Multi-modal ultrasonic flow nonlinear error compensation method and system

Abstract

The invention relates to the technical field of multi-mode ultrasonic flow nonlinear error compensation scheme design, in particular to a multi-mode ultrasonic flow nonlinear error compensation method and system. The method comprises the following steps: acquiring propagation time data of different sound wave paths through a multi-ultrasonic transducer array; extracting a multi-modal characteristic parameter reflecting the nonlinear characteristic of the fluid; abnormal types such as bubble interference, solid impurities or turbulence are identified based on the characteristic parameters; calling a preset compensation model which is matched abnormally to generate a dynamic correction parameter; and fusing the parameter and the path weight coefficient to reconstruct an accurate flow value. The system is correspondingly provided with a sensing module, a feature extraction module, an interference identification module, a dynamic compensation module and a flow reconstruction module. The measuring bottleneck of a traditional flowmeter in complex fluid is broken through, and through the synergistic effect of interference type recognition and a dynamic compensation model, the measuring precision, the environmental adaptability and the long-term stability under the complex working condition are remarkably improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of multi-modal ultrasonic flow nonlinear error compensation scheme design, and particularly relates to a multi-modal ultrasonic flow nonlinear error compensation method and system. BACKGROUND

[0002] Ultrasonic flow measurement technology is widely used in the industrial field due to its non-contact and high-precision characteristics, but existing solutions have significant limitations when facing complex fluid media. The traditional single-path measurement model relies on the ideal fluid assumption. When bubble groups, solid particles are suspended or in a turbulent state in the medium, the sound wave propagation path will produce nonlinear distortion, resulting in a deviation of the propagation time measurement value from the true physical model. Existing compensation methods mostly use static environmental parameter correction, which cannot dynamically respond to fluid transient changes, especially in multiphase flow conditions, the error is aggravated. The more essential defect is that the current technical architecture lacks spatial perception ability of flow distribution - a single measurement path cannot capture the flow velocity gradient of the pipe cross section, and dynamic events such as bubble coalescence and rupture, random distribution of particles further destroy the uniformity of the flow velocity field, making the standard time difference method based on the central symmetric flow field derivation produce systematic deviation. Although in recent years there have been attempts to introduce a multi-sensor array, the data processing only simply superimposes the results of each path, without establishing a quantitative correlation mechanism between abnormal interference and measurement error, and without developing adaptive compensation strategies for different interference types. In summary, there is an urgent need for a systematic solution in the industrial field that can analyze fluid nonlinear characteristics in real time, intelligently identify interference sources, and dynamically reconstruct the flow calculation model.

[0003] Therefore, the prior art still needs further development. SUMMARY

[0004] The purpose of the present application is to overcome the above technical deficiencies and provide a multi-modal ultrasonic flow nonlinear error compensation method and system to solve the problems existing in the prior art.

[0005] To achieve the above technical purpose, according to the first aspect of the present application, the present application provides a multi-modal ultrasonic flow nonlinear error compensation method, the method comprising:

[0006] (a) acquiring original propagation time data of fluid on at least two different sound wave propagation paths through a sensor array composed of multiple ultrasonic transducers;

[0007] (b) extracting multi-modal feature parameters reflecting the nonlinear characteristics of the fluid based on the original propagation time data;

[0008] (c) identifying the type of abnormal interference in the fluid according to the multi-modal feature parameters;

[0009] (d) calling a preset compensation model matching the abnormal interference type to generate a dynamic correction parameter;

[0010] (e) inputting the dynamic correction parameter into a flow calculation function to output an accurate flow value after error compensation.

[0011] Specifically, the multi-modal feature parameters in step (b) include path-to-path propagation time standard deviation σ t and acoustic wave signal attenuation coefficient α:

[0012]

[0013] Wherein:

[0014] σ t : path-to-path propagation time standard deviation;

[0015] t i : propagation time of the i-th path;

[0016] average propagation time of all paths;

[0017] N: total number of acoustic wave propagation paths;

[0018]

[0019] Wherein:

[0020] α: acoustic wave signal attenuation coefficient;

[0021] A0: amplitude of the signal at the transmitting end;

[0022] A r : amplitude of the signal at the receiving end;

[0023] L: acoustic wave propagation distance.

[0024] Specifically, the interference type identification in step (c) is realized by a KNN classifier.

[0025] Specifically, the compensation model in step (d) includes:

[0026]

[0027] Wherein:

[0028] Q comp : compensated flow;

[0029] Q raw : original flow;

[0030] C b : bubble concentration;

[0031] d p: pipe inner diameter;

[0032] k1, k2: calibration coefficients.

[0033] Specifically, C b is calculated by the following formula:

[0034]

[0035] wherein:

[0036] S(f): signal power spectral density;

[0037] f0: ultrasonic base frequency;

[0038] f1-f3: bubble resonance frequency bands.

[0039] Specifically, step (e) adopts dynamic weighted fusion:

[0040]

[0041] wherein:

[0042] w i : the i-th path weight factor;

[0043] D: pipe inner diameter;

[0044] θ i : sound wave incidence angle;

[0045] t u,i ,t d,i : upstream and downstream propagation time.

[0046] Specifically, w i is determined by the following formula:

[0047]

[0048] wherein:

[0049] SNR i : the i-th path signal-to-noise ratio;

[0050] σ t,i : time measurement standard deviation.

[0051] Specifically, it further includes the step of periodically calibrating the compensation model.

[0052] Specifically, the calibration adopts a recursive least squares algorithm:

[0053]

[0054] wherein:

[0055] θ k: model parameter vector;

[0056] K k+1 : Kalman gain matrix;

[0057] y k+1 : flow observation bias.

[0058] According to a second aspect of the present application, a multi-modal ultrasonic flow nonlinear error compensation system is provided, comprising:

[0059] A multi-path ultrasonic sensing module is used to collect multi-path ultrasonic propagation time signals.

[0060] A fluid feature extraction module is used to calculate fluid nonlinear characteristic quantities.

[0061] An interference type identification module is used to output interference category labels.

[0062] A dynamic compensation engine is used to generate type-matched compensation parameters.

[0063] A flow reconstruction module is used to output corrected flow values.

[0064] Beneficial effects:

[0065] The present application creatively constructs a multi-modal sensing-interference classification-model self-adaptive technical closed loop, breaking through the static limitations of traditional measurement frameworks:

[0066] Firstly, through the spatial sampling mechanism of the multi-path ultrasonic sensing network, the discretization analysis of the global flow field of the pipeline is realized for the first time, solving the inherent defect of insufficient representativeness of single-point measurement. Based on the collaborative analysis of characteristic parameters such as ultrasonic propagation time group, signal attenuation spectrum, and flow velocity distribution uniformity, a quantitative evaluation system of fluid nonlinear degree is formed, providing multi-dimensional criteria for error tracing.

[0067] Secondly, the interference type identification module uses an intelligent classification model trained under a large number of working conditions to accurately distinguish typical abnormal scenarios such as bubble interference, solid impurities, and turbulent flow changes, overcoming the hysteresis and one-sidedness of empirical threshold criteria. This mechanism ensures the precise directionality of the compensation strategy and avoids the overfitting or undercompensation problems caused by traditional "one-size-fits-all" compensation.

[0068] Finally, the dynamic compensation engine generates targeted correction parameters in real time through the matching mechanism of interference type-compensation model. This parameter cooperates with the multi-path weight distribution algorithm to achieve the optimal balance between data effectiveness and reliability in the flow reconstruction stage. The entire technical solution significantly improves three core performances: the measurement accuracy remains stable under extreme working conditions, the system robustness has strong inhibition ability to industrial noise and medium disturbance, and the self-calibration mechanism maintains the consistency of the measurement reference during long-term use. BRIEF DESCRIPTION OF DRAWINGS

[0069] Figure 1 is a flowchart of a multi-modal ultrasonic flow nonlinear error compensation method provided in specific embodiments of the present application;

[0070] Figure 2 is a system composition diagram of a multi-modal ultrasonic flow nonlinear error compensation system provided in specific embodiments of the present application. DETAILED DESCRIPTION

[0071] In order for those skilled in the art to better understand the technical solutions of the present application, the technical solutions of the present application will be described clearly and completely below in combination with the drawings of the present application. Based on the embodiments in the present application, other similar embodiments obtained by those skilled in the art without making creative efforts should all belong to the scope of protection of the present application. In addition, the directional words mentioned in the following embodiments, such as "up", "down", "left", "right", etc. are only the directions of the drawings, therefore, the directional words used are used to illustrate but not to limit the present application.

[0072] The present application will be further described below in combination with the drawings and preferred embodiments.

[0073] Please refer to Figures 1-2 , the present application provides a multi-modal ultrasonic flow nonlinear error compensation method, comprising:

[0074] (a) acquiring original propagation time data of a fluid on at least two different sound wave propagation paths through a sensor array composed of multiple ultrasonic transducers.

[0075] It should be noted that the system is composed of a multi-path ultrasonic sensing module 101, a fluid feature extraction module 102, an interference type identification module 103, a dynamic compensation engine 104 and a flow reconstruction module 105. The sensor array adopts 4 groups of 1MHz ultrasonic transducers, which are installed in a DN50 pipeline (inner diameter dp=52.5mm), the sound path angle θi=45°, and the propagation distance L=52.5mm / sin45°≈74.2mm.

[0076] (b) based on the original propagation time data, extracting multi-modal feature parameters reflecting the nonlinear characteristics of the fluid.

[0077] Specifically, the multi-modal feature parameters in step (b) include the inter-path propagation time standard deviation σ t and the sound wave signal attenuation coefficient α:

[0078]

[0079] wherein:

[0080] σ t: Standard deviation of inter-path propagation time (μs) ;

[0081] t i : Propagation time of the i-th path (μs) ;

[0082] : Average propagation time of all paths (μs) ;

[0083] N: Total number of sound wave propagation paths;

[0084] In the preferred embodiment of the present application, the preferred number of paths is 4, and it can be understood that N≥4 can cover more than 80% of the flow rate distribution (according to ISO standards).

[0085]

[0086] Wherein:

[0087] α: Attenuation coefficient of sound wave signal (dB / m) ;

[0088] A0: Signal amplitude at the transmitting end (V) ;

[0089] A r : Signal amplitude at the receiving end (V) ;

[0090] L: Sound wave propagation distance (m) ;

[0091] In the preferred embodiment of the present application, the preferred signal amplitude at the transmitting end is 5V, and the preferred sound wave propagation distance is 74.2mm.

[0092] Further, the multi-modal characteristic parameter further comprises a flow rate uniformity index:

[0093]

[0094] Wherein:

[0095] Single-path flow rate (m / s) ; D = 52.5mm.

[0096] (c) Identifying the type of abnormal interference in the fluid according to the multi-modal characteristic parameter.

[0097] Specifically, the interference type identification of step (c) is realized by a KNN classifier.

[0098] It needs to be further explained that, as for the training of the interference type identification model, the present application designs the following scheme:

[0099] 1. Training data collection, as shown in Table 1:

[0100] Table 1 Training data collection

[0101]

[0102] Sample size: 500 groups for each type (total 1500 groups).

[0103] 2. KNN parameter configuration:

[0104] k = 7 (number of neighbors);

[0105] Distance function: Euclidean distance;

[0106] Weight function: inverse distance weighting;

[0107] It can be understood that the cross-validation accuracy is 98.7% when k = 7 (k = 5 is 97.2%, and k = 9 is 98.5%).

[0108] (d) calling a preset compensation model matched with the abnormal interference type to generate a dynamic correction parameter.

[0109] Specifically, the compensation model of step (d) includes:

[0110]

[0111] Wherein:

[0112] Q comp : compensated flow (m 3 / h);

[0113] Q raw : original flow (m 3 / h);

[0114] C b : bubble concentration (pieces / mL);

[0115] d p : pipe inner diameter (mm);

[0116] k1, k2: calibration coefficients.

[0117] Specifically, C b is calculated by the following formula:

[0118]

[0119] Wherein:

[0120] S(f): signal power spectral density (V 2 / Hz);

[0121] f0: ultrasonic base frequency (kHz);

[0122] f1-f3: bubble resonance frequency band (kHz).

[0123] It should be further noted that the following parameters have been optimized in this invention:

[0124] f0 = 1.0MHz (fundamental frequency);

[0125] f1 = 1.2MHz, f2 = 2.0MHz, f3 = 3.0MHz (bubble resonance band);

[0126] Understandably, the optimal parameters are based on the coverage of the 0.1-10 micrometer bubble resonance frequency band (according to the Minnaert equation).

[0127] It should be further noted that the following parameters have been optimized in this invention:

[0128] k1 = 0.35, k2 = 0.021 (calibration coefficients);

[0129] Experimentally determined value: at C b =0.5 cells / mL, error <0.5% (uncompensated error >8%).

[0130] (e) Input the dynamic correction parameters into the flow calculation function and output the accurate flow value after error compensation.

[0131] Specifically, step (e) employs dynamic weighted fusion:

[0132]

[0133] in:

[0134] w i : Weight factor for the i-th path;

[0135] D: Pipe inner diameter (mm);

[0136] θ i : Angle of incidence of sound wave (°);

[0137] t u,i ,t d,i : Upstream and downstream propagation time (μs).

[0138] Specifically, w i Determined by the following formula:

[0139]

[0140] in:

[0141] SNR i : Signal-to-noise ratio (dB) of the i-th path;

[0142] σ t,i Standard deviation of time measurement (μs).

[0143] It needs to be further explained that the core algorithm steps of the above design include:

[0144] ① Single path weight calculation:

[0145]

[0146] Among them:

[0147] (δ n =0.1V background noise);

[0148] σ t,i : Time measurement standard deviation (10 consecutive measurements).

[0149] ② Flow fusion output:

[0150]

[0151] Specifically, it also includes the steps of periodically calibrating the compensation model.

[0152] Specifically, the calibration uses the recursive least squares algorithm:

[0153]

[0154] Among them:

[0155] θ k : Model parameter vector;

[0156] K k+1 : Kalman gain matrix;

[0157] y k+1 : Flow observation deviation (%).

[0158] It needs to be further explained that regarding the recursive least squares (RLS) implementation, the scheme designed by the application includes:

[0159] 1. Calibration trigger condition:

[0160] Flow change <1% / min (for 5 minutes);

[0161] Zero flow reference time >30 seconds.

[0162] It can be understood that the calibration duration is set to 5 minutes, which is based on the ASTM stability criterion.

[0163] 2. Parameter update algorithm:

[0164]

[0165] Among them:

[0166] θ = [k1, k2] T (model parameter vector);

[0167] φ = [C b , d p ] T (characteristic vector);

[0168] y: measured flow deviation (%);

[0169] λ = 0.95 (forgetting factor);

[0170] It can be understood that the parameter is preferably based on the weight ratio of new and old data of 1:19 when λ = 0.95.

[0171] It should be further explained that the present application has carried out technical effect verification, and the results are shown in Table 2:

[0172] Table 2 Technical effect verification

[0173] Interference scenario Error before compensation Error after compensation Bubble concentration 0.3 per mL 6.8％ 0.4％ Solid impurities 0.5 mm 4.2％ 0.7％ Turbulence intensity 0.25 9.1％ 1.2％

[0174] It should be further explained that the KNN classifier is realized by using Python scikit-learn, and the RLS algorithm is calculated by the floating point unit of the STM32F407 chip.

[0175] It can be understood that the present application creatively constructs a multi-modal perception-interference classification-model adaptive technical closed loop, breaking through the static limitations of the traditional measurement framework:

[0176] Firstly, through the spatial sampling mechanism of the multi-path ultrasonic sensing network, the discretization analysis of the global flow field of the pipeline is realized for the first time, solving the inherent defect of insufficient representativeness of single-point measurement. Based on the cooperative analysis of characteristic parameters such as sound wave propagation time group, signal attenuation spectrum and flow velocity distribution uniformity, a quantitative evaluation system of fluid nonlinearity is formed, providing multi-dimensional criteria for error tracing.

[0177] Secondly, the interference type recognition module adopts an intelligent classification model trained by a large number of working conditions, which can accurately distinguish typical abnormal scenes such as bubble interference, solid impurities and turbulent excitation, and overcome the hysteresis and one-sidedness of the empirical threshold criterion. This mechanism ensures the precise directivity of the compensation strategy and avoids the overfitting or undercompensation problems caused by the traditional "one-size-fits-all" compensation.

[0178] Finally, the dynamic compensation engine generates a targeted correction parameter in real time through the matching mechanism of the interference type-compensation model. This parameter works in conjunction with the multipath weight distribution algorithm to achieve the optimal balance between data effectiveness and reliability during the flow reconstruction phase. The complete technical solution significantly improves three core performance: the measurement accuracy remains stable under extreme conditions, the system robustness has strong suppression ability for industrial noise and medium disturbance, and the self-calibration mechanism maintains the consistency of the measurement reference during long-term use.

[0179] Referring to Figure 2 The present application provides another embodiment, which provides a multi-modal ultrasonic flow nonlinear error compensation system, comprising:

[0180] A multi-path ultrasonic sensing module 101 is configured to collect multi-path ultrasonic propagation time signals.

[0181] A fluid feature extraction module 102 is configured to calculate fluid nonlinear feature quantities.

[0182] An interference type identification module 103 is configured to output interference category labels.

[0183] A dynamic compensation engine 104 is configured to generate type-matched compensation parameters.

[0184] A flow reconstruction module 105 is configured to output corrected flow values.

[0185] It should be noted that the present application creatively constructs a multi-modal perception-interference classification-model self-adaptive technical closed loop, breaking through the static limitations of traditional measurement frameworks:

[0186] First, through the spatial sampling mechanism of the multi-path ultrasonic sensing network, the discretization and analysis of the global flow field in the pipeline are realized for the first time, solving the inherent defect of insufficient representative of single-point measurement. Based on the collaborative analysis of characteristic parameters such as sound wave propagation time group, signal attenuation spectrum, and flow velocity distribution uniformity, a quantitative evaluation system for the degree of fluid nonlinearity is formed, providing multi-dimensional criteria for error tracing.

[0187] Second, the interference type identification module uses an intelligent classification model trained under a large number of working conditions to accurately distinguish typical abnormal scenarios such as bubble interference, solid impurities, and turbulent flow changes, overcoming the hysteresis and one-sidedness of empirical threshold criteria. This mechanism ensures the precise directionality of the compensation strategy and avoids the overfitting or undercompensation problems caused by traditional "one-size-fits-all" compensation.

[0188] Finally, the dynamic compensation engine generates a targeted correction parameter in real time through the matching mechanism of the interference type-compensation model. The parameter cooperates with the multipath weight distribution algorithm to achieve the optimal balance between data effectiveness and reliability in the flow reconstruction stage. The complete technical solution significantly improves three core performances: the measurement accuracy remains stable output under extreme working conditions, the system robustness has strong inhibition ability to industrial noise and medium disturbance, and the consistency of the measurement reference is maintained through the self-calibration mechanism in long-term use.

[0189] In a preferred embodiment, the present application also provides an electronic device, comprising:

[0190] a memory, and a processor, wherein computer readable instructions are stored on the memory, and the computer readable instructions, when executed by the processor, implement the multi-modal ultrasonic flow nonlinear error compensation method. The computer device can be a server, a terminal, or any other electronic device with necessary computing and / or processing capabilities. In one embodiment, the computer device can include a processor, a memory, a network interface, a communication interface, and the like connected by a system bus. The processor of the computer device can be used to provide necessary computing, processing, and / or control capabilities. The memory of the computer device can include a non-volatile storage medium and an internal memory. The non-volatile storage medium or the non-volatile storage medium can store an operating system, a computer program, and the like. The internal memory can provide an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The network interface and the communication interface of the computer device can be used to connect and communicate with external devices through a network. The computer program is executed by the processor to execute the steps of the method of the present application.

[0191] The present application can be implemented as a computer readable storage medium having a computer program stored thereon, which, when executed by a processor, causes the steps of the method of the embodiments of the present application to be performed. In one embodiment, the computer program is distributed on multiple computer devices or processors coupled by a network, so that the computer program is stored, accessed and executed by one or more computer devices or processors in a distributed manner. A single method step / operation, or two or more method steps / operations, can be performed by a single computer device or processor, or by two or more computer devices or processors. One or more method steps / operations can be performed by one or more computer devices or processors, and one or more other method steps / operations can be performed by one or more other computer devices or processors. One or more computer devices or processors can perform a single method step / operation, or perform two or more method steps / operations.

[0192] As will be appreciated by one of ordinary skill in the art, the steps of the methods of the present application can be directed to relevant hardware, such as computer devices or processors, by way of computer program instructions stored in a non-transitory computer-readable storage medium that, when executed, cause the steps of the present application to be performed. Any reference to memory, storage, a database, or another medium herein can include non-volatile and / or volatile storage. Examples of non-volatile storage include read-only memory (ROM), programmable ROM (PROM), electronically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), flash memory, magnetic tape, floppy disks, magnetic

[0193] It can be understood that the present application creatively constructs a multi-modal perception-interference classification-model adaptive technical closed loop, breaking through the static limitations of the traditional measurement framework:

[0194] Firstly, through the spatial sampling mechanism of the multi-path ultrasonic sensing network, the discretization analysis of the global flow field of the pipeline is realized for the first time, solving the inherent defect of insufficient representative of single-point measurement. Based on the collaborative analysis of characteristic parameters such as sound wave propagation time group, signal attenuation spectrum, and flow velocity distribution uniformity, a quantitative evaluation system for the nonlinearity degree of fluid is formed, providing multi-dimensional criteria for error tracing.

[0195] Secondly, the interference type recognition module adopts an intelligent classification model trained by a large number of working conditions, which accurately distinguishes typical abnormal scenarios such as bubble interference, solid impurities, and turbulent flow changes, overcoming the hysteresis and one-sidedness of empirical threshold criteria. This mechanism ensures the precise directionality of the compensation strategy and avoids the overfitting or undercompensation problems caused by traditional "one-size-fits-all" compensation.

[0196] Finally, the dynamic compensation engine generates targeted correction parameters in real time through the matching mechanism of the interference type-compensation model. The parameters work together with the multi-path weight allocation algorithm to achieve the optimal balance between data effectiveness and reliability in the flow reconstruction stage. The entire technical solution significantly improves three core performances: the measurement accuracy remains stable under extreme working conditions, the system robustness has strong inhibition ability to industrial noise and medium disturbance, and the self-calibration mechanism maintains the consistency of the measurement reference during long-term use.

[0197] The technical features described above can be combined in any way. Although not all possible combinations of the technical features are described, any combination of the technical features should be considered as covered by the present specification, as long as such a combination does not result in a contradiction.

[0198] The above description of the specific embodiments of the present application is not intended to limit the scope of the present application. Any other corresponding changes and modifications made according to the technical concept of the present application should be included in the scope of protection of the claims of the present application.

Claims

1. A multi-modal ultrasonic flow nonlinear error compensation method, characterized in that, The method comprises: (a) acquiring original propagation time data of the fluid on at least two different sound wave propagation paths through a sensor array composed of multiple ultrasonic transducers; (b) extracting multi-modal feature parameters reflecting the nonlinear characteristics of the fluid based on the original propagation time data; (c) identifying the type of abnormal interference in the fluid according to the multi-modal feature parameters; (d) calling a preset compensation model matched with the type of abnormal interference to generate a dynamic correction parameter; (e) inputting the dynamic correction parameter into a flow calculation function to output an accurate flow value after error compensation.

2. The multi-modal ultrasonic flow non-linearity error compensation method of claim 1, wherein, The multi-modal feature parameters in step (b) include inter-path travel time standard deviation σ t and acoustic wave signal attenuation coefficient α: Wherein: σ t : inter-path propagation time standard deviation; t i : propagation time of the i-th path; t: average propagation time of all paths; N: total number of sound wave propagation paths; Wherein: α: sound wave signal attenuation coefficient; A0: signal amplitude at the transmitting end; A r : received end signal amplitude; L: sound wave propagation distance.

3. The multi-modal ultrasonic flow non-linearity error compensation method of claim 1, wherein, The interference type identification of step (c) is realized through a KNN classifier.

4. The multi-modal ultrasonic flow non-linearity error compensation method of claim 1, wherein, The compensation model of step (d) comprises: Wherein: Q comp : compensated flow rate; Q raw : raw flow; C b : bubble concentration; d p : pipe inside diameter; k1, k2: calibration coefficients.

5. The multi-modal ultrasonic flow non-linearity error compensation method of claim 4, wherein, C b By the following formula: Wherein: S(f): signal power spectral density; f0: ultrasonic base frequency; f1-f3: bubble resonance frequency band.

6. The multi-modal ultrasonic flow non-linearity error compensation method of claim 1, wherein, Step (e) adopts dynamic weighted fusion: Wherein: w i : ith path weight factor; D: inner diameter of the pipeline; θ i : angle of incidence of the sound wave; t u,i ,t d,i : upstream and downstream propagation times.

7. The multi-modal ultrasonic flow non-linearity error compensation method of claim 6, wherein, w i was determined by the formula: Wherein: SNR i : ith path signal-to-noise ratio; σ t,i : Time measurement standard deviation.

8. The multi-modal ultrasonic flow non-linearity error compensation method of claim 1, wherein, It also includes the step of periodically calibrating the compensation model.

9. The multi-modal ultrasonic flow non-linearity error compensation method of claim 8, wherein, The calibration adopts a recursive least squares algorithm: Wherein: θ k : model parameter vector; K k+1 : Kalman gain matrix; y k+1 : flow observation bias.

10. A multi-modal ultrasonic flow nonlinear error compensation system, characterized by, It includes: A multi-path ultrasonic sensing module (101) for collecting multi-sound path propagation time signals; A fluid feature extraction module (102) for calculating fluid nonlinear characteristic quantities; An interference type identification module (103) for outputting interference category labels; A dynamic compensation engine (104) for generating type-matched compensation parameters; A flow reconstruction module (105) for outputting corrected flow values.

Images