Rheometer signal processing method, program product and storage medium

Abstract

The invention discloses a rheometer signal processing method, a program product and a storage medium, and the method comprises the steps: obtaining a plurality of torque data of a rheometer, detecting and replacing an outlier in the torque data through a window moving method, carrying out the noise reduction of the processed torque data through a wavelet threshold, and outputting the data. According to the method, abnormal fluctuation data is processed by adopting a moving window method, and the influence of high-frequency noise on a result is reduced by utilizing wavelet threshold noise reduction; according to the data processed by the method, signal fluctuation is removed, and the interference of special conditions such as aggregate interlocking on the data collected by the rheometer is reduced, so that the data better conforms to the test condition.

Description

Technical Field

[0001] The present invention relates to the technical field of signal processing, and in particular to a signal processing method, a program product and a storage medium for a rheometer. Background Art

[0002] With the introduction of fluid mechanics to the in-depth study of the flow properties of concrete, concrete is often considered as a Bingham fluid, which means that it will flow under the action of shear stress, and the relationship between its shear stress and shear rate shows a linear relationship. Based on the Bingham model principle, rheometers are developed and manufactured to test the rheological properties of concrete, and their performance is evaluated by plastic viscosity and dynamic shear stress. The basic principle of the rheometer is to set different speeds, measure the torque at the corresponding speed, fit the corresponding curve of torque-speed, and convert the relationship between torque-speed into the relationship between shear stress-shear rate through the mathematical equation of the rheological model to obtain the rheological parameters, namely yield stress and plastic viscosity.

[0003] Coaxial twin-cylinder rheometers are often designed with a size four times the maximum nominal particle size in order to reduce the situation in which aggregate jamming occurs during actual use, which would lead to a sudden increase in collected data. However, concrete is a large-particle fluid, and the aggregate moves outward during the test, making it difficult for the rheometer to ensure a stable signal during actual use, and data distortion caused by signal fluctuations and noise is prone to occur. Summary of the invention

[0004] In view of the above-mentioned deficiencies in the prior art, the technical problem to be solved by the present invention is: how to provide a signal processing method for a rheometer that can remove fluctuation signals, reduce noise interference, and improve data acquisition accuracy.

[0005] In order to solve the above technical problems, the present invention adopts the following technical solutions:

[0006] A signal processing method for a rheometer obtains a plurality of torque data of the rheometer, detects and replaces outliers in the torque data by using a window shifting method, and then uses a wavelet threshold to reduce noise on the processed torque data before outputting the data.

[0007] As an optimization, in the process of using the moving window method to detect and replace outliers in the torque data, the moving median method is used as the evaluation method for outliers to detect abnormal torque data in the torque data. The data filling method uses backward value taking to ensure the smoothness of the torque data, which is used to eliminate abnormal torque data with large fluctuations. The trial and error method is used to determine the window size of the moving window method.

[0008] As an optimization, when performing wavelet threshold denoising, db2 is selected as the wavelet basis function for wavelet decomposition. After a certain number of decompositions, threshold processing is performed. The threshold processing uses two threshold functions, four threshold selection methods, and two threshold readjustment methods to combine parameters for denoising. The threshold function includes a soft threshold function and a hard threshold function. The formula of the soft threshold function is:

[0009]

[0010] Where sgn(w) is 1 when w>0 and -1 when w<0, w is a variable, and thr is a non-negative threshold;

[0011] The hard threshold function formula is:

[0012]

[0013] The four threshold selection methods include rigrsure, minimaxi, sqtwolog and heursure. The two threshold readjustment methods include one, which does not change with noise, and mln, which adjusts according to the noise estimation of each layer of wavelet decomposition. After wavelet threshold processing, the signal is reconstructed. The signal-to-noise ratio and mean square error of the reconstructed signal are used as the fitting evaluation indicators of the torque data. The denoised signal with the largest noise ratio and the smallest mean square error is selected as the wavelet threshold reconstruction result, and the denoised signal is output. The signal-to-noise ratio calculation formula is:

[0014]

[0015] Where P signal is a useful signal; P noise is the noise signal; Y i Predicted value; a i Actual value;

[0016] Mean square error calculation formula:

[0017]

[0018] where Y i Predicted value; a i Actual value.

[0019] A computer program product comprises a computer program, and when the computer program is executed by a computer, the method described above is implemented.

[0020] A computer-readable storage medium stores a computer program, which, when executed by a computer, implements the method described above.

[0021] Compared with the prior art, the present invention has the following advantages: the present invention adopts a moving window method to process abnormal fluctuation data, and utilizes wavelet threshold noise reduction to reduce the influence of high-frequency noise on the results. Compared with the traditional data processing method of the rheometer which takes the average by time period, the data processed by the present invention removes signal fluctuations and reduces the interference of special situations such as aggregate embedding on the data collected by the rheometer, making the data more consistent with the test situation. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 It is a schematic diagram of the process of the present invention;

[0023] Figure 2 This is a schematic diagram of the wavelet threshold denoising principle in the present invention;

[0024] Figure 3 This is a diagram showing the effect of different window sizes on data smoothing when the trial and error method is used in the present invention;

[0025] Figure 4 This is a comparison diagram between the wavelet packet denoising result and the original signal in the present invention;

[0026] Figure 5 The results of matching pursuit analysis of four groups of wavelet basis vectors in the present invention are shown in FIG.

[0027] Figure 6 Each component diagram of the wavelet analysis in the present invention;

[0028] Figure 7 This is the integrated diagram of each component of the wavelet analysis in the present invention. DETAILED DESCRIPTION

[0029] In order to make the purpose, technical scheme and advantages of the embodiments of the present invention clearer, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. The components of the embodiments of the present invention generally described and shown in the drawings here can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present invention provided in the drawings is not intended to limit the scope of the claimed invention, but merely represents selected embodiments of the present invention. Based on the embodiments in the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work belong to the scope of protection of the present invention.

[0030] like Figure 1 As shown, the signal processing method of the rheometer in this specific embodiment obtains several torque data of the rheometer, which may be relevant data of concrete or similar large-particle fluids, and uses a shifting window method to detect and replace outliers in the torque data. The processed torque data is then denoised using a wavelet threshold and then output.

[0031] In this specific implementation, in the process of using the moving window method to detect and replace outliers in the torque data, the moving median method is used as the outlier evaluation method to detect abnormal torque data in the torque data. The data filling method uses backward value taking to ensure the smoothness of the torque data, which is used to eliminate abnormal torque data with large fluctuations. The trial and error method is used to determine the window size of the moving window method.

[0032] In this specific implementation, when performing wavelet threshold denoising, db2 is selected as the wavelet basis function for wavelet decomposition, and threshold processing is performed after a certain number of decompositions. The threshold processing uses two threshold functions, four threshold selection methods, and two threshold readjustment methods to combine parameters for denoising. The threshold function includes a soft threshold function and a hard threshold function. The formula of the soft threshold function is:

[0033]

[0034] Where sgn(w) is 1 when w>0 and -1 when w<0, w is a variable, and thr is a non-negative threshold;

[0035] The hard threshold function formula is:

[0036]

[0037] The four threshold selection methods include rigrsure, minimaxi, sqtwolog and heursure. The two threshold readjustment methods include one, which does not change with noise, and mln, which adjusts according to the noise estimation of each layer of wavelet decomposition. After wavelet threshold processing, the signal is reconstructed. The signal-to-noise ratio and mean square error of the reconstructed signal are used as the fitting evaluation indicators of the torque data. The denoised signal with the largest noise ratio and the smallest mean square error is selected as the wavelet threshold reconstruction result, and the denoised signal is output. The signal-to-noise ratio calculation formula is:

[0038]

[0039] Where P signal is a useful signal; P noise is the noise signal; Y i Predicted value; a i Actual value;

[0040] Mean square error calculation formula:

[0041]

[0042] where Y i Predicted value; a i Actual value.

[0043] A computer program product comprises a computer program, and when the computer program is executed by a computer, the method described above is implemented.

[0044] A computer-readable storage medium stores a computer program, which, when executed by a computer, implements the method described above.

[0045] Take a section of rheometer torque data, the total amount of data is 3500. There are fluctuation noises caused by aggregate embedding and noises in the test process in the data. First, the moving window method is used to find and eliminate the outliers in the signal (the outliers are generated by aggregate embedding). The moving median method is used as the evaluation method for outliers. The data filling method uses backward value to ensure data smoothness. The trial and error method is used to determine the window size of the moving window method. The window size is set to 10, 100, 500, and 1000 for trial and error. Figure 3 As shown in the figure, the window selection of the moving window method is selected by trial and error. The window should not be too small, otherwise the noise reduction purpose cannot be achieved. The window selection should not be too large, otherwise the data will be too smooth and the signal characteristics of each stage will not be reflected. It is found that when the window size is 500, the signal characteristics can be retained and outliers can be effectively removed.

[0046] The processed signal is subjected to secondary processing using wavelet threshold noise reduction. Before wavelet analysis, it is necessary to select a suitable wavelet basis function. In this embodiment, four groups of wavelet functions, haar, db2, sym4, and coif2, are selected to process the signal. The four groups of wavelet functions are made into wavelet packets, such as Figure 4 As shown in Figure 2, the matching pursuit method is used to select the wavelet basis function suitable for the rheometer torque signal, such as Figure 5 As shown, db2 is finally selected as the wavelet basis function of wavelet analysis, and db2 is used to perform wavelet decomposition on the primary processing signal.

[0047] When performing wavelet decomposition on a signal, the larger the number of decomposition layers, the more obvious the different characteristics of noise and signal performance, and the length also decreases accordingly. Therefore, it is necessary to select an appropriate number of decomposition layers. After the torque data is decomposed by wavelet six times, the signal characteristics are obvious. The component signals after six decompositions are thresholded, such as Figure 6 and Figure 7 As shown in the figure. Two threshold functions (soft threshold function and hard threshold function), four threshold processing methods (rigrsure, minimaxi, sqtwolog and heursure) and two threshold adjustment methods (one and mln) are used to combine 16 signal processing methods. The signal-to-noise ratio and mean square error are used as the evaluation indicators of the signal. After processing the signal, 16 output results can be obtained, and the signal with high signal-to-noise ratio and small mean square error is selected as the final output signal.

[0048]

[0049] M represents the maximum signal-to-noise ratio, and U represents the combination number of the maximum signal-to-noise ratio and the minimum mean square error. By processing the torque data of the rheometer, the aggregate embedding problem existing in the coaxial double-cylinder rheometer test process can be effectively solved, and the noise signal can be effectively reduced, making the signal more accurate and the data more precise.

[0050] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described with reference to the preferred embodiments of the present invention, it should be understood by those skilled in the art that various changes may be made in form and details without departing from the spirit and scope of the present invention as defined in the appended claims.

Claims

1. A signal processing method for a rheometer, characterized in that: Several torque data of the rheometer are obtained, and the outliers in the torque data are detected and replaced by the window shifting method. Then the processed torque data is denoised by wavelet threshold and then output.

2. The signal processing method of the rheometer according to claim 1, characterized in that: In the process of using the moving window method to detect and replace outliers in torque data, the moving median method is used as the evaluation method for outliers to detect abnormal torque data in the torque data. The data filling method uses backward value taking to ensure the smoothness of the torque data, which is used to eliminate abnormal torque data with large fluctuations. The trial and error method is used to determine the window size of the moving window method.

3. The signal processing method of the rheometer according to claim 1, characterized in that: When performing wavelet threshold denoising, db2 is selected as the wavelet basis function for wavelet decomposition. After a certain number of decompositions, threshold processing is performed. The threshold processing uses two threshold functions, four threshold selection methods, and two threshold readjustment methods to combine parameters for denoising. The threshold function includes a soft threshold function and a hard threshold function. The formula of the soft threshold function is: Where sgn(w) is 1 when w>0 and -1 when w<0, w is a variable, and thr is a non-negative threshold; The hard threshold function formula is: Four threshold selection methods include rigrsure, minimaxi, sqtwolog and heursure. Two threshold readjustment methods include one, which does not change with noise, and mln, which adjusts according to the noise estimation of each layer of wavelet decomposition. After wavelet threshold processing, the signal is reconstructed. The signal-to-noise ratio and mean square error of the reconstructed signal are used as the fitting evaluation indicators of the torque data. The denoised signal with the largest noise ratio and the smallest mean square error is selected as the wavelet threshold reconstruction result, and the denoised signal is output. Signal-to-noise ratio calculation formula: Where P signal is a useful signal; P noise is the noise signal; Y i Predicted value; a i Actual value; Mean square error calculation formula: where Y i Predicted value; a i Actual value.

4. A computer program product, characterized in that: The method comprises a computer program, which, when executed by a computer, implements the method according to any one of claims 1 to 3.

5. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a computer, the method according to any one of claims 1 to 3 is implemented.

Images