbulence intensity characteristics are greater than the preset turbulence threshold; and generate specific diagnostic codes to indicate valve noise exceeding limits when the signal-to-noise ratio characteristics are less than the preset signal-to-noise ratio threshold.

[0086] This embodiment refines the mapping logic of the fault diagnosis function, establishing a correspondence between features and specific fault modes; the module internally stores a fault diagnosis code table; the system executes parallel logic judgments: responding to gain value features. If the gain exceeds the preset normal range, it is determined that the signal attenuation is severe, and a specific diagnostic code characterizing fouling or aging on the transducer surface is generated; in response to turbulence characteristics If the turbulence exceeds a preset turbulence threshold, the flow field is determined to be extremely unstable, and specific diagnostic codes characterizing upstream rectifier failure or insufficient straight pipe section are generated; this is in response to signal-to-noise ratio characteristics. If the signal-to-noise ratio is less than the preset threshold, the background noise is determined to be too high, and a specific diagnostic code characterizing the noise interference of the pressure regulating valve is generated.

[0087] This embodiment, through classification and diagnosis based on characteristic fingerprints, can not only identify equipment abnormalities, but also accurately distinguish specific causes such as fouling, flow field distortion, or noise interference. This precise fault location capability allows technicians to prepare the corresponding tools and spare parts before arriving at the site, significantly improving fault handling efficiency and reducing gas outage maintenance time.

[0088] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A full-waveform diagnostic and adaptive gas flow metering system, characterized in that, include: The full waveform acquisition module is used to digitally sample the received ultrasonic echoes to obtain a full waveform digital signal; An acoustic feature extraction module is used to perform multi-dimensional feature analysis on the full-waveform digital signal to obtain an acoustic channel feature set. The diagnostic analysis module is used to perform matching analysis on the acoustic channel feature set based on pre-stored benchmark waveform data to obtain a waveform similarity index, and to check it in conjunction with the theoretical sound velocity to obtain a sound velocity deviation index. The dynamic weighted flow calculation module is used to dynamically adjust the calculation weight of each channel according to the waveform similarity index and the sound speed deviation index, and calculate standard flow data based on the adjusted calculation weight and the full waveform digital signal. The metrological health assessment module is used to generate a metrological health index and output diagnostic status data based on the acoustic channel feature set and the waveform similarity index.

2. The full-waveform diagnostic and adaptive gas flow metering system according to claim 1, characterized in that, The acoustic feature extraction module is used for: Extract the gain value feature reflecting the signal attenuation from the full waveform digital signal; Calculate the signal-to-noise ratio characteristics that reflect the background noise level; Analysis of waveform width and peak value ratio characteristics, which reflect the degree of waveform distortion; Calculate the turbulence intensity characteristics that reflect the stability of the flow field; The gain value feature, signal-to-noise ratio feature, bandwidth feature, peak value ratio feature, and turbulence feature are combined to obtain the acoustic channel feature set.

3. The full-waveform diagnostic and adaptive gas flow metering system according to claim 2, characterized in that, The diagnostic analysis module includes: The waveform matching unit is used to call the cross-correlation algorithm to compare the current full waveform digital signal with the reference waveform data and calculate the waveform similarity index. The sound velocity verification unit is used to calculate the difference between the measured sound velocity based on the full waveform digital signal and the theoretical sound velocity calculated based on the current operating condition parameters, so as to obtain the sound velocity deviation index.

4. The full-waveform diagnostic and adaptive gas flow metering system according to claim 3, characterized in that, The sound speed verification unit is configured as follows: Obtain the current pressure, temperature, and gas composition parameters; The theoretical speed of sound is calculated based on the pressure parameters, temperature parameters, and gas composition parameters. In response to the sound speed deviation index exceeding a preset safety threshold, warning data indicating a risk of measurement inaccuracy is generated; In response to the sound speed deviation index being less than or equal to the safety threshold, confirmation data indicating normal measurement is generated.

5. The full-waveform diagnostic and adaptive gas flow metering system according to claim 1, characterized in that, The dynamic weighted flow calculation module is used for: In a multi-channel system, the waveform similarity index and signal-to-noise ratio characteristics of each channel are acquired in real time. In response to a certain channel having a waveform similarity index lower than a preset quality threshold or a signal-to-noise ratio feature lower than a preset noise threshold, the calculated weight of that channel is set to a preset degraded weight value that is lower than the standard weight. The standard flow rate data is obtained by weighted averaging the flow rate data of each channel based on the adjusted channel calculation weights.

6. The full-waveform diagnostic and adaptive gas flow metering system according to claim 2, characterized in that, The calculation of turbulence intensity features, which reflect the stability of the flow field, in the acoustic feature extraction module is specifically used for: Collect flow velocity data sequences within a preset time window; Calculate the normalized variance of the flow velocity data sequence; The normalized variance of the flow velocity is used as the turbulence intensity feature to characterize the degree of disturbance in the flow field.

7. The full-waveform diagnostic and adaptive gas flow metering system according to claim 2, characterized in that, The measured health assessment module is used for: The preset multidimensional scoring logic is invoked, which includes normalization algorithms for gain value, signal-to-noise ratio, waveform distortion degree and turbulence degree and corresponding weight coefficients; The acoustic channel feature set is input into the multidimensional scoring logic for normalization and weighted summation to generate the measurement health index; When the measured health index falls below a preset maintenance threshold, an alarm signal is generated.

8. The full-waveform diagnostic and adaptive gas flow metering system according to claim 7, characterized in that, The measurement health assessment module is also used for: In response to the gain value characteristic exceeding the preset normal gain range, a specific diagnostic code characterizing the transducer fouling warning is generated; In response to the turbulence intensity characteristic being greater than a preset turbulence threshold, a specific diagnostic code characterizing the vortex disturbance in the flow field is generated; In response to the signal-to-noise ratio feature being less than a preset signal-to-noise ratio threshold, a specific diagnostic code characterizing excessive valve noise is generated.

Images