Online evaluation method for performance of ultrasonic gas flowmeter

通过贝叶斯信息融合方法,结合先验信息和实时数据,实时更新不确定性评估,解决了超声波流量计在动态条件下的性能评估难题,提高了测量精度和可靠性,适应流体流动和环境 Changes have improved industrial production efficiency.

CN118913411APending Publication Date: 2024-11-08CHINA JILIANG UNIV
View PDF 0 Cites 2 Cited by

Patent Information

Application Number
CN202410950813.0
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-16
Publication Date
2024-11-08

AI Technical Summary

Technical Problem

It is difficult to evaluate performance in real time under dynamic conditions. The measurement uncertainty is mainly applicable to static conditions and cannot effectively deal with real-time performance fluctuations caused by changes in external conditions and internal structures.

Method used

By using Bayesian information fusion method, the uncertainty evaluation is updated in real time through the combination of prior information and real-time data, the posterior mean and standard deviation are calculated, the impact of errors of a single data source is reduced, and the real-time evaluation and accuracy of flowmeter performance is achieved. calibration.

Benefits of technology

It improves the measurement accuracy and reliability of ultrasonic flowmeters, adapts to the dynamic changes of fluid flow and environmental conditions, enhances the real-time and accuracy of the flowmeter, and improves the quality and efficiency of industrial production.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118913411A_ABST
    Figure CN118913411A_ABST
Patent Text Reader

Abstract

The invention relates to an online performance evaluation method for an ultrasonic gas flowmeter, and belongs to the field of ultrasonic flow measurement. Comprising the following steps: S1, calculating a priori mean value mu0 and a priori standard deviation sigma0; s2, calculating a posterior mean value mu 1 and a posterior standard deviation sigma 1; s3, the uncertainty u1 is calculated, the steps are repeated, and the posterior distribution serves as posterior distribution in sequence and is fused with next data until all real-time data are fused; and S4, obtaining the mean values [mu] up1, [mu] up2,... [mu] upn, [mu] down 1, [mu] down 2,... [mu] down n and the standard deviations [sigma] up1, [sigma] up2,... [sigma] upn, [sigma] down 1, [sigma] down 2,... [sigma] down n of the clockwise and countercurrent time, and calculating the standard uncertainty of the clockwise and countercurrent time as [mu] up1, [mu] up2,... [mu] upn, [mu] down 1, [mu] down 2,... [sigma] down n. According to the method, real-time collection of online measurement data is achieved by establishing the change point detection model and obtaining the change trend of the measurement data, and the specific implementation process is that prior information and real-time sample information are combined to obtain posterior distribution so as to more accurately evaluate the performance of the flowmeter. By means of the method, real-time performance evaluation and precision calibration can be conducted on the ultrasonic flowmeter, and then the quality and efficiency of industrial production are improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to an online performance evaluation method of an ultrasonic gas flow meter, belonging to the field of ultrasonic flow meters. Background Art

[0002] The time difference method ultrasonic flowmeter usually measures the time from the emission to the reception of the ultrasonic signal, which is called the transit time. The propagation time of the ultrasonic signal from the upstream sensor to the downstream sensor will decrease due to the flow of the fluid, while the propagation time in the reverse direction will increase. By calculating the time difference between the upstream and reverse flow, the flow velocity of the fluid can be obtained, and then the flow rate can be calculated. How to process the data of the upstream and reverse flow time difference is particularly important.

[0003] The current commonly used measurement uncertainty assessment is mainly applicable to static measurement uncertainty, but in reality, the ultrasonic flowmeter is affected by external conditions and internal structure, and its performance changes in real time with the measurement time. Therefore, it is of great significance to propose a method for dynamic ultrasonic flowmeter data processing. Summary of the invention

[0004] The purpose of the present invention is to provide a method for online monitoring of ultrasonic flowmeter transit time by using a Bayesian information fusion method. The specific technical solution is as follows:

[0005] S1, let the population X~N(μ,σ 2 ), sample x=(x1,x2,…x n ). Calculate the prior mean μ0 and the prior standard deviation σ0. First, the first 100 upstream and downstream time data collected are taken as prior information. The prior information collected here belongs to historical information.

[0006]

[0007]

[0008] Where M is the number of trials, y r is the rth data point, is the population mean. The prior mean μ0 and the prior standard deviation σ0 are obtained by formulas (1) and (2).

[0009] S2. Calculate the posterior mean μ1 and the posterior standard deviation σ1. Continue to collect the upstream and downstream time, take the 101th data as sample information, and merge it with the first 100 prior information data to obtain the posterior distribution N~(μ1,σ1 2 ), where σ 2 Unknown, the mean μ is unknown:

[0010] The parameters to be estimated are (μ,σ 2 ), where the estimated parameter σ2 =α, given a sample, the joint density of α is:

[0011]

[0012] in,

[0013] The marginal distribution of parameter β is set to the normal distribution N~(μ0,σ0 2 ), the marginal distribution of α is set to the normal distribution N~(μ1,σ1 2 ), if β and α are unrelated, a priori joint distribution is performed on (β, α), and a binormal distribution is selected. At this time, p(β, α)~N(μ0,μ1,σ1 2 ,σ2 2 ),but:

[0014]

[0015] According to p(θ|x)∝l(x|θ)p(θ), the posterior distribution kernel function of the estimated parameter (β, α) is:

[0016]

[0017] Process the above formula β and α, take their derivatives and equal them to zero, and then solve:

[0018]

[0019]

[0020] in,

[0021] Then the maximum a posteriori estimate of α is:

[0022]

[0023] Among them, σ1 2 is the prior variance of variance α, Determined by the current data, n is the number of current measurement data.

[0024] From formulas (6) and (7), we can obtain the posterior mean μ1 and the posterior standard deviation σ1.

[0025] S3, and calculate the uncertainty u1. If the random variable X is set to obey the normal distribution N~(μ,σ 2 ), x=(x1,x2,…x n ) is the sample size, and the Bayesian uncertainty can be expressed by calculating the variance of the posterior distribution and finally by the square root of the maximum a posteriori estimate, where the variance σ 2Unknown, the mean μ is unknown:

[0026]

[0027] in,

[0028] The uncertainty u1 can be obtained from formula (9). This cycle is repeated, and the posterior distribution is used as the posterior distribution in turn and fused with the next data until all real-time data are fused.

[0029] S4, the mean of the forward and reverse flow time is μ up1 ,μ up2 ,...μ upn , μ down1 ,μ down2 ,...μ downn and standard deviation σ up1 ,σ up2 ,...σ upn , σ down1 ,σ down2 ,...σ downn , and calculate the standard uncertainty of the forward and reverse flow time as u up1 ,u up2 ...u upn 、u down1 ,u down2 ...u downn .

[0030] The beneficial effects of the present invention are as follows: through the Bayesian information fusion method, the uncertainty assessment can be updated in real time when new data arrives, adapting to the dynamic changes of fluid flow and environmental conditions, fusing multi-source information, reducing the impact of single data source errors on the calculation of upstream and downstream time differences, and improving the real-time and accuracy of the measurement.

[0031] The solution establishes a change point detection model to obtain the change trend of the measurement data and realize the real-time collection of online measurement data. The specific implementation process is to combine the prior information with the real-time sample information to obtain the posterior distribution to more accurately evaluate the performance of the flow meter. In this way, the flow meter is evaluated and calibrated in real time using the prior information and real-time data. This provides a new method and approach for the online monitoring of ultrasonic flow meters, which helps to improve the measurement accuracy and reliability of flow meters, thereby improving the quality and efficiency of industrial production. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] When considered in conjunction with the accompanying drawings, the advantages of the present invention can be better understood by referring to the following detailed description, but the drawings described herein are used to provide a further understanding of the present invention and constitute a part of the present invention. The exemplary embodiments and descriptions of the present invention are used to explain the present invention and do not constitute an improper limitation of the present invention, wherein:

[0033] Figure 1 It is the flow chart of the algorithm of the present invention.

[0034] Figure 2 This is a priori information data collection diagram of the present invention.

[0035] Figure 3 It is the posterior mean map of the present invention.

[0036] Figure 4 This invention 3 / hStandard uncertainty.

[0037] Figure 5 The standard uncertainty diagram of the present invention at different flow points.

[0038] Figure 6 The data collection interface diagram of the present invention is shown. DETAILED DESCRIPTION

[0039] The present invention is further described below with reference to the accompanying drawings using an embodiment, but this embodiment is merely illustrative and the protection scope of the present invention is not limited by this embodiment.

[0040] In this embodiment, a gas ultrasonic flow meter prototype is selected and placed on a sonic nozzle device table for experimental testing. Figure 1 Take 0m 3 The 100 real-time upstream and downstream time data collected at 1 / h are used as prior information, and the results are as follows: Figure 2 shown.

[0041] The prior mean μ0 and the prior standard deviation σ0 of the first 100 data are calculated according to formulas (1) and (2):

[0042]

[0043]

[0044]

[0045]

[0046] The prior distribution of the backflow time is N~(153.5569,0.001739 2 ), the prior distribution of downstream time is N~(153.5549,0.0057842 ).

[0047] Then take the 101th data as the current measurement sample information and integrate it into the prior distribution to obtain the posterior mean μ1 and posterior standard deviation σ1:

[0048]

[0049]

[0050]

[0051]

[0052] According to formula (9), the standard uncertainty of upstream time and downstream time is:

[0053] u up1 =0.0001749 (18)

[0054] u down1 =0.0001631 (19)

[0055] The first set of measurement results is used as the prior distribution N~(153.5589,0.001749 2 ), the prior distribution of the countercurrent time is N~(153.5541,0.001631 2 ), according to formulas (6) and (7), the posterior mean of the reverse flow time can be obtained as μ up2 =154.1968, the posterior mean of the downstream time is μ down2 =153.4846, the posterior standard deviation of the counterflow time is σ up2 =0.0008994, the standard deviation of the posterior mean of the downstream time is σ down2 =0.001009, then the estimated value of the second set of upstream time measurement results is 156.1968μs, with a standard uncertainty of 0.0008994, and the estimated value of the downstream time measurement results is 153.4846μs, with a standard uncertainty of 0.001009, and so on. The posterior mean is as follows Figure 3 As shown, 0m 3 / hStandard uncertainty is as follows Figure 4 shown.

[0056] The mean of the upstream time is 153.5575 μs, and the mean of the downstream time is 153.5568 μs. The standard deviation of the mean of the downstream time and the upstream time is:

[0057]

[0058]

[0059] Therefore, the uncertainty component caused by the upstream time is 0.00001344μs, and the uncertainty component caused by the downstream time is 0.00001056μs.

[0060] Figure 5 The standard uncertainty diagram at different flow points is shown. In order to further observe the possible problems of the ultrasonic flow meter during the test, the ultrasonic gas flow meter at different flow points is monitored online based on the Bayesian information fusion method. Each flow point is set to 10m 3 / h、20m 3 / h、30m 3 / h、40m 3 / h、50m 3 / h、60m 3 / h、70m 3 / h, it can be clearly seen in the figure that the standard uncertainty of the forward and reverse flow time has a "spike", indicating that there are abnormal and sudden situations. The reason is that the change of the nozzle combination, the closing and opening of the control valve and the gas flow characteristics inside the device affect the accuracy of the forward and reverse flow time, causing its fluctuation, and the standard uncertainty of the forward and reverse flow time becomes larger. The Bayesian information fusion method is used to monitor this abnormal situation, which can respond to the problem in time and solve the abnormal situation, thereby improving the accuracy and stability of the ultrasonic gas flow meter.

[0061] Figure 6 This is the data acquisition interface diagram, which collects the values ​​of upstream and downstream transit time in real time and performs online performance evaluation on the upstream and downstream transit time. It is mainly divided into four modules. The upper left part is the upstream and downstream transit time value collected online, the upper right part is the calculation of the a posteriori mean of the upstream and downstream transit time collected online, the lower left part is the standard uncertainty evaluation of the online upstream and downstream transit time, and the lower right part is the flow rate at each flow point. The overall software interface realizes the online collection of upstream and downstream transit time, and calculates the a priori mean and standard uncertainty in real time online to realize the online performance evaluation of the ultrasonic gas flowmeter.

[0062] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solution of the present invention can be modified or replaced by equivalents without departing from the purpose and scope of the technical solution, which should be included in the scope of the claims of the present invention.

Claims

1. An online performance evaluation method for ultrasonic gas flowmeters, which monitors the transit time of ultrasonic flowmeters online. When the overall sample obeys a normal distribution, the measurement uncertainty is evaluated based on Bayesian information fusion. It is characterized by Follow these steps: S1. Determine the prior probability distribution; S2. Collect sample data and update the prior distribution to the posterior distribution; S3 and S4. Calculate the Bayesian estimate and uncertainty.

2. The method for online performance evaluation of an ultrasonic gas flowmeter according to claim 1, characterized in that: In step S1, the priori mean μ0 and the priori standard deviation σ0 are calculated. The first 100 upstream and downstream time data collected are taken as priori information. The priori information collected here belongs to historical information.

3. The method for online performance evaluation of an ultrasonic gas flowmeter according to claim 1, characterized in that: In step S2, the posterior mean μ1 and the posterior standard deviation σ1 are calculated. The upstream and downstream times are continuously collected, and the 101th data is used as sample information and fused with the first 100 prior information data to obtain the posterior distribution N~(μ1,σ1 2 ).

4. The method for online performance evaluation of an ultrasonic gas flowmeter according to claim 1, characterized in that: In step S3, the uncertainty u1 is calculated. This cycle is repeated, and the posterior distribution is used as the posterior distribution in turn and fused with the next data until all real-time data are fused.

5. The method for online performance evaluation of an ultrasonic gas flowmeter according to claim 1, characterized in that: In step S4, the mean of the forward and reverse flow time is μ up1 ,μ up2 ,...μ upn , μ down1 ,μ down2 ,...μ downn and standard deviation σ up1 ,σ up2 ,…σ upn , σ down1 ,σ down2 ,…σ downn , and calculate the standard uncertainty of the forward and reverse flow time as u up1 ,u up2 …u upn 、u down1 ,u down2 ...u downn .

Citation Information

Cited By

  • Ultrasonic metering instrument online self-checking method based on waveform compensation

    CN122130141A

  • An online self-checking method for ultrasonic metering instrument based on waveform compensation

    CN122130141B