Pipeline pressure signal denoising method

Abstract

The invention discloses a pipeline pressure signal denoising method. The method comprises the following steps: S1, acquiring a noise-containing pressure signal c in a pipeline; s2, calculating a mean value signal E of the noise-containing pressure signal c, replacing the noise-free signal a with the mean value signal E, and calculating an estimated signal # imgabs0 # of the noise signal b; s3, introducing a denoising regulation factor k, and calculating estimation signals # imgabs1 # and # imgabs2 # = c-k # imgabs3 #, 0lt of the noise pressure signal a; klt; 1; s4, calculating a noise level s according to the estimated signal # imgabs4 # and the estimated signal # imgabs5 #; comparing the noise level s with a set threshold value q; if s is smaller than q, the denoising process is ended, and the estimated signal # imgabs6 # is a noiseless pressure signal a; otherwise, the # imgabs7 is used for replacing c in # imgabs8 # = c-k # imgabs9 # in the step S3, iteration is repeated until the obtained noise level meets the condition that s is smaller than q, and the finally obtained estimated signal # imgabs10 # is the noise-free pressure signal a. According to the method, high-frequency noise components can be reduced, low-frequency useful signal components are reserved, sudden change and distortion are avoided, and compressor pipeline pressure pulsation evaluation is more accurate.

Description

Technical Field

[0001] The present invention relates to the technical field of pressure signal data processing, and more specifically, to a pipeline pressure signal denoising method. Background Art

[0002] Compressors are widely used in hydrogenation stations, high-purity gas, and chemical industries, and are key equipment. However, for compressors with reciprocating motion, such as mechanical reciprocating, piston reciprocating, and diaphragm compressors, there are pressure fluctuations in the pipeline, and it is necessary to test the size of the pressure fluctuations and evaluate the size of the pipeline pulsation to prevent excessive pulsation from causing damage to the pipeline system and its accessories. However, the pipeline system is a complex system, and the measured pipeline pressure signal noise usually contains a variety of noise components. The presence of noise will affect the evaluation of pulsation. Therefore, after measuring the pressure signal, it is necessary to denoise it. In the prior art, the CEEMDAN method decomposes the signal and removes internal noise. The WPT method is to set a wavelet energy threshold to remove smaller energy components, thereby achieving denoising. The SVD method is to set a singular value energy threshold to remove smaller energy components, thereby achieving denoising. The above denoising method still has the following problems: 1. The denoised curve still contains a lot of noise components; 2. The denoised curve may produce false mutations in local non-stationary areas, causing distortion and resulting in large errors; 3. Some useful components in the signal are removed, resulting in translation of the denoised signal. Summary of the invention

[0003] An object of the present invention is to solve at least the above problems and to provide at least the advantages which will be described hereinafter.

[0004] In order to achieve these purposes and other advantages according to the present invention, a pipeline pressure signal denoising method is provided, comprising the following steps: S1. Obtain the noisy pressure signal in the pipeline c , which contains the noise-free signal a and noise signal b ; S2. Calculate the noisy pressure signal c The mean signal E， And the mean signal E Replacement of noise-free signal a , calculate the noise signal b The estimated signal ; S3. Introducing denoising adjustment factors k , calculate the noise pressure signal a The estimated signal , =c - k , 0<k<1 ; S4. Calculate the noise level based on the estimated signal and the estimated signal ; the noise level is used to evaluate the denoising effect of the current estimated signal s relative to the noisy pressure signal ; c S5. Compare the noise level with the set threshold s ; if q < s < q , the denoising process ends, and the estimated signal is the noise-free pressure signal a ; otherwise, replace in step S3 =c - k with c , and repeat the iteration until the calculated noise level satisfies s < q , and the finally obtained estimated signal is the noise-free pressure signal a .

[0005] Preferably, the mean signal c of the noisy pressure signal E is calculated specifically through the following steps: S21. Calculate the upper envelope c and the lower envelope E 1 of the noisy pressure signal E 2 ; S22. Calculate the mean signal E , E= ( E 1 + E 2) / 2.

[0006] Preferably, the upper envelope E1 and the lower envelope E2 are constructed by extracting the upper and lower poles of the noisy pressure signal c and then using the interpolation method.

[0007] Preferably, in step S2, the estimated signal b of the noise signal is calculated by the following formula: = ( c^ 2 - E^ 2) / (2 E ).

[0008] Preferably, in step S4, the noise levels Calculated by the following formula: s =log 10 (P( ) / P( )) In the above formula, P( ) and P( ) are respectively the sum of the squares of the signal values in the estimated signal and the estimated signal .

[0009] The present invention has at least the following beneficial effects: For the pipeline pressure signal denoising method provided by the present invention, first calculate the mean signal of the pipeline pressure signal with noise, and on the basis of the mean signal, calculate the estimated signal b of the noise signal ; then calculate the estimated signal after denoising and the noise level s , and finally judge whether the current estimated signal after denoising meets the requirements 。 The above process uses the approximation method to gradually iterate and remove the small noise components in the pressure signal, so that the curve of the finally obtained pipeline pressure signal without noise is smooth, the high-frequency noise components are reduced, the low-frequency useful signal components are retained, no mutation and distortion occur, and the evaluation of the compressor pipeline pressure pulsation is more accurate.

[0010] Other advantages, objectives and features of the present invention will be partially reflected by the following description, and partially will be understood by those skilled in the art through the research and practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] Figure 1 is a flowchart of the pipeline pressure signal denoising method described in the present invention; Figure 2 is a schematic diagram of the pressure signal with noise in the pipeline in an embodiment of the present invention c ; Figure 3 is the upper envelope c of the pressure signal with noise in the above embodiment of the present invention E 1 and the lower envelope E 2 schematic diagram; Figure 4 and Figure 5 are the mean signals c of the pressure signal with noise in the above embodiment of the present invention E schematic diagram; Figure 6 and Figure 7The estimated signal after the first denoising in the above embodiments of the present invention Schematic diagram; Figure 8 The noise-free signal finally obtained after denoising in the above embodiments of the present invention a Schematic diagram; Specific implementation manner

[0012] The present invention will be further described in detail below with reference to the accompanying drawings, so that those skilled in the art can implement it according to the description in the specification.

[0013] It should be noted that the experimental methods described in the following embodiments are all conventional methods unless otherwise specified, and the reagents and materials can be obtained from commercial sources unless otherwise specified; in the description of the present invention, the terms "horizontal", "longitudinal", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", etc. indicate the orientation or positional relationship based on the orientation or positional relationship shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the present invention.

[0014] As Figure 1 shown, the present invention provides a method for denoising pipeline pressure signals, including the following steps: S1. Obtain the noisy pressure signal in the pipeline c , which includes the noise-free signal a and the noise signal b ; a, b and c are all equal to Non-zero vectors, the noisy pressure signal c can be expressed by the following formula: c = a + b (1) S2. Calculate the mean signal c of the noisy pressure signal E， and replace the mean signal E with the noise-free signal a , and calculate the estimated signal b of the noise signal ; Among them, the mean signal c of the noisy pressure signal E is specifically calculated through the following steps: S21. Calculate the upper envelope c and the lower envelope E 1 and the lower envelope E 2; The upper envelope E1 and the lower envelope E2 are extracted by extracting the noisy pressure signal c The upper and lower poles of are then interpolated to construct the upper and lower envelopes.

[0015] S22, calculate the mean signal E , E= ( E 1+ E 2) / 2 (2) From the above formula (1), we can get: c^2 = a^2 + 2a*b + b^2 (3) Considering that the noise component in the compressor pipeline pressure signal is a high-frequency and small energy, it is a small noise, that is, |b (i)|<<|a(i)| , and then we can get the following formula (3): b ≈( c^ 2- a^ 2) / (2 a ) (4) In step S2, the mean signal E Replacement of noise-free signal a Calculating the Noise Signal b The estimated signal , which is calculated by the following formula: = ( c^ 2- E^ twenty two E ) (5) S3. Introducing denoising adjustment factors k , calculate the noise pressure signal a The estimated signal , =c - k (6) In formula (6), 0 <k<1 ; k The larger the value, the faster the calculation process, but it may cause signal distortion.

[0016] S4. Based on the estimated signal and the estimated signal Calculating noise level s , the noise level is used to evaluate the current estimated signal Relative to the noisy pressure signal c The denoising effect; specifically, the noise level s Calculated by the following formula: s =log10 (P( ) / P( )) (7) In Equation (7), P( ) and P( ) are respectively the sum of the squares of each signal value in the estimated signal and the estimated signal . P( ) = Σ 2 (i) (8) P( ) = Σ 2 (i) (9) S5. Compare the noise level s with the set threshold q ; if s < q , the denoising process ends, and the estimated signal is the noise-free pressure signal a ; otherwise, replace in =c - k in step S3, and repeat steps S3 to S5 until the calculated noise level satisfies c < s < q . The finally obtained estimated signal is the noise-free pressure signal a . The denoising adjustment factor k and the threshold q are both empirical values, which can be obtained based on the test calculation results of historical data and combined with expert experience.

[0017] Taking the denoising process of the pressure signal in a certain compressor pipeline as an example, the denoising method of the pipeline pressure signal is specifically described. First, obtain the noisy pressure signal c in the pipeline, as shown in Figure 2 ; calculate the upper envelope c and the lower envelope E 1 of the noisy pressure signal E 2 , as shown in Figure 3 ; then calculate the mean signal E, as shown in Figure 4 and Figure 5 . In Figure 4 , the mean signal E and the original noisy pressure signal c are superimposed and displayed. Figure 5The average signal E is separately shown; the estimated signal after the first denoising is calculated according to Equations (5) and (6). 1 , as Figure 6 and Figure 7 shown, Figure 6 the estimated signal after the first denoising is 1 superposed and shown with the original noisy pressure signal c ; Figure 7 the estimated signal after the first denoising is 1 separately shown; then the noise level s 1 is calculated and compared with the set threshold q. In this example, after the first denoising s 1 ＞q , the estimated signal after the first denoising 1 is substituted into Equation (6) to successively calculate the estimated signal 2 after the second denoising and the noise level s2, completing the second iterative calculation; after multiple iterative calculations, finally satisfying s < q the estimated signal is as Figure 8 shown, and the finally denoised signal curve is smooth and retains the low-frequency useful signal components. The pipeline pressure signal denoising method can be applied in a compressor monitoring system to process the monitored compressor pipeline pressure signal, thereby further performing data evaluation and analysis, evaluating the pipeline pulsation magnitude, and preventing excessive pulsation from causing harm to the pipeline system and its accessories.

[0018] Although the embodiments of the present invention have been disclosed above, they are not limited to only the applications listed in the specification and embodiments. It can be fully applied to various fields suitable for the present invention. For those familiar with the field, additional modifications can be easily achieved. Therefore, without departing from the general concept defined by the claims and the equivalent scope, the present invention is not limited to the specific details and the illustrated examples here.

Claims

1. A pipeline pressure signal denoising method, characterized in that including the following steps: S1. Obtain the noisy pressure signal in the pipeline c , which includes the noise-free signal a and the noise signal b ; S2. Calculate the noisy pressure signal c of the mean signal E， and use the mean signal E to replace the noise-free signal a , and calculate the estimated signal b of the noise signal ; S3. Introduce a denoising adjustment factor k , calculate the estimated signal of the noise pressure signal a , , =c-k , 0<k<1 ; S4. Calculate the noise level based on the estimated signal and the estimated signal s , where the noise level is used to evaluate the denoising effect of the current estimated signal relative to the noisy pressure signal c ; S5. Compare the noise level s with the set threshold q ; If s < q , the denoising process ends, and the estimated signal is the noise-free pressure signal a ; otherwise, replace in step S3 =c-k with c , and repeat the iteration until the calculated noise level satisfies s < q . The finally obtained estimated signal is the noise-free pressure signal a .

2. The pipeline pressure signal denoising method according to claim 1, wherein Noisy pressure signal c mean signal E Specifically, it is calculated through the following steps: S21. Calculate the noisy pressure signal c of the upper envelope E 1 and the lower envelope E 2 ; S22. Calculate the mean signal E , E= ( E 1 + E 2) / 2 3. The pipeline pressure signal denoising method according to claim 2, characterized in that, The upper envelope E1 and the lower envelope E2 are constructed by extracting the respective upper and lower poles of the noisy pressure signal c and then using the interpolation method to construct the upper envelope and the lower envelope.

4. The pipeline pressure signal denoising method according to claim 1, characterized in that, In step S2, the estimated signal of the noise signal b is calculated by the following formula: ​ = ( c^ 2- E^ 2) / (2 E )。 5. The pipeline pressure signal denoising method according to claim 1, wherein In step S4, the noise level s is calculated by the following formula: s =log 10 (P( ) / P( )) In the above formula, P( ) and P( ) are respectively the sum of the squares of the values of each signal in the estimated signal and the estimated signal .

Images