Non-straight pipeline gas flow velocity measuring system

Abstract

The invention relates to the technical field of flow velocity measurement, in particular to a non-straight pipeline gas flow velocity measurement system, comprising: an acoustic detection module comprising a plurality of acoustic sensors arranged outside a target pipeline section and used for forming an acoustic detection path outside the pipeline; the signal processing module is used for receiving the acoustic signal sent by the acoustic detection module, performing amplitude analysis and transit time calculation on the acoustic signal, and obtaining gas flow rate path integral measurement data of each path; the flow field reconstruction module is used for performing inversion calculation on path integral measurement data by adopting an improved acoustic tomography algorithm fused with a non-straight pipeline turbulence model, and reconstructing a two-dimensional flow velocity distribution field of the cross section to be measured; and the flow field analysis module is used for performing intelligent data analysis on the two-dimensional flow velocity distribution field and extracting and outputting flow velocity key information. The technical problem of providing reliable section flow velocity distribution information in a non-straight pipeline in the prior art is solved.

Description

Technical Field

[0001] This application relates to the field of flow velocity measurement technology, and in particular to a gas flow velocity measurement system for non-straight pipes. Background Technology

[0002] In industrial production, energy transmission, and environmental monitoring, accurate measurement of gas velocity within pipelines is crucial. This is especially true when pipelines have non-straight sections such as bends, diameter changes, or confluences, where the internal flow field exhibits complex spatial distribution characteristics due to inertia, centrifugal force, and wall effects, such as asymmetrical velocity profiles, secondary flows, and even vortex structures. Accurately understanding the two-dimensional velocity distribution at such cross-sections is of great significance for assessing flow losses, optimizing system energy efficiency, and ensuring equipment safety.

[0003] Currently, existing technologies for measuring gas velocity in pipelines mainly rely on single-point measurements or line-averaging measurements based on specific flow assumptions. Specifically, one common technique involves using invasive probes such as Pitot tubes and thermal anemometers. This method requires extending the probe into the flow field, which not only interferes with or even alters the structure of the flow field being measured, making it difficult to reflect the true state, but also only obtains the velocity at a local point at the probe's location, failing to efficiently and unobstructedly acquire velocity distribution information across the entire cross-section. Another widely used technique is the ultrasonic time-of-flight flowmeter. This technique relies on the time difference between the propagation of sound waves in the forward and reverse directions to calculate the average velocity along the acoustic path; the measurement result is essentially a line integral value along an acoustic path. To obtain the cross-sectional average velocity and then calculate the volumetric flow rate, this technique typically assumes that the velocity distribution within the pipe cross-section is axisymmetric and regular. However, in non-straight pipes, due to the aforementioned complex flow phenomena, the velocity distribution deviates significantly from the regularity assumption. Therefore, ultrasonic time-of-flight methods based on such assumptions introduce significant measurement errors and cannot provide reliable cross-sectional velocity distribution information.

[0004] Therefore, there is an urgent need for a non-straight pipe gas velocity measurement system. Summary of the Invention

[0005] (a) Technical problems to be solved

[0006] In view of the above-mentioned shortcomings and deficiencies of the prior art, this application provides a non-straight pipe gas velocity measurement system, which solves the technical problem of providing reliable cross-sectional velocity distribution information in non-straight pipes in the prior art.

[0007] (II) Technical Solution

[0008] To achieve the above objectives, the main technical solutions adopted in this application include:

[0009] This application provides a non-straight pipe gas flow velocity measurement system, including:

[0010] The acoustic detection module includes multiple acoustic sensors deployed outside the target pipe section. The multiple acoustic sensors are paired in pairs to form multiple sensor pairs, which are used to form multiple acoustic detection paths outside the pipe that cross the same cross section inside the pipe from different spatial angles.

[0011] The signal processing module is used to receive the acoustic signal emitted by the acoustic detection module, perform amplitude analysis and transit time calculation on the acoustic signal, and obtain the gas flow velocity path integral measurement data for each path.

[0012] The flow field reconstruction module is used to perform inversion calculations on the path integral measurement data using an improved acoustic tomography algorithm that integrates a non-straight pipe turbulence model, and reconstruct the two-dimensional velocity distribution field of the cross section to be measured.

[0013] The flow field analysis module is used to perform intelligent data analysis on the two-dimensional velocity distribution field, extract and output key velocity information; the key velocity information includes: statistical characteristic values ​​of velocity covering a specified area within the cross-section to be measured and a visual cloud map of velocity distribution.

[0014] Optionally, in some embodiments of this application, the plurality of acoustic sensors are arranged on the outside of the target pipe section using an adjustable mounting bracket. The plurality of acoustic sensors are arranged in a ring array, the center of which coincides with the pipe axis, and the angle between adjacent sensors is less than or equal to 15°.

[0015] Optionally, in some embodiments of this application, the signal processing module includes:

[0016] The signal conditioning unit is used to filter and amplify the acoustic signal received by the acoustic sensor to obtain the preprocessed signal.

[0017] The data acquisition unit is used to convert the preprocessed signal into a digital signal;

[0018] The transit time calculation unit is used to analyze the digital signal using a cross-correlation algorithm, calculate the transit time of the sound wave propagating along each sensor in the downstream and upstream directions respectively, and obtain the transit time difference;

[0019] The path integral calculation unit is used to obtain the gas velocity path integral measurement data on each acoustic detection path based on the transit time difference and the corresponding spatial geometric relationship, according to the principle of acoustic time difference method.

[0020] Optionally, in some embodiments of this application, the transit time calculation unit is specifically used for:

[0021] For each sensor pair, perform the following steps:

[0022] Acquire both forward-flow and reverse-flow digital signals;

[0023] Calculate the first cross-correlation function between the downstream digital signal and the standard reference signal, and the second cross-correlation function between the upstream digital signal and the standard reference signal, respectively.

[0024] The main peak values ​​of the first cross-correlation function and the second cross-correlation function are detected respectively to determine the downstream transit time and the upstream transit time.

[0025] The crossing time difference is obtained based on the downstream crossing time and the upstream crossing time.

[0026] Optionally, in some embodiments of this application, the flow field reconstruction module includes:

[0027] The flow field initialization unit is used to establish an initial two-dimensional velocity distribution field based on the geometric structure parameters and fluid property parameters of the target pipe section.

[0028] The model fusion unit is used to embed the non-straight pipe turbulence model into the acoustic tomography inversion algorithm in the form of prior constraints. The non-straight pipe turbulence model is used to describe the flow field asymmetry and vortex characteristics caused by pipe bending or structural abrupt changes.

[0029] The iterative inversion unit is used to take the path integral measurement data as input and use an acoustic tomography inversion algorithm with an embedded turbulence model to iteratively correct the initial two-dimensional velocity distribution field until the error between the theoretical path integral value and the actual measurement data corresponding to the reconstructed two-dimensional velocity distribution field is less than a set threshold, and outputs the final reconstructed two-dimensional velocity distribution field.

[0030] Optionally, in some embodiments of this application, the iterative inversion unit is specifically used to perform inversion calculations through the following steps:

[0031] The iterative process begins based on the initial two-dimensional velocity distribution hypothesis field and the path integral measurement data; in each iteration step, the following operations are performed:

[0032] Based on the estimated value of the two-dimensional velocity distribution field in the current iteration, a forward calculation is performed to obtain the corresponding theoretical path integral value set; the theoretical path integral value set is compared with the path integral measurement data, and the data fitting residual is calculated;

[0033] Based on the prior constraints of the non-straight pipe turbulence model, the correction amount of the current two-dimensional velocity distribution field estimate is obtained through optimization algorithm, and the estimate of the two-dimensional velocity distribution field is updated.

[0034] The iteration process is repeated until the data fitting residual is less than a set threshold or the number of iterations reaches a preset upper limit, and the estimated two-dimensional velocity distribution obtained at this time is output as the final reconstructed two-dimensional velocity distribution field.

[0035] Optionally, in some embodiments of this application, the flow field analysis module includes:

[0036] The flow field feature extraction unit is used to identify and quantify specific flow field features caused by the non-straight section structure of the pipe from the two-dimensional velocity distribution field; the specific flow field features include: the asymmetry index of the velocity distribution, the center position and intensity of one or more vortex structures, and the actual flow direction angle synthesized by the main flow and secondary flow;

[0037] The integrated diagnosis and output unit is used to obtain the flow statistics of a specified area within the cross section to be measured based on the two-dimensional velocity distribution field and the specific flow field characteristics; at the same time, it generates a velocity distribution visualization cloud map that integrates the annotations of the specific flow field characteristics.

[0038] Optionally, in some embodiments of this application, the integrated diagnosis and output unit is used to obtain the flow statistics of a specified area within the cross-section to be measured based on the two-dimensional velocity distribution field and the specific flow field characteristics, specifically including:

[0039] Based on the actual flow direction angle in the specific flow field characteristics, the two-dimensional planar velocity vector of each grid cell in the two-dimensional velocity distribution field is converted and synthesized into the axial velocity component of each grid cell along the pipe axis, thereby generating the axial velocity component distribution field of the cross section to be measured.

[0040] Based on the specific flow field characteristics, the designated region is defined within the cross section to be measured. The designated region includes: the core flow region indicated by the asymmetry index, the vortex influence region centered on the center position of the vortex structure, and the secondary flow significant region defined by the actual flow direction angle.

[0041] Based on the axial velocity component distribution field, the velocity statistical feature values ​​of each specified region are extracted and calculated. The velocity statistical feature values ​​include: the region average axial velocity, the region axial velocity standard deviation, and the region maximum reverse velocity.

[0042] Optionally, in some embodiments of this application, the integrated diagnosis and output unit is further used for:

[0043] When the asymmetry index exceeds the first threshold, the core flow region containing the mainstream high-speed region and the corresponding low-speed return region are defined as the designated region.

[0044] When the intensity of the vortex structure exceeds the second threshold, the vortex influence area within a preset radius is defined as the designated area, with the vortex center as the center.

[0045] Based on the spatial distribution of the actual flow direction angle within the cross-section, the region where the actual flow direction angle deviates from the pipe axis by an angle greater than the third threshold is defined as the secondary flow significant region as the designated region.

[0046] Optionally, in some embodiments of this application, the system further includes: a real-time adaptive calibration module, the real-time adaptive calibration module comprising:

[0047] The model-flow field matching degree evaluation unit is used to continuously monitor the time-varying characteristics of the two-dimensional velocity distribution field output by the flow field reconstruction module, and dynamically evaluate the matching degree between the non-straight pipe turbulence model used in the current inversion calculation and the current actual flow field.

[0048] The online model parameter optimization unit is used to trigger an online calibration process when the matching degree is lower than a preset threshold, and to perform inverse problem optimization on the key model parameters in the non-straight pipe turbulence model to obtain the optimized model parameters; the key model parameters include one or more of the following: boundary layer velocity profile exponent, turbulent eddy viscosity coefficient, and secondary flow intensity coefficient.

[0049] The model parameter dynamic update unit is used to feed back the optimized model parameters to the flow field reconstruction module in real time to update the embedded prior constraints.

[0050] (III) Beneficial Effects

[0051] The beneficial effects of this application are as follows: The non-straight pipe gas velocity measurement system of this application, by adopting a detection method with externally deployed multi-angle intersecting acoustic paths and integrating a non-straight pipe turbulence model and an improved acoustic tomography algorithm for flow field inversion and reconstruction, can achieve high-precision, non-contact measurement of two-dimensional velocity distribution in complex curved or non-straight pipe sections without intruding into the pipe interior or interfering with the flow field, compared with the prior art. It achieves the effect of accurately obtaining the statistical characteristics of velocity in a specified area and the visualization of the entire field distribution cloud map, effectively solving the technical problem that traditional single-point measurement or internal probe methods are difficult to accurately obtain global information of complex turbulent flow fields in non-straight pipes. Attached Figure Description

[0052] Figure 1 This is a schematic diagram of a non-straight pipe gas velocity measurement system according to an embodiment of this application;

[0053] Figure 2 This is a layout diagram of multiple sensors in a non-straight pipe gas flow velocity measurement system according to an embodiment of this application;

[0054] Figure 3 This is an internal flowchart of the real-time adaptive calibration module of a non-straight pipe gas velocity measurement system according to an embodiment of this application. Detailed Implementation

[0055] To better explain and facilitate understanding of this application, the following detailed description of the application is provided in conjunction with the accompanying drawings and specific embodiments.

[0056] In existing technologies, measuring gas velocity inside non-straight pipes typically faces significant challenges. Mainstream methods fall into two categories: First, invasive point measurements, such as using probes like Pitot tubes or hot-film anemometers to measure specific points inside the pipe. This method interferes with the original flow field and only obtains discrete point data, failing to reflect the complex turbulent structure of the entire cross-section (such as separated vortices and secondary flows). Second, ultrasonic or optical measurements based on the assumption of a straight pipe, such as monochannel ultrasonic flowmeters or laser Doppler velocimetry. These methods perform well in straight pipe sections, but their measurement models and algorithms heavily rely on the assumption of uniform and symmetrical fluid flow along the pipe axis. In non-straight pipes, due to the presence of strong three-dimensional turbulence and secondary flows, this assumption fails, leading to a sharp increase in measurement error and even failing to provide a meaningful cross-sectional velocity distribution.

[0057] Therefore, existing technologies lack an effective means to accurately and comprehensively measure the two-dimensional velocity distribution within a complex turbulent cross section of a non-straight pipe without intruding into or interfering with the flow field.

[0058] To address this, this application proposes a gas velocity measurement system for non-straight pipes, particularly suitable for monitoring gas velocity in critical process pipelines such as deodorization pipes and aeration pipes in confined spaces like underground sewage treatment plants. The system utilizes multiple pairs of acoustic sensors deployed outside the pipe, intersecting the same cross-section from different spatial angles, forming a spatial coverage network to acquire velocity integral information along multiple paths in a non-contact manner. Then, an improved acoustic tomography inversion algorithm incorporating a unique turbulence physical model specific to non-straight pipes is used to process this path integral data. This technical solution overcomes the traditional method's assumption of relying on the flow field in straight pipes, accurately reconstructing the complete two-dimensional velocity distribution field within the cross-section of a non-straight pipe from limited path measurement data. Ultimately, it achieves intelligent extraction of the statistical characteristics of velocity in a specified area within the cross-section and visualized output of the overall velocity distribution, providing a revolutionary measurement tool for flow field diagnosis and optimization of complex pipeline systems.

[0059] To better understand the above technical solutions, exemplary embodiments of this application will be described in more detail below with reference to the accompanying drawings. Although exemplary embodiments of this application are shown in the drawings, it should be understood that this application can be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this application can be understood more clearly and thoroughly, and that the scope of this application can be fully conveyed to those skilled in the art.

[0060] Figure 1 This is a schematic diagram of a non-straight pipe gas velocity measurement system according to an embodiment of this application. Figure 1 As shown, the non-straight pipe gas velocity measurement system includes:

[0061] The acoustic detection module includes multiple acoustic sensors deployed outside the target pipe section. The multiple acoustic sensors are paired in pairs to form multiple sensor pairs, which are used to form multiple acoustic detection paths outside the pipe that cross the same cross section inside the pipe from different spatial angles.

[0062] Multiple acoustic sensors are mounted on the outside of the target pipe section using adjustable mounting brackets. The sensors are arranged in a ring array, with the center of the array coinciding with the pipe axis, and the angle between adjacent sensors being less than or equal to 15° (e.g., ...). Figure 2 (As shown).

[0063] Specifically, the adjustable mounting bracket is designed with an angle and distance fine-tuning structure and a locking function. It is designed to overcome engineering challenges such as the curved surface of the pipe outer wall and the limitation of on-site installation space, ensuring that each sensor can be closely attached to or aligned with the pipe outer wall at its optimal working angle and spacing. At the same time, it ensures that the position remains stable under on-site environments such as vibration and temperature difference, thereby maintaining the long-term reliability of the measurement reference.

[0064] Multiple acoustic sensors are arranged in an optimized spatial array configuration. In a preferred embodiment of this application, they are arranged in a dense ring array on the outer periphery corresponding to the cross-section of the pipe to be measured. The geometric center of this ring array is precisely calibrated to ensure that it coincides with the theoretical axis of the pipe at that cross-section, thereby constructing a symmetrical measurement coordinate system based on the pipe axis. This is an important spatial constraint for the accuracy of subsequent flow field inversion algorithms.

[0065] This dense array arrangement ensures the formation of a sufficient number (N≥8 pairs, preferably 12 or more) acoustic detection paths outside the pipe, intersecting from different spatial angles and passing through the same cross-section inside the pipe. These paths form a dense, intersecting measurement network within the cross-section, much like performing a "CT scan" on the fluid, thus acquiring sufficient data for subsequent high-precision reconstruction of the two-dimensional velocity distribution field.

[0066] The aforementioned high-density, multi-angle acoustic path network provides spatially sufficient and informationally complete observation data for subsequent acoustic tomography inversion, serving as the data foundation for achieving high-precision flow field reconstruction.

[0067] The signal processing module is used to receive the acoustic signals emitted by the acoustic detection module, perform amplitude analysis and transit time calculation on the acoustic signals, and obtain the gas velocity path integral measurement data for each path.

[0068] The signal processing module includes:

[0069] The signal conditioning unit is used to filter and amplify the acoustic signal received by the acoustic sensor to obtain the preprocessed signal.

[0070] The data acquisition unit is used to convert the preprocessed signal into a digital signal;

[0071] The transit time calculation unit is used to analyze digital signals using a cross-correlation algorithm, calculate the transit time of the sound wave propagating along each sensor in the downstream and upstream directions respectively, and obtain the transit time difference.

[0072] The transit time calculation unit is specifically used for:

[0073] For each sensor pair, perform the following steps:

[0074] Acquire both forward-flow and reverse-flow digital signals;

[0075] Calculate the first cross-correlation function between the forward-flow digital signal and the standard reference signal, and the second cross-correlation function between the reverse-flow digital signal and the standard reference signal, respectively.

[0076] The main peak values ​​of the first cross-correlation function and the second cross-correlation function are detected respectively to determine the downstream transit time and the upstream transit time.

[0077] The time difference between crossing the current and crossing it upstream is obtained.

[0078] For example, the acoustic environment is particularly harsh when monitoring the flow velocity in critical process pipelines (such as sludge return pipes and aeration pipes) of underground wastewater treatment plants. The continuous vibration of pumps and mixing equipment, the complex flow of multiphase media (solid, liquid, and gas) in the fluid, and the uneven adhesion of biofilm and sediment to the inner wall of the pipes all contribute to high-intensity, wide-spectrum background noise and severe sound wave scattering attenuation. Under such conditions, traditional ultrasonic measurement methods are highly susceptible to failure in transit time calculation or significant errors due to the effective signal being submerged by noise and the sound wave path being distorted, making reliable measurements difficult to achieve.

[0079] At this point, the signal conditioning unit employs a hybrid processing strategy combining adaptive filtering and lock-in amplification. First, a programmable bandpass filter bank (center frequency 100kHz, bandwidth ±5kHz) filters out low-frequency components of pump vibration (<500Hz) and high-frequency power line interference (>1MHz). Simultaneously, an adaptive notch filter is designed to track and eliminate interference at specific mechanical resonant frequencies in real time. The amplification stage uses a programmable gain amplifier and an automatic gain control loop to achieve stable amplification with a dynamic range of 120dB under weak nanovolt-level signals, ensuring signal amplitude consistency. A specially designed temperature compensation circuit compensates for changes in sensor sensitivity with ambient temperature, guaranteeing measurement stability throughout the year.

[0080] The data acquisition unit employs a multi-channel synchronous sampling ADC architecture, simultaneously acquiring signals from all sensor channels at a sampling rate of 10MS / s and a resolution of 16 bits, with a synchronization error between channels of less than 100ps. Its unique clock tree distribution design and shielded drive technology effectively suppress channel crosstalk to below -90dB. The data acquisition unit also incorporates a first-in-first-out buffer and a real-time data verification mechanism to ensure that no valid acoustic pulse data is lost during continuous monitoring.

[0081] Furthermore, the transit time calculation unit extracts the digital signal waveforms acquired by the current sensor in both the forward and reverse flow directions. By performing synchronous DC bias correction and consistent windowing preprocessing on these two digital signals, a foundation is laid for subsequent accurate comparison. Subsequently, the crucial generalized cross-correlation calculation stage begins. In this stage, the system calculates the cross-power spectrum of the forward and reverse flow digital signals and the reference waveform of the standard transmitted signal, based on the digital signals provided by the data acquisition unit, and applies phase transformation weighting. This unique step effectively suppresses frequency-related environmental noise and pipe structure reverberation interference in the digital signals, highlighting the pure phase information contained in the signals, thereby generating two cross-correlation function curves with exceptionally sharp main peaks, corresponding to the forward and reverse flow propagation, respectively.

[0082] Next, the transit time calculation unit automatically identifies the global dominant peak on the cross-correlation function curve and filters out false peaks caused by echoes or sudden noises using strict peak shape criteria. After locking onto the true dominant peak, the system uses parabolic curves or more precise mathematical functions to perform fitting interpolation between discrete sampling points, thereby calculating the time position corresponding to the peak with an accuracy far exceeding that of a single sampling interval, i.e., achieving subsampling interpolation. Thus, high-precision downstream and upstream transit times are determined. Furthermore, this process includes validity self-checking; if two time values ​​show a physically unreasonable difference, the system will initiate a retest or call a backup intelligent algorithm for arbitration.

[0083] Finally, after calculating the transit time difference, the transit time calculation unit automatically subtracts the inherent hardware delay bias of the specific sensor measured under zero-flow static conditions from the pre-stored calibration database. This step eliminates systematic errors caused by electronic asymmetry in the data acquisition chain and individual sensor differences, ultimately outputting a pure, effective transit time difference solely due to fluid flow. This highly purified time difference data is the most fundamental and reliable input for subsequent reconstruction of the cross-sectional velocity distribution field, and its accuracy directly determines the final measurement performance of the entire system.

[0084] The signal processing module in this embodiment also includes:

[0085] The path integral calculation unit is used to obtain the gas velocity path integral measurement data on each acoustic detection path based on the transit time difference and the corresponding spatial geometric relationship, according to the principle of acoustic time difference method.

[0086] Specifically, the spatial geometric relationships are pre-established and stored in the system through high-precision calibration of the sensor array. During installation, the system uses a laser tracker or photogrammetry system to accurately determine the absolute coordinates of each acoustic sensor in three-dimensional space. Combining this with the three-dimensional CAD model of the target pipeline or the actual pipeline point cloud data obtained through laser scanning, the precise spatial parameters of the acoustic detection path formed by each pair of sensors are calculated. These parameters include: the effective sound path length L inside the pipeline (i.e., the actual distance the sound wave travels in the fluid, compensated for by the pipeline wall thickness and refraction effects), and the angle θ between the path and the pipeline axis projected onto the cross-section under test (i.e., the angle between the path direction and the mainstream direction), etc.

[0087] In this application, based on the principle of acoustic time-of-flight (TOF) method, the gas velocity path integral measurement data for each acoustic detection path are obtained specifically as follows: First, based on precisely calibrated spatial geometric relationships, the effective length and direction of each acoustic path are determined; then, the actual average sound velocity of the fluid in the current state is calculated by combining the propagation time in both directions; finally, the time difference, path length, and sound velocity are substituted into the core relationship of the TOF method for calculation, and the output result is the average value of the line integral of the gas velocity vector along the corresponding acoustic path direction. This value fully characterizes the average effect of the velocity projection along the spatial straight line, providing the most direct physical observation data for subsequent flow field tomography inversion.

[0088] The formula for the acoustic time difference method is:

[0089] ;

[0090] Among them, V pathThis is the average line integral of the gas velocity vector along the corresponding acoustic path direction, which is still the average velocity component along the path direction. c is the actual sound velocity in the fluid, dynamically calculated from the average propagation time in both the forward and reverse directions. L is the effective sound path length of this acoustic path. This refers to the measured transit time difference.

[0091] The aforementioned signal processing module, through a multi-stage precision processing chain of "signal conditioning - data acquisition - time calculation - path integration," transforms the weak analog signals received by the acoustic sensor into path integral data reflecting flow velocity information. This module achieves nanosecond-level transit time difference measurement in noisy environments using a high-precision cross-correlation algorithm and subsampling interpolation technology. Combined with a time-difference method model with dynamic sound velocity compensation, it accurately calculates the flow velocity line integral values ​​along each acoustic path, providing a high signal-to-noise ratio and high-reliability underlying data foundation for subsequent flow field reconstruction. This ensures the measurement accuracy and stability of the entire system in complex industrial environments.

[0092] The non-straight pipe gas velocity measurement system in this application embodiment also includes:

[0093] The flow field reconstruction module is used to perform inversion calculations on path integral measurement data using an improved acoustic tomography algorithm that integrates a non-straight pipe turbulence model, and reconstruct the two-dimensional velocity distribution field of the cross section to be measured.

[0094] The flow field reconstruction module includes:

[0095] The flow field initialization unit is used to establish an initial two-dimensional velocity distribution field based on the geometric parameters and fluid property parameters of the target pipe section.

[0096] Based on the geometric and fluid property parameters of the target pipe section, the initial two-dimensional velocity distribution field is established, specifically including:

[0097] By inputting the geometric and fluid property parameters of the target pipe section into an embedded, validated parametric empirical model library or performing rapid simplified CFD pre-calculation, a two-dimensional velocity field with a pre-defined typical flow structure is intelligently generated. For example, for a 90-degree bend, the system will automatically generate a characteristic velocity profile based on the Dean number, where the velocity peak region deviates from the geometric center and is biased towards the outside of the bend, and a pair of symmetrical secondary flow vortices (i.e., Dean vortices) have been initially formed. This initial field is not simply a uniform or axisymmetric distribution, but directly reflects the core physical mechanism that "pipe bending inevitably leads to flow field distortion and secondary flow," thus providing a high-order, physically interpretable iterative starting point for subsequent acoustic tomography inversion algorithms that integrate turbulence models. This effectively overcomes the ill-posedness of the inversion problem and significantly improves the accuracy and computational efficiency of complex flow field reconstruction.

[0098] The geometric parameters of the target pipeline section are typically derived from design drawings or 3D geometric information reconstructed through on-site 3D laser scanning. These parameters include pipeline type (e.g., elbows, tees), bending radius, cross-sectional shape and dimensions, and flow direction change angle. These parameters determine the spatial constraints and boundary conditions of the flow field. Fluid property parameters include the fluid's density and viscosity. These properties can be calculated based on known fluid types (e.g., air, natural gas) and real-time temperature and pressure data provided by additional sensors, by querying a built-in property database or state equations.

[0099] The model fusion unit is used to embed the non-straight pipe turbulence model into the acoustic tomography inversion algorithm in the form of prior constraints. The non-straight pipe turbulence model is used to describe the flow field asymmetry and vortex characteristics caused by pipe bending or structural abrupt changes.

[0100] Specifically, the model fusion unit transforms the non-straight pipe turbulence model into mathematical constraints on the inversion solution space. These constraints include: regularization constraints: adding terms based on turbulent energy spectrum or gradient priors to the objective function to penalize high-frequency oscillations or non-smooth velocity fields that do not conform to physical laws; parameterization constraints: representing the velocity field as a linear combination of several basis functions conforming to the secondary flow structure of a curved pipe (such as the modes of the Dean vortex), transforming the inversion problem into a better-state problem of solving for the coefficients of these basis functions; and hard boundary constraints: directly limiting the asymmetry index of the velocity distribution, the range of vortex center locations, etc. These constraints are treated as prior knowledge and, together with the measurement data, constitute the objective function of the inversion algorithm.

[0101] Taking the 90-degree bend at the outlet of the sludge return pump as an example, the fluid flowing inside is activated sludge with a solids content of approximately 2%. This fluid has high viscosity, complex properties, and its flow is affected by pump pulses. In this case, the model fusion unit first selects a corresponding analysis template from a model library specifically designed for wastewater flow, based on the pipe shape and real-time estimated parameters such as velocity and density. This template indicates that when this high-viscosity sludge flows through the bend, the region with the fastest flow velocity will be biased towards the outside of the pipe, but the deviation will be smaller than that of water flow; simultaneously, due to the influence of solid particles, the secondary vortices formed inside the pipe will exhibit an asymmetrical structure, stronger at the top and weaker at the bottom.

[0102] In the inversion calculation, the system transforms the above physical characteristics into computational constraints: first, it restricts the velocity distribution that does not conform to the characteristics of sludge flow; second, it limits the solution process to a set of typical sewage bend flow patterns, which already include special vortex structures caused by solid-liquid two-phase separation.

[0103] In this way, the algorithm not only matches the limited acoustic measurement data during iteration, but is also guided to converge toward results that conform to the actual sludge flow pattern, making the reconstructed flow field closer to the real situation.

[0104] In the specific implementation process, the physically interpretable initial field generated by the flow field initialization unit, which contains typical characteristics such as secondary flow in bends, is combined with the turbulence prior constraints provided by the model fusion unit to form a powerful physical guide for the inversion algorithm. This ensures that the iteration starts from a reasonable starting point and converges in a direction that conforms to the laws of fluid mechanics, thereby significantly improving the accuracy, speed and stability of reconstructing complex flow fields with limited data.

[0105] The flow field reconstruction module in this embodiment also includes:

[0106] The iterative inversion unit is used to take path integral measurement data as input and use an acoustic tomography inversion algorithm with embedded turbulence model to iteratively correct the initial two-dimensional velocity distribution field until the error between the theoretical path integral value and the actual measurement data corresponding to the reconstructed two-dimensional velocity distribution field is less than a set threshold, and outputs the final reconstructed two-dimensional velocity distribution field.

[0107] The iterative inversion unit is specifically used to perform inversion calculations through the following steps:

[0108] The iterative process begins based on the initial two-dimensional velocity distribution assumption field and path integral measurement data; in each iteration step, the following operations are performed:

[0109] Based on the estimated value of the two-dimensional velocity distribution field in the current iteration, a forward calculation is performed to obtain the corresponding set of theoretical path integral values; the set of theoretical path integral values ​​is compared with the path integral measurement data, and the data fitting residual is calculated.

[0110] By combining the prior constraints of the non-straight pipe turbulence model, the correction amount of the current two-dimensional velocity distribution field estimate is obtained through optimization algorithm, and the estimate of the two-dimensional velocity distribution field is updated.

[0111] Repeat the iterative process until the data fitting residual is less than the set threshold or the number of iterations reaches the preset upper limit, and then use the estimated value of the two-dimensional velocity distribution obtained at this time as the output of the final reconstructed two-dimensional velocity distribution field.

[0112] The iterative inversion unit uses path integral measurement data as an objective benchmark and an optimization algorithm incorporating turbulence models as its core engine. In each iteration, it mandates that the reconstructed velocity field must simultaneously meet two conditions: first, its theoretical predictions must continuously approach the measured data to minimize residuals; second, its spatial structure must conform to the physical prior laws of non-straight pipe flow. This dual constraint ensures that the reconstruction process will not fall into mathematical local optima or produce absurd solutions that violate the common sense of fluid mechanics. Thus, it can stably and efficiently output high-fidelity two-dimensional velocity distribution fields in industrial environments with limited data and significant noise, achieving an accurate transformation from sparse and indirect line integral measurements to reliable and intuitive full-field visualization.

[0113] The non-straight pipe gas velocity measurement system in this application embodiment also includes:

[0114] The flow field analysis module is used to perform intelligent data analysis on the two-dimensional velocity distribution field, extract and output key velocity information, including: statistical characteristic values ​​of velocity covering a specified area within the cross-section to be measured, and a visual cloud map of velocity distribution.

[0115] The flow field analysis module includes:

[0116] The flow field feature extraction unit is used to identify and quantify specific flow field features caused by the non-straight section structure of the pipe from the two-dimensional velocity distribution field. The specific flow field features include: the asymmetry index of the velocity distribution, the center position and intensity of one or more vortex structures, and the actual flow direction angle synthesized by the main flow and secondary flow.

[0117] The asymmetry index for flow field distribution is determined by comparing the average velocity or momentum flux in symmetrical regions about the geometric center or a specific axis within a cross section. Specifically, the cross section is divided into multiple symmetrical sectors or pairs of symmetrical points, the statistical differences in velocity values ​​within the corresponding regions (such as root mean square difference or maximum deviation) are calculated, and these differences are normalized to a dimensionless index between 0 and 1. This index is the asymmetry index.

[0118] For the identification and quantification of vortex structures, the element employs a method combining vortex dynamics and vector field topology analysis. First, the vorticity distribution of the velocity field is calculated. Then, by finding vorticity extrema and combining them with the rotational direction of the surrounding velocity vectors, the core locations of one or more secondary flow vortices are automatically located. For each identified vortex core, the element defines a physically scaled closed path around it, and the intensity of the vortex is accurately quantified by calculating the circulation of the velocity vector along this path. Simultaneously, by analyzing the spatial distribution of the vorticity field, the influence range (equivalent radius) of each vortex can be estimated. These parameters comprehensively describe the spatial location, rotational intensity, and size of the secondary flow vortices.

[0119] The actual flow direction angle is calculated by vector synthesis. At each computational grid point within the cross section or along a specified line of interest, the axial velocity component and the transverse (radial / circumferential) velocity component are synthesized to obtain the resultant velocity vector at that point. The angle between this vector and the pipe axis (or local tangential direction) is then calculated. By statistically analyzing the flow direction angle distribution across the entire cross section or a specific region (such as the near-wall region or the central region), flow direction angle contour maps or distribution curves can be generated, clearly revealing how the secondary flow distorts the mainstream direction, forming a complex three-dimensional helical flow structure.

[0120] The integrated diagnostic and output unit is used to obtain the flow statistics of a specified area within the cross section to be measured based on the two-dimensional velocity distribution field and specific flow field characteristics; at the same time, it generates a visual cloud map of velocity distribution that integrates specific flow field feature annotations.

[0121] The integrated diagnostic and output unit is used to obtain flow statistics for a specified region within the cross-section under test based on a two-dimensional velocity distribution field and specific flow field characteristics. Specifically, it includes:

[0122] Based on the actual flow direction angle in the specific flow field characteristics, the two-dimensional planar velocity vector of each grid cell in the two-dimensional velocity distribution field is transformed and synthesized into the axial velocity component of each grid cell along the pipe axis, thereby generating the axial velocity component distribution field of the cross section to be measured.

[0123] Based on specific flow field characteristics, a designated region is defined within the cross section to be measured. The designated region includes: the core flow region indicated by the asymmetry index, the vortex influence region centered on the center position of the vortex structure, and the significant secondary flow region defined by the actual flow direction angle.

[0124] Based on the axial velocity component distribution field, the velocity statistical characteristic values ​​of each specified region are extracted and calculated. The velocity statistical characteristic values ​​include: the region's average axial velocity, the region's standard deviation of axial velocity, and the region's maximum reverse velocity.

[0125] Furthermore, when generating the visualization cloud map, the integrated diagnosis and output unit performs the following steps: Based on the grid data of the reconstructed two-dimensional velocity distribution field or axial velocity component distribution field, a scientifically calculated color scheme, such as a rainbow color system or gradient color system, is used to color map the velocity magnitude, generating a core velocity distribution color cloud map. On top of this cloud map, the system automatically overlays key physical feature annotations identified by the flow field feature extraction unit: for example, using contour lines or specific color bands to highlight the main high-speed region and the low-speed backflow region; using striking symbols or animated vortex lines to clearly mark the center position and rotation direction of each secondary flow vortex; and simultaneously, using semi-transparent fills or specific boundary lines to delineate the range of the "significant secondary flow region." Finally, the system integrates and generates a comprehensive flow field visualization cloud map that integrates velocity distribution, feature regions, and key physical structure annotations, and supports output in the form of high-resolution images or dynamic interactive charts, providing intuitive and comprehensive visual insights for flow field diagnosis.

[0126] The integrated diagnostic and output unit is also used for:

[0127] When the asymmetry index exceeds the first threshold, the core flow region containing the mainstream high-speed region and the corresponding low-speed return region are defined as the designated region.

[0128] When the intensity of the vortex structure exceeds the second threshold, the vortex influence area within a preset radius is defined as the designated area, with the vortex center as the center.

[0129] Based on the spatial distribution of the actual flow direction angle within the cross-section, the region where the actual flow direction angle deviates from the pipe axis by a greater than the third threshold is defined as the significant secondary flow region and designated as the specified region.

[0130] The non-straight pipe gas velocity measurement system of this application embodiment further includes: a real-time adaptive calibration module, which includes:

[0131] The model-flow field matching evaluation unit is used to continuously monitor the time-varying characteristics of the two-dimensional velocity distribution field output by the flow field reconstruction module, and dynamically evaluate the matching degree between the non-straight pipe turbulence model used in the current inversion calculation and the current actual flow field.

[0132] The online model parameter optimization unit is used to trigger the online calibration process when the matching degree is lower than a preset threshold. It performs inverse problem optimization on the key model parameters in the non-straight pipe turbulence model to obtain the optimized model parameters. The key model parameters include one or more of the following: boundary layer velocity profile exponent, turbulent eddy viscosity coefficient, and secondary flow intensity coefficient.

[0133] The model parameter dynamic update unit is used to feed back the optimized model parameters to the flow field reconstruction module in real time to update the embedded prior constraints.

[0134] In the specific implementation process, such as Figure 3 As shown, the real-time adaptive calibration module establishes an online closed loop of "monitoring-evaluation-optimization-updating," endowing the entire measurement system with the ability to continuously self-optimize and adapt to changes in operating conditions. This module can evaluate the matching degree between the current flow field characteristics and the preset turbulence model in real time. Once it detects changes in flow characteristics due to fluid property variations, significant deviations in operating conditions, or long-term pipeline operation, it automatically triggers online inverse problem optimization of key physical parameters in the model (such as boundary layer velocity profile exponents and secondary flow intensity coefficients), and feeds the optimized parameters back to the flow field reconstruction module in real time, thereby dynamically correcting the prior physical constraints in the algorithm. This mechanism effectively overcomes the limitations of fixed models in complex and variable industrial scenarios, ensuring that the physical priors upon which the inversion reconstruction relies always closely match the actual flow state. This maintains high accuracy and high reliability during long-term operation, achieving an intelligent upgrade from "preset model guiding measurement" to "measurement data continuously calibrating the model, and the model optimizing measurement in real time," significantly improving the system's adaptability and robustness throughout its entire lifecycle.

[0135] The non-straight pipe gas velocity measurement system of this application, through an innovative scheme combining external acoustic tomography and model-constrained inversion, achieves non-invasive, high-precision, full-field measurement of the complex three-dimensional turbulent flow field inside non-straight pipes. Firstly, the system achieves a breakthrough in measurement method by using a ring-shaped acoustic array outside the pipe for non-contact measurement, fundamentally eliminating probe interference with the flow field and providing a non-invasive observation window for flow field diagnosis. Secondly, by embedding physical models such as "secondary flow caused by pipe bending" as prior knowledge into the acoustic tomography inversion algorithm, it completely eliminates the traditional assumption of "straight pipe, uniform flow," accurately reconstructing the asymmetric velocity distribution and vortex structure in non-straight pipes, filling a gap in existing technology. Thirdly, highly robust signal processing and real-time adaptive calibration closed-loop ensure stable and reliable full-field velocity distribution and statistical characteristics even in noisy and variable industrial environments, providing unprecedentedly accurate data support for process optimization, fault early warning, and energy efficiency management, and possessing significant engineering application value.

[0136] In the description of this application, it should be understood that the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this application, "multiple" means two or more, unless otherwise explicitly specified.

[0137] In this application, unless otherwise expressly specified and limited, the terms "installation," "connection," "linking," and "fixing," etc., should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral part; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; they can refer to the internal communication of two components or the interaction between two components. Those skilled in the art can understand the specific meaning of the above terms in this application according to the specific circumstances.

[0138] In this application, unless otherwise expressly specified and limited, "above" or "below" the second feature can mean that the first and second features are in direct contact, or that they are in indirect contact through an intermediate medium. Furthermore, "above," "on top of," and "over" the second feature can mean that the first feature is directly above or diagonally above the second feature, or simply that the first feature is at a higher horizontal level than the second feature. "Below," "below," and "under" the second feature can mean that the first feature is directly below or diagonally below the second feature, or simply that the first feature is at a lower horizontal level than the second feature.

[0139] In the description of this specification, the terms "one embodiment," "some embodiments," "embodiment," "example," "specific example," or "some examples," etc., refer to specific features, structures, materials, or characteristics described in connection with that embodiment or example, which are included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0140] Although embodiments of this application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting this application. Those skilled in the art can make modifications, alterations, substitutions and variations to the above embodiments within the scope of this application.

Claims

1. A gas velocity measurement system for non-straight pipes, characterized in that, include: The acoustic detection module includes multiple acoustic sensors deployed outside the target pipe section. The multiple acoustic sensors are paired in pairs to form multiple sensor pairs, which are used to form multiple acoustic detection paths outside the pipe that cross the same cross section inside the pipe from different spatial angles. The signal processing module is used to receive the acoustic signal emitted by the acoustic detection module, perform amplitude analysis and transit time calculation on the acoustic signal, and obtain the gas flow velocity path integral measurement data for each path. The flow field reconstruction module is used to perform inversion calculations on the path integral measurement data using an improved acoustic tomography algorithm that integrates a non-straight pipe turbulence model, and reconstruct the two-dimensional velocity distribution field of the cross section to be measured. The flow field analysis module is used to perform intelligent data analysis on the two-dimensional velocity distribution field, extract and output key velocity information; the key velocity information includes: statistical characteristic values ​​of velocity covering a specified area within the cross-section to be measured and a visual cloud map of velocity distribution.

2. The non-straight pipe gas velocity measurement system according to claim 1, characterized in that, The multiple acoustic sensors are mounted on an adjustable bracket outside the target pipe section. The multiple acoustic sensors are arranged in a ring array, with the center of the ring array coinciding with the pipe axis, and the angle between adjacent sensors is less than or equal to 15°.

3. The non-straight pipe gas velocity measurement system according to claim 1, characterized in that, The signal processing module includes: The signal conditioning unit is used to filter and amplify the acoustic signal received by the acoustic sensor to obtain the preprocessed signal. The data acquisition unit is used to convert the preprocessed signal into a digital signal; The transit time calculation unit is used to analyze the digital signal using a cross-correlation algorithm, calculate the transit time of the sound wave propagating along each sensor in the downstream and upstream directions respectively, and obtain the transit time difference; The path integral calculation unit is used to obtain the gas velocity path integral measurement data on each acoustic detection path based on the transit time difference and the corresponding spatial geometric relationship, according to the principle of acoustic time difference method.

4. The non-straight pipe gas velocity measurement system according to claim 3, characterized in that, The transit time calculation unit is specifically used for: For each sensor pair, perform the following steps: Acquire both forward-flow and reverse-flow digital signals; Calculate the first cross-correlation function between the downstream digital signal and the standard reference signal, and the second cross-correlation function between the upstream digital signal and the standard reference signal, respectively. The main peak values ​​of the first cross-correlation function and the second cross-correlation function are detected respectively to determine the downstream transit time and the upstream transit time. The crossing time difference is obtained based on the downstream crossing time and the upstream crossing time.

5. The non-straight pipe gas velocity measurement system according to claim 1, characterized in that, The flow field reconstruction module includes: The flow field initialization unit is used to establish an initial two-dimensional velocity distribution field based on the geometric structure parameters and fluid property parameters of the target pipe section. The model fusion unit is used to embed the non-straight pipe turbulence model into the acoustic tomography inversion algorithm in the form of prior constraints. The non-straight pipe turbulence model is used to describe the flow field asymmetry and vortex characteristics caused by pipe bending or structural abrupt changes. The iterative inversion unit is used to take the path integral measurement data as input and use an acoustic tomography inversion algorithm with an embedded turbulence model to iteratively correct the initial two-dimensional velocity distribution field until the error between the theoretical path integral value and the actual measurement data corresponding to the reconstructed two-dimensional velocity distribution field is less than a set threshold, and outputs the final reconstructed two-dimensional velocity distribution field.

6. The non-straight pipe gas velocity measurement system according to claim 5, characterized in that, The iterative inversion unit is specifically used to perform inversion calculations through the following steps: The iterative process begins based on the initial two-dimensional velocity distribution hypothesis field and the path integral measurement data; in each iteration step, the following operations are performed: Based on the estimated value of the two-dimensional velocity distribution field in the current iteration, a forward calculation is performed to obtain the corresponding theoretical path integral value set; the theoretical path integral value set is compared with the path integral measurement data, and the data fitting residual is calculated; Based on the prior constraints of the non-straight pipe turbulence model, the correction amount of the current two-dimensional velocity distribution field estimate is obtained through optimization algorithm, and the estimate of the two-dimensional velocity distribution field is updated. The iteration process is repeated until the data fitting residual is less than a set threshold or the number of iterations reaches a preset upper limit, and the estimated two-dimensional velocity distribution obtained at this time is output as the final reconstructed two-dimensional velocity distribution field.

7. The non-straight pipe gas velocity measurement system according to claim 1, characterized in that, The flow field analysis module includes: The flow field feature extraction unit is used to identify and quantify specific flow field features caused by the non-straight section structure of the pipe from the two-dimensional velocity distribution field; the specific flow field features include: the asymmetry index of the velocity distribution, the center position and intensity of one or more vortex structures, and the actual flow direction angle synthesized by the main flow and secondary flow; The integrated diagnosis and output unit is used to obtain the flow statistics of a specified area within the cross section to be measured based on the two-dimensional velocity distribution field and the specific flow field characteristics; at the same time, it generates a velocity distribution visualization cloud map that integrates the annotations of the specific flow field characteristics.

8. The non-straight pipe gas velocity measurement system according to claim 7, characterized in that, The integrated diagnosis and output unit is used to obtain the flow statistics of a specified region within the cross-section to be measured based on the two-dimensional velocity distribution field and the specific flow field characteristics. Specifically, it includes: Based on the actual flow direction angle in the specific flow field characteristics, the two-dimensional planar velocity vector of each grid cell in the two-dimensional velocity distribution field is converted and synthesized into the axial velocity component of each grid cell along the pipe axis, thereby generating the axial velocity component distribution field of the cross section to be measured. Based on the specific flow field characteristics, the designated region is defined within the cross section to be measured. The designated region includes: the core flow region indicated by the asymmetry index, the vortex influence region centered on the center position of the vortex structure, and the secondary flow significant region defined by the actual flow direction angle. Based on the axial velocity component distribution field, the velocity statistical feature values ​​of each specified region are extracted and calculated. The velocity statistical feature values ​​include: the region average axial velocity, the region axial velocity standard deviation, and the region maximum reverse velocity.

9. The non-straight pipe gas velocity measurement system according to claim 8, characterized in that, The integrated diagnostic and output unit is also used for: When the asymmetry index exceeds the first threshold, the core flow region containing the mainstream high-speed region and the corresponding low-speed return region are defined as the designated region. When the intensity of the vortex structure exceeds the second threshold, the vortex influence area within a preset radius is defined as the designated area, with the vortex center as the center. Based on the spatial distribution of the actual flow direction angle within the cross-section, the region where the actual flow direction angle deviates from the pipe axis by an angle greater than the third threshold is defined as the secondary flow significant region as the designated region.

10. The non-straight pipe gas velocity measurement system according to claim 1, characterized in that, The system further includes: a real-time adaptive calibration module, the real-time adaptive calibration module comprising: The model-flow field matching degree evaluation unit is used to continuously monitor the time-varying characteristics of the two-dimensional velocity distribution field output by the flow field reconstruction module, and dynamically evaluate the matching degree between the non-straight pipe turbulence model used in the current inversion calculation and the current actual flow field. The online model parameter optimization unit is used to trigger an online calibration process when the matching degree is lower than a preset threshold, and to perform inverse problem optimization on the key model parameters in the non-straight pipe turbulence model to obtain the optimized model parameters; the key model parameters include one or more of the following: boundary layer velocity profile exponent, turbulent eddy viscosity coefficient, and secondary flow intensity coefficient. The model parameter dynamic update unit is used to feed back the optimized model parameters to the flow field reconstruction module in real time to update the embedded prior constraints.

Images