Ultrasonic water meter metering method based on EMD (Empirical Mode Decomposition) de-noising

Abstract

The invention belongs to the field of flow metering, and particularly discloses an ultrasonic water meter metering method based on EMD decomposition denoising, which comprises the following steps of: S1, giving a value interval of temperature in an ultrasonic water meter and a value interval of intensity of external electromagnetic interference; s2, when water flow in the ultrasonic water meter moves, extracting signal features and water flow velocity, and constructing a signal sample; s3, for the combination of the value interval of each temperature and the value interval of the electromagnetic interference intensity, adding all obtained signal samples into the same set to form a sample set; s4, constructing a water flow velocity prediction model based on the deep neural network, and training the water flow velocity prediction model by using the sample set; s5, in the actual ultrasonic water meter metering process, ultrasonic water meter metering is completed. According to the method, the mapping relation among the signal characteristics, the temperature, the external electromagnetic environment and the water flow velocity after EMD denoising is established, and accurate ultrasonic water meter metering is achieved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of flow measurement, in particular to an ultrasonic water meter measurement method based on EMD decomposition denoising. BACKGROUND

[0002] The time difference method ultrasonic water meter is an important measurement method, which detects the water flow velocity by detecting the time difference generated by the change of velocity when the ultrasonic beam propagates in the water, so as to realize flow calculation; when detecting the time difference of the forward flow and the reverse flow, it needs to rely on the received ultrasonic signal, but the ultrasonic signal may be disturbed by various noises in the transmission process, which may come from the environment, the equipment itself or other external factors, which has a great influence on the accuracy of the time difference method; in addition, the propagation speed of ultrasonic wave in medium has a nonlinear relationship with temperature, and the traditional detection method is to compensate the temperature by looking up the table and combining interpolation, but interpolation method is essentially a linear method, and its precision has certain defects. SUMMARY

[0003] The purpose of the present application is to overcome the shortcomings of the prior art, provide an ultrasonic water meter measurement method based on EMD decomposition denoising, effectively filter out the noise in the ultrasonic signal through EMD decomposition, and consider the influence of temperature and external electromagnetic intensity, establish the relationship between the signal characteristics after EMD denoising, temperature, external electromagnetic environment and water flow velocity, and realize accurate ultrasonic water meter measurement.

[0004] The purpose of the present application is achieved by the following technical scheme: an ultrasonic water meter measurement method based on EMD decomposition denoising, comprising the following steps:

[0005] Step S1. Given the temperature value interval and the intensity value interval of external electromagnetic interference in the ultrasonic water meter;

[0006] Step S2. For any combination of temperature value interval and electromagnetic interference intensity value interval, when the water flow moves in the ultrasonic water meter, the ultrasonic signal is transmitted between the first transducer and the second transducer, and the signal characteristics and water flow velocity are extracted to construct the signal sample;

[0007] Step S3. For each combination of temperature value interval and electromagnetic interference intensity value interval, repeat step S2, and add all the signal samples obtained to the same set to form a sample set;

[0008] Step S4. Construct a water flow velocity prediction model based on deep neural network, and train the water flow velocity prediction model using the sample set;

[0009] Step S5. In the actual ultrasonic water meter measurement process, the ultrasonic signals are transmitted between the first transducer and the second transducer, the signal characteristics are extracted, and the current temperature value range and the electromagnetic interference intensity value range are tested to form a training sample, which is input into the trained water flow speed prediction model to obtain the water flow speed, and then the water flow speed is multiplied by the cross-sectional area of the fluid in the ultrasonic water meter to complete the ultrasonic water meter measurement.

[0010] The beneficial effects of the present application are: (1) The present application calculates the cross-correlation coefficient of each component obtained by EMD decomposition and the transmitted ultrasonic signal, selects the IMF component with a cross-correlation coefficient greater than a threshold value and parameters for signal reconstruction, effectively filters out the noise in the ultrasonic signal, and retains the important components in the signal, which helps to improve the accuracy of ultrasonic measurement;

[0011] (2) The present application performs correlation peak detection on the ultrasonic signal after EMD decomposition and denoising and the ultrasonic transmission signal, can accurately obtain the correlation peak to obtain the corresponding propagation time, improve the accuracy of the downstream and upstream time acquisition, and thus help to improve the measurement accuracy;

[0012] (3) The present application establishes the relationship between the signal characteristics after EMD denoising, temperature, external electromagnetic environment and water flow speed through the training of deep neural network. Since the signal characteristics contain denoising information associated with propagation time and flow rate, the direct operation of water flow speed is ingeniously converted into prediction based on deep neural network, and the influence information of temperature and external electromagnetic environment can be added in the prediction process. Therefore, it is not necessary to compensate by table lookup and interpolation, but the water flow speed prediction under different temperatures and external electromagnetic environments is completed in the model, and the measurement result is calculated through the water flow speed, which effectively improves the measurement accuracy. BRIEF DESCRIPTION OF DRAWINGS

[0013] Figure 1 The method flowchart of the present application. DETAILED DESCRIPTION

[0014] The technical solutions of the present application will be described in further detail below in combination with the drawings, but the protection scope of the present application is not limited to the following description.

[0015] As shown in the drawings, an ultrasonic water meter measurement method based on EMD decomposition and denoising includes the following steps: Figure 1

[0016] Step S1. Given the temperature value range and the intensity value range of external electromagnetic interference in the ultrasonic water meter.

[0017] ​Step S2. For any combination of temperature range and electromagnetic interference intensity range, when water flows in the ultrasonic water meter, ultrasonic signals are transmitted between the first transducer and the second transducer, and signal features and water flow velocity are extracted to construct signal samples.

[0018] In embodiments of this application, step S2 includes:

[0019] S201. For any combination of temperature range and electromagnetic interference intensity range, when water flows in the ultrasonic water meter, an ultrasonic signal is transmitted from the first transducer to the second transducer located downstream of the first transducer. The received ultrasonic signal is decomposed and denoised using EMD, and the first signal feature is extracted.

[0020] Then, an ultrasonic signal is transmitted from the second transducer to the first transducer. The received ultrasonic signal is decomposed and denoised using EMD, and the second signal features are extracted.

[0021] S202. For the denoised ultrasonic signal corresponding to the ultrasonic signal received by the second transducer, perform correlation peak detection with the ultrasonic signal emitted by the first transducer to obtain the position of the maximum correlation peak, and take the corresponding time as the downstream propagation time t1.

[0022] S203. For the denoised ultrasonic signal corresponding to the ultrasonic signal received by the first transducer, perform correlation peak detection with the ultrasonic signal emitted by the second transducer to obtain the position of the maximum correlation peak, and take the corresponding time as the reverse propagation time t2.

[0023] S204. Calculate the detected water flow velocity under the combined range of temperature and electromagnetic interference intensity:

[0024] ;

[0025] Where L represents the distance between the first transducer and the second transducer;

[0026] S205. Construct a signal sample by using the vector formed by the first signal feature, the second signal feature, the current temperature range, and the electromagnetic interference intensity range as the sample feature, and the water flow velocity as the sample label.

[0027] S206. Adjust the water flow speed multiple times, and repeat steps S201~S205 after each adjustment to obtain multiple signal samples under the combination of temperature range and electromagnetic interference intensity range.

[0028] In embodiments of this application, the EMD decomposition and denoising of the received ultrasonic signal includes:

[0029] The received ultrasonic signal is decomposed by EMD to obtain multiple IMF components and residuals.

[0030] The cross-correlation coefficient of each IMF component with the corresponding ultrasonic transmission signal is calculated. IMF components with cross-correlation coefficients greater than a preset threshold are selected and reconstructed with the residual to obtain the ultrasonic signal after EMD decomposition and denoising.

[0031] This denoising method can effectively remove irrelevant noise, thereby obtaining a denoised signal that is highly correlated with the ultrasonic emission signal. In subsequent feature construction or flow velocity calculation, it can make the construction or calculation results more accurate.

[0032] In an embodiment of this application, the first signal feature includes: a vector formed by the residual obtained after EMD decomposition of the ultrasonic signal received by the second transducer and the IMF component with a cross-correlation coefficient greater than a preset threshold.

[0033] In an embodiment of this application, the second signal feature includes: a vector formed by the residual obtained after EMD decomposition of the ultrasonic signal received by the first transducer and the IMF component with a cross-correlation coefficient greater than a preset threshold.

[0034] Step S3. For each combination of temperature range and electromagnetic interference intensity range, repeat step S2 to add all the obtained signal samples into the same set to form a sample set;

[0035] Step S4. Construct a water flow velocity prediction model based on a deep neural network, and train the water flow velocity prediction model using a sample set;

[0036] During the training process of the deep neural network, signal samples from the sample set are used as network inputs, and corresponding labels are used as expected outputs. The loss function is calculated, and network training is completed through backpropagation or gradient descent. This establishes the relationship between the signal features after EMD denoising, temperature, external electromagnetic environment, and water flow velocity. Since the signal features correspond to denoising information related to propagation time and flow velocity, the direct calculation of water flow velocity is cleverly transformed into prediction based on the deep neural network. Furthermore, the influence information of temperature and external electromagnetic environment can be incorporated into the prediction process, eliminating the need for table lookup interpolation compensation. Instead, the prediction of water flow velocity under different temperatures and external electromagnetic environments is completed directly in the model, and the measurement result is calculated from the water flow velocity, effectively improving the accuracy of the measurement.

[0037] Step S5. In the actual process of ultrasonic water meter measurement, ultrasonic signals are transmitted between the first transducer and the second transducer to extract signal features, and the current temperature range and electromagnetic interference intensity range are tested to form training samples. These samples are then input into the trained water flow velocity prediction model to obtain the water flow velocity. Finally, the water flow velocity is multiplied by the cross-sectional area of ​​the fluid in the ultrasonic water meter to complete the ultrasonic water meter measurement.

[0038] In the embodiments of this application, the first transducer is located upstream of the second transducer, and the signal transceiver terminals of the first and second transducers are directly opposite each other; the angle between the ultrasonic transceiver beam direction and the pipe axis direction is... The pipe axis is the direction of water flow in the ultrasonic water meter.

[0039] The foregoing description illustrates and describes a preferred embodiment of the present invention. However, as previously stated, it should be understood that the present invention is not limited to the forms disclosed herein and should not be construed as excluding other embodiments. It can be used in various other combinations, modifications, and environments, and can be altered within the scope of the inventive concept described herein through the foregoing teachings or techniques or knowledge in related fields. Any modifications and variations made by those skilled in the art that do not depart from the spirit and scope of the present invention should be within the protection scope of the appended claims.

Claims

1. An ultrasonic water meter metering method based on EMD decomposition and noise reduction, characterized in that: Includes the following steps: Step S1. Given the temperature range and the intensity range of external electromagnetic interference in the ultrasonic water meter; Step S2. For any combination of temperature range and electromagnetic interference intensity range, when water flows in the ultrasonic water meter, ultrasonic signals are transmitted between the first transducer and the second transducer, and signal features and water flow velocity are extracted to construct signal samples. Step S3. For each combination of temperature range and electromagnetic interference intensity range, repeat step S2 to add all the obtained signal samples into the same set to form a sample set; Step S4. Construct a water flow velocity prediction model based on a deep neural network, and train the water flow velocity prediction model using a sample set; Step S5. In the actual process of ultrasonic water meter measurement, ultrasonic signals are transmitted between the first transducer and the second transducer to extract signal features, and the current temperature range and electromagnetic interference intensity range are tested to form training samples. These samples are then input into the trained water flow velocity prediction model to obtain the water flow velocity. Finally, the water flow velocity is multiplied by the cross-sectional area of ​​the fluid in the ultrasonic water meter to complete the ultrasonic water meter measurement.

2. The ultrasonic water meter metering method based on EMD decomposition and noise reduction according to claim 1, characterized in that: The first transducer is located upstream of the second transducer, and the signal transceiver terminals of the first and second transducers are directly opposite each other; the angle between the ultrasonic transceiver beam direction and the pipe axis direction is... The pipe axis is the direction of water flow in the ultrasonic water meter.

3. The ultrasonic water meter metering method based on EMD decomposition and noise reduction according to claim 2, characterized in that: Step S2 includes: S201. For any combination of temperature range and electromagnetic interference intensity range, when water flows in the ultrasonic water meter, an ultrasonic signal is transmitted from the first transducer to the second transducer located downstream of the first transducer. The received ultrasonic signal is decomposed and denoised using EMD, and the first signal feature is extracted. Then, an ultrasonic signal is transmitted from the second transducer to the first transducer. The received ultrasonic signal is decomposed and denoised using EMD, and the second signal features are extracted. S202. For the denoised ultrasonic signal corresponding to the ultrasonic signal received by the second transducer, perform correlation peak detection with the ultrasonic signal emitted by the first transducer to obtain the position of the maximum correlation peak, and take the corresponding time as the downstream propagation time t1. S203. For the denoised ultrasonic signal corresponding to the ultrasonic signal received by the first transducer, perform correlation peak detection with the ultrasonic signal emitted by the second transducer to obtain the position of the maximum correlation peak, and take the corresponding time as the reverse propagation time t2. S204. Calculate the detected water flow velocity under the combined range of temperature and electromagnetic interference intensity: ； Where L represents the distance between the first transducer and the second transducer; S205. Construct a signal sample by using the vector formed by the first signal feature, the second signal feature, the current temperature range, and the electromagnetic interference intensity range as the sample feature, and the water flow velocity as the sample label. S206. Adjust the water flow speed multiple times, and repeat steps S201~S205 after each adjustment to obtain multiple signal samples under the combination of temperature range and electromagnetic interference intensity range.

4. The ultrasonic water meter metering method based on EMD decomposition and noise reduction according to claim 3, characterized in that: The EMD decomposition and denoising of the received ultrasonic signal includes: The received ultrasonic signal is decomposed by EMD to obtain multiple IMF components and residuals. The cross-correlation coefficient of each IMF component with the corresponding ultrasonic transmission signal is calculated. IMF components with cross-correlation coefficients greater than a preset threshold are selected and reconstructed with the residual to obtain the ultrasonic signal after EMD decomposition and denoising.

5. The ultrasonic water meter metering method based on EMD decomposition and noise reduction according to claim 4, characterized in that: The first signal feature includes: a vector consisting of the residual obtained after EMD decomposition of the ultrasonic signal received by the second transducer and the IMF components with a cross-correlation coefficient greater than a preset threshold.

6. The ultrasonic water meter metering method based on EMD decomposition and noise reduction according to claim 4, characterized in that: The second signal feature includes: a vector consisting of the residual obtained after EMD decomposition of the ultrasonic signal received by the first transducer and the IMF component with a cross-correlation coefficient greater than a preset threshold.

Images