Step signal detection method and system

Abstract

The invention provides a step signal detection method and system, and the method comprises the steps: S1, carrying out the stability test of a first test condition and a second test condition, and entering a next step if the test passes; wherein the noise of the first test condition is smaller than that of the second test condition; s2, according to the result of the stability test, calculating to obtain a calibration value; s3, performing step detection under a second test condition to obtain first detection data, and further obtaining a fitting function; s4, performing data screening according to the calibration value and the fitting function, and determining abnormal data; s5, performing regression estimation on the abnormal data to obtain a regression value, and replacing the abnormal data with the regression value to obtain second detection data; s6, performing step feature extraction on the second detection data to obtain third detection data; and S7, performing calculation according to the third detection data to obtain a plurality of step heights, and further obtaining a target step signal. The method has the advantages of good universality, obvious steps, few miscellaneous points, no staggered layer and the like.

Description

Technical Field

[0001] The present disclosure relates to the technical field of data processing, and in particular to a step signal detection method and system. Background Art

[0002] The control system changes the control signal through the driving mechanism, and then controls the detection mechanism to detect the control signal and collect the test signal. Due to the existence of objective factors such as interference, the test signal cannot truly reflect the control signal. For example, in the test of the step amount, the test signal is in the shape of a step. Since the test signal contains noise, the step feature is often not obvious. The conventional step signal test method is aimed at signals with large step amounts and cannot measure step signals with small step amounts. Even when the step amount is large enough, although obvious step features can be generated, this method is only applicable to the case where the test signal changes linearly. Once the test signal changes nonlinearly, a step signal that does not match the actual control signal will be collected, that is, it is impossible to extract a test signal that matches the actual signal. Therefore, the traditional step signal test method cannot be used for high-precision testing of control signals.

[0003] Generally speaking, during the step signal detection process, the situation with less noise is usually when a single system is working or when other systems have very little interference with the system. This situation is generally rare in actual work. It is more common when multiple systems work together or other systems have external interference with the system. When testing a single system, there is generally only internal noise and the noise energy is small. At this time, the signal detection will be relatively accurate. However, when multiple systems work together, there may be large-scale external interference, with many and complex interference sources, large signal energy, strong randomness, long duration, and large interference amplitude. The step signal generated by a small step amount originally has the characteristic of an unclear step shape. A large amplitude of interference will make the step shape completely non-existent.

[0004] The time window and frequency window of wavelet transform can be changed. When low-frequency information is needed, a long time window is used, and when high-frequency information is needed, a short time window is used. Therefore, wavelet transform can effectively extract information. However, wavelet transform is mainly used to suppress noise and interference in the overall signal data. It is difficult to extract local large-amplitude interference data and retain large areas of original data.

[0005] Since the steps of the test signal generated by the small step amount are not obvious, it is impossible to confirm whether the screened data is step data or interference data, so the interference data cannot be directly screened out and discarded, but the normal value needs to be estimated using the current interference data. In the conventional regression method, the average value of the estimated quantity remains unchanged, but the overall continuous step signal shows a certain trend (the average value of each step is increasing or decreasing within a certain range), so a method is needed to estimate the mean of the current data. The conventional regression method uses the method of averaging multiple groups of data before and after, which is likely to blur the step characteristics, making the step assessment more difficult. Due to the aforementioned reasons, when the control signal is small, the step increment of the corresponding test signal will have a nonlinear growth. Due to the fuzzy step characteristics, the conventional regression method will extract a test signal that does not match the actual control signal and cannot extract an accurate step test signal.

[0006] Based on the above analysis, how to remove large-scale interference data and retain a large area of ​​original data in order to retain the step characteristics as much as possible and achieve high-precision step signal detection is a topic that R&D personnel continue to pay attention to.

[0007] The above information disclosed in this section is only for understanding the background of the inventive concept of the present disclosure and therefore the above information may contain information that does not constitute the prior art. Summary of the invention

[0008] In view of this, the first aspect of the present disclosure provides a step signal detection method, including: S1, performing stability tests on the first test condition and the second test condition respectively, and proceeding to the next step if the test passes; wherein, the noise of the first test condition is less than the noise of the second test condition; S2, calculating and obtaining a calibration value according to the result of the stability test; S3, performing step detection under the second test condition to obtain first detection data, and then obtaining a fitting function; S4, performing data screening according to the calibration value and the fitting function to determine abnormal data; S5, performing regression estimation on the abnormal data to obtain a regression value, replacing the abnormal data with the regression value, and obtaining second detection data; S6, performing step feature extraction on the second detection data to obtain third detection data; S7, calculating and obtaining multiple step heights according to the third detection data, and then obtaining a target step signal.

[0009] According to an embodiment of the present disclosure, S1 includes: S11, performing stability tests on the first test condition and the second test condition respectively to obtain multiple segments of first test data and multiple segments of second test data; S12, calculating the standard deviation according to the multiple segments of first test data and the multiple segments of second test data respectively to obtain multiple segments of first standard deviation and multiple segments of second standard deviation; S13, calculating the range according to the multiple segments of first standard deviation and the multiple segments of second standard deviation respectively to obtain the first range and the second range; S14, judging whether the first range and the second range meet the preset conditions respectively, and if so, the test passes and proceeds to the next step.

[0010] According to an embodiment of the present disclosure, S14 includes: S141, judging whether the first extreme difference is less than the first threshold; S142, judging whether the second extreme difference is less than the second threshold; S143, when the judgment results of S141 and S142 are both yes, it is judged that the stability test has passed and proceeds to the next step.

[0011] According to an embodiment of the present disclosure, S2 includes: S21, calculating the mean according to multiple segments of first standard deviations to obtain a first mean; S22, calculating the mean according to multiple segments of second standard deviations to obtain a second mean; S23, calculating the ratio of the second mean to the first mean to obtain a calibration value.

[0012] According to an embodiment of the present disclosure, S3 includes: S31, performing step detection under second test conditions to obtain first detection data; S32, performing function fitting on the overall data according to the trend of the first detection data to obtain a fitting function; wherein the fitting function includes one of a straight line function, a curve function, a step function, and a piecewise function.

[0013] According to an embodiment of the present disclosure, S4 includes: S41, traversing the horizontal coordinate values ​​in the first detection data, and calculating the corresponding function values ​​according to the fitting function; S42, determining the data filtering range according to the function value and the calibration value; S43, judging whether the signal data corresponding to the horizontal coordinate value in the first detection data is within the data filtering range, if not, it is abnormal data.

[0014] According to an embodiment of the present disclosure, S5 includes: S51, calculating the correlation coefficients of multiple segments of data according to multiple segments of first test data and multiple segments of second test data respectively; S52, calculating the mean according to the correlation coefficients of the multiple segments of data; S53, calculating the corresponding regression value according to the function value corresponding to the abnormal data and the mean; S54, replacing the abnormal data with the regression value to obtain the second test data.

[0015] According to an embodiment of the present disclosure, S6 includes: S61, performing discrete wavelet transform on the second detection data to obtain denoised data; S62, reconstructing the denoised data to obtain third detection data.

[0016] According to an embodiment of the present disclosure, S7 includes: S71, using a unilateral method or a histogram method to calculate the third detection data to obtain a plurality of step amounts; S72, subtracting adjacent step amounts to obtain a step height, and then obtaining a target step signal.

[0017] According to an embodiment of the present disclosure, after S7, the step further includes: S8, calculating an average value of a plurality of step heights, where the average value is used to evaluate the estimated resolution of the step signal.

[0018] The second aspect of the present disclosure provides a step signal detection system, including: a stability test module, used to perform stability tests on a first test condition and a second test condition respectively, and proceed to the next step if the test passes; wherein the noise of the first test condition is less than the noise of the second test condition; a first calculation module, used to calculate and obtain a calibration value according to the result of the stability test; a step detection module, used to perform step detection under the second test condition, obtain first detection data, and then obtain a fitting function; a data screening module, used to perform data screening according to the calibration value and the fitting function, and determine abnormal data; a regression estimation module, used to perform regression estimation on the abnormal data to obtain a regression value, and replace the abnormal data with the regression value to obtain second detection data; a feature extraction module, used to perform step feature extraction on the second detection data to obtain third detection data; a second calculation module, used to calculate multiple step heights according to the third detection data, and then obtain a target step signal.

[0019] The step signal detection method and system disclosed in the present invention firstly judge the validity of the test signal through stability test to confirm that the test data is available; then calculate the calibration value and fitting function to filter out abnormal data; adopt regression estimation to ensure that the overall trend of the signal remains unchanged and the step characteristics are retained; finally, extract and calculate the step characteristics to obtain the target step signal, thereby realizing high-precision step signal detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] The above and other objects, features and advantages of the present disclosure will become more apparent through the following description of the embodiments of the present disclosure with reference to the accompanying drawings, in which:

[0021] Figure 1 The flowchart of the step signal detection method according to the embodiment of the present disclosure is schematically shown;

[0022] Figure 2 A schematic diagram schematically shows step signal detection of force and deformation according to an embodiment of the present disclosure;

[0023] Figure 3 A schematic diagram schematically shows first test data in an embodiment of the present disclosure;

[0024] Figure 4A schematic diagram schematically shows second test data in an embodiment according to the present disclosure;

[0025] Figure 5 A schematic diagram schematically shows first detection data under a first working condition according to an embodiment of the present disclosure;

[0026] Figure 6 A schematic diagram of a fitting function curve under a first working condition according to an embodiment of the present disclosure is schematically shown;

[0027] Figure 7 A schematic diagram schematically shows abnormal data screened out under a first working condition according to an embodiment of the present disclosure;

[0028] Figure 8 A schematic diagram schematically shows second detection data under a first working condition according to an embodiment of the present disclosure;

[0029] Fig. 9A A schematic diagram showing the result of step extraction using the method of the present disclosure under the first working condition in an embodiment of the present disclosure is shown;

[0030] Fig. 9B A schematic diagram showing the result of directly using the original signal to perform step extraction under the first working condition according to an embodiment of the present disclosure is shown;

[0031] Fig. 10A A schematic diagram showing the result of performing step amount statistics using the method of the present disclosure under the first working condition according to an embodiment of the present disclosure is shown;

[0032] Fig. 10B A schematic diagram schematically shows the result of directly using the original signal to perform step amount statistics under the first working condition according to an embodiment of the present disclosure;

[0033] Fig.11A A schematic diagram showing the result of step extraction using the method of the present disclosure under the second working condition in an embodiment of the present disclosure is shown;

[0034] Fig. 11B A schematic diagram schematically shows the result of directly using the original signal to perform step extraction under the second working condition according to an embodiment of the present disclosure;

[0035] Fig. 12A The schematic diagram schematically shows the result of step extraction using the method of the present disclosure under the third working condition in an embodiment of the present disclosure;

[0036] Fig. 12B The schematic diagram schematically shows the result of directly using the original signal to perform step extraction under the third working condition according to the embodiment of the present disclosure;

[0037] Fig.13A The schematic diagram schematically shows the result of step extraction using the method of the present disclosure under the fourth working condition in an embodiment of the present disclosure;

[0038] Fig. 13B The schematic diagram schematically shows the result of directly using the original signal to perform step extraction under the fourth working condition according to the embodiment of the present disclosure;

[0039] Fig.14A A schematic diagram schematically shows the result of performing step amount statistics using the method of the present disclosure under the second working condition according to an embodiment of the present disclosure;

[0040] Fig. 14B A schematic diagram schematically shows the result of directly using the original signal to perform step amount statistics under the second working condition according to an embodiment of the present disclosure;

[0041] Fig.15A The schematic diagram schematically shows the result of performing step amount statistics using the method of the present disclosure under the third working condition according to an embodiment of the present disclosure;

[0042] Fig. 15B The schematic diagram schematically shows the result of directly using the original signal to perform step amount statistics under the third working condition according to the embodiment of the present disclosure;

[0043] Fig.16A The schematic diagram schematically shows the result of performing step amount statistics using the method of the present disclosure under the fourth working condition according to an embodiment of the present disclosure;

[0044] Fig. 16B The schematic diagram schematically shows the result of directly using the original signal to perform step amount statistics under the fourth working condition according to the embodiment of the present disclosure;

[0045] Fig.17 The structural block diagram of the step signal detection system according to the embodiment of the present disclosure is schematically shown. DETAILED DESCRIPTION

[0046] In order to make the objectives, technical solutions and advantages of the present disclosure more clearly understood, the present disclosure is further described in detail below in combination with specific embodiments and with reference to the accompanying drawings.

[0047] The terms used herein are only for describing specific embodiments and are not intended to limit the present disclosure. The terms "comprise", "include", etc. used herein indicate the existence of the features, steps, operations and / or components, but do not exclude the existence or addition of one or more other features, steps, operations or components.

[0048] All terms (including technical and scientific terms) used herein have the meanings commonly understood by those skilled in the art unless otherwise defined. It should be noted that the terms used herein should be interpreted as having a meaning consistent with the context of this specification and should not be interpreted in an idealized or overly rigid manner.

[0049] Some block diagrams and / or flow charts are shown in the accompanying drawings. It should be understood that some boxes or combinations thereof in the block diagrams and / or flow charts can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that these instructions can create a device for implementing the functions / operations described in these block diagrams and / or flow charts when executed by the processor. The technology of the present disclosure can be implemented in the form of hardware and / or software (including firmware, microcode, etc.). In addition, the technology of the present disclosure can take the form of a computer program product on a computer-readable storage medium storing instructions, which can be used by an instruction execution system or in combination with an instruction execution system.

[0050] The ordinal numbers used in the specification and claims, such as "first", "second", "third", etc., to modify the corresponding elements, do not themselves mean or represent any ordinal number of the element, nor do they represent the order of one element and another element, or the order of manufacturing methods. The use of these ordinal numbers is only used to clearly distinguish a component with a certain name from another component with the same name.

[0051] Figure 1 The flowchart of the step signal detection method according to the embodiment of the present disclosure is schematically shown.

[0052] like Figure 1 As shown, the step signal detection method may include operations S1 to S7.

[0053] In operation S1 , stability tests are performed on the first test condition and the second test condition respectively, and if the tests pass, the next step is entered; wherein the noise of the first test condition is smaller than the noise of the second test condition.

[0054] It should be noted that the first test condition includes the test condition when a single system is working or other systems have very little interference with the system, and the noise is relatively small under the first test condition; the second test condition includes the test condition when multiple systems are working in coordination or other systems have greater external interference with the system, and the noise is relatively large under the second test condition. Operation S1 quantifies the noise and interference introduced by the coordinated operation of multiple systems through stability testing, confirms that the interference is within a controllable range, and confirms that the collected test data is meaningful. This operation is used to determine the validity of the test signal. When the interference is large enough, if the valid data cannot be identified, then the obtained test signal loses its meaning. This operation can determine the validity of the test signal and confirm that the test data is available through stability testing.

[0055] In operation S2, a calibration value is calculated based on the result of the stability test.

[0056] The calibration value represents the difference between the data dispersion when multiple systems work together and when a single system works. By calculating the difference range through the difference, it is judged whether the detected data exceeds the difference range, and abnormal data can be easily screened out. The purpose of obtaining the calibration value is to screen out abnormal data.

[0057] In operation S3, step detection is performed under the second test condition to obtain first detection data, and then obtain a fitting function.

[0058] There are many optional signal fitting functions, which can be selected as needed. According to the signal trend, a more appropriate function can be selected for fitting, which is conducive to improving the accuracy of data screening and regression estimation.

[0059] In operation S4, data screening is performed according to the calibration value and the fitting function to determine abnormal data.

[0060] In data screening, the function value of the fitting function plus or minus multiples of the calibration value are used as the upper and lower thresholds of data screening. This multiple can be selected according to the required data stability.

[0061] In operation S5, regression estimation is performed on the abnormal data to obtain a regression value, and the regression value is used to replace the abnormal data to obtain second detection data.

[0062] Using the function value of the fitting function instead of the mean in the regression estimation can ensure that the overall trend of the signal remains unchanged and retain the step characteristics. If regression estimation is not performed, the original signal may be misaligned, that is, the step signal cannot reflect the detection signal, such as the detection signal is increasing, while the step signal shows an abnormal decrease.

[0063] In operation S6, step feature extraction is performed on the second detection data to obtain third detection data.

[0064] The signal processed as above filters out most of the interference information, retains a large area of ​​original data and highlights the overall step signal change trend. However, the step signal is still not obvious, so step feature extraction is performed to obtain data with more obvious step features.

[0065] In operation S7 , a plurality of step heights are calculated according to the third detection data, and then a target step signal is obtained.

[0066] Through the above detection method, high-precision step signal detection is achieved.

[0067] The detection method of the step signal is described in detail below.

[0068] On the basis of the above embodiment, S1 includes: S11, performing stability tests on the first test condition and the second test condition respectively to obtain multiple segments of first test data and multiple segments of second test data; S12, calculating the standard deviation according to the multiple segments of first test data and the multiple segments of second test data respectively to obtain multiple segments of first standard deviation and multiple segments of second standard deviation; S13, calculating the range according to the multiple segments of first standard deviation and the multiple segments of second standard deviation respectively to obtain the first range and the second range; S14, judging whether the first range and the second range meet the preset conditions respectively, and if so, the test passes and proceeds to the next step.

[0069] Collect multiple sections of first test data with low noise and second test data with high noise to conduct stability test. Stability test includes collecting n sections of test data when single system (Single) works and when multiple systems (Multi) work together under the same test environment, calculating the standard deviation of each section of test data, and then calculating the range of the standard deviation. When the range meets the set threshold, the stability test passes and data calibration can be performed.

[0070] For the n segments of first test data collected when a single system (Single) is working, a set of n segments of first test data is obtained as shown in formula (1).

[0071] (1)

[0072] The amount of data in each of the above test data may be different. In this embodiment, 300 test data are taken for convenience of calculation.

[0073] Calculate the standard deviation of each segment of the first test data, as shown in formula (2). is the mean of the first test data in each segment, is the number of the first test data in each segment.

[0074] (2)

[0075] The first standard deviation set when the single system works is obtained as shown in formula (3), and the first range is calculated as shown in formula (4). The first threshold is set as , when the first extreme value ( ) is less than the first threshold, it indicates that the test signal can be stabilized within the first threshold range, and the size of the first threshold is determined by the specific system conditions.

[0076] (3)

[0077] (4)

[0078] For the n segments of second test data collected when multiple systems (Multi) work together, a set of n segments of second test data is obtained as shown in formula (5).

[0079] (5)

[0080] Calculate the second standard deviation of each segment of the second test data, as shown in formula (6). is the mean of the second test data in each segment, is the number of second test data in each segment.

[0081] (6)

[0082] The second standard deviation set when multiple systems work together is obtained as shown in formula (7), and the second range is calculated as shown in formula (8). The second threshold is set as , when the second extreme ( ) is less than the second threshold, it indicates that the test signal can be stabilized within the second threshold range, and the size of the second threshold is determined by the specific system conditions.

[0083] (7)

[0084] (8)

[0085] Based on the above embodiment, S14 includes: S141, judging whether the first extreme difference is less than the first threshold; S142, judging whether the second extreme difference is less than the second threshold; S143, when the judgment results of S141 and S142 are both yes, it is judged that the stability test has passed and proceeds to the next step.

[0086] When formula (9) holds true, the stability test passes. This test is to ensure that the test signal and the noise and interference therein fluctuate within a fixed range. Only such test data is meaningful.

[0087] (9)

[0088] Based on the above embodiment, S2 includes: S21, calculating the mean according to multiple segments of first standard deviation to obtain a first mean; S22, calculating the mean according to multiple segments of second standard deviation to obtain a second mean; S23, calculating the ratio of the second mean to the first mean to obtain a calibration value.

[0089] Perform data calibration and determine the calibration value. Calculate the means of the two standard deviation sets as shown in formulas (10) and (11).

[0090] (10)

[0091] (11)

[0092] The ratio of the mean of the standard deviation set of multi-system signal data and single-system signal data is the calibration value. , as shown in formula (12). Used to determine the difference in data dispersion between multi-system collaborative work and single system work.

[0093] (12)

[0094] Based on the above embodiment, S3 includes: S31, performing step detection under the second test condition to obtain first detection data; S32, performing function fitting on the overall data according to the trend of the first detection data to obtain a fitting function; wherein the fitting function includes one of a straight line function, a curve function, a step function, and a piecewise function.

[0095] Perform overall signal fitting. Under the condition of multi-system collaborative work, perform step detection to obtain the first detection data of step detection, whose expression is shown in formula (13).

[0096] (13)

[0097] According to the general trend of the signal, a function is fitted to the overall signal data to obtain the fitting function expression as shown in formula (14), where x is the horizontal coordinate and y is the fitting value corresponding to the horizontal coordinate.

[0098] (14)

[0099] Based on the above embodiment, S4 includes: S41, traversing the horizontal coordinate values ​​in the first detection data, and calculating the corresponding function values ​​according to the fitting function; S42, determining the data filtering range according to the function value and the calibration value; S43, judging whether the signal data corresponding to the horizontal coordinate value in the first detection data is within the data filtering range, if not, it is abnormal data.

[0100] Data screening is performed according to the fitting function, the horizontal coordinate values ​​in the first detection data are traversed, and the corresponding function value is calculated according to the fitting function. The multiples of the function value plus or minus the calibration value are used as the upper and lower thresholds for data screening. If the signal data is greater than the upper threshold or less than the lower threshold, it is determined to be abnormal data, as shown in formula (15). Wherein, It is a multiple of the calibration value and can be set according to the discreteness of the data. When the stability requirement is high, You can choose a multiple less than 1, and when some data values ​​​​change suddenly, You can choose a multiple greater than 1. It should be noted that when the multiple is too small, the step features may be compressed.

[0101] (15)

[0102] On the basis of the above embodiment, S5 includes: S51, calculating the correlation coefficients of multiple segments of data according to multiple segments of first test data and multiple segments of second test data respectively; S52, calculating the mean according to the correlation coefficients of the multiple segments of data; S53, calculating the corresponding regression value according to the function value corresponding to the abnormal data and the mean; S54, replacing the abnormal data with the regression value to obtain the second test data.

[0103] Regression estimation is performed on abnormal data. Based on the stability test data of single system and multiple systems, the correlation coefficients of multiple test data are calculated, as shown in formula (16). The set of correlation coefficients obtained according to formula (16) is shown in formula (17).

[0104] (16)

[0105] (17)

[0106] Among them, P z ≤P j , P z It can be selected as needed, that is, a section of continuous test data is selected from multiple sections of first test data and multiple sections of second test data.

[0107] Calculating Collections The mean of is shown in formula (18). Perform regression estimation.

[0108] (18)

[0109] The fitting function value corresponding to the horizontal coordinate of the current abnormal data is used as the average value of the estimated quantity in the regression method, and the regression value is calculated. Abnormal, that is, the horizontal coordinate value x in the first detection data is k, and the signal data corresponding to the horizontal coordinate value is , then the regression value is calculated as shown in formula (19), where Indicates the first detection data The mean of Indicates the corresponding fitting function value when the horizontal axis is k.

[0110] (19)

[0111] Using regression values Instead of this abnormal data, all the screened abnormal data are regressed and estimated, and the corresponding abnormal data are replaced by the regression value to obtain the second detection data, whose expression is As shown in formula (20), the second detection data can keep the overall signal trend unchanged, ensuring more accurate step assessment.

[0112] (20)

[0113] Wherein, m is the number of data of the first detection data. The number of data of the first detection data is the same as the number of data of the second detection data.

[0114] On the basis of the above embodiment, S6 includes: S61, performing discrete wavelet transform on the second detection data to obtain denoised data; S62, reconstructing the denoised data to obtain third detection data.

[0115] After denoising and reconstructing the second detection data using a wavelet threshold method, a step waveform with more obvious step features can be obtained. Select a suitable wavelet basis and decomposition series to decompose and reconstruct the regression data.

[0116] The basic expression of Discrete Wavelet Transform (DWT) is shown in formula (21), where: are the wavelet coefficients, is the second detection data and the signal to be decomposed. is the wavelet basis function.

[0117] (twenty one)

[0118] (twenty two)

[0119] The basic expression of the reconstruction process is shown in formula (22), where: is the decomposition series, is the jth level, kth wavelet coefficient. The process is: use low-pass and high-pass decomposition filters to decompose the signal layer by layer, and then use low-pass and high-pass reconstruction filters to reconstruct the signal layer by layer.

[0120] On the basis of the above embodiment, S7 includes: S71, using a unilateral method or a histogram method to calculate the third detection data to obtain a plurality of step amounts; S72, subtracting adjacent step amounts to obtain a step height, and then obtaining a target step signal.

[0121] The step height can be analyzed by the unilateral method or the histogram method, and multiple step quantities can be obtained. The subtraction of the two step quantities before and after can obtain multiple step heights, forming a height set. .

[0122] Based on the above embodiment, after S7, the following further includes: S8, calculating an average value of a plurality of step heights, where the average value is used to evaluate the estimated resolution of the step signal.

[0123] The detection method disclosed in the present invention has clear step data, which has obvious advantages for systems that need to measure nonlinear resolution. The method also has the advantages of good universality, obvious steps, few noise points, and no mismatch.

[0124] The present disclosure is further described below through specific implementations. The above-mentioned step signal detection method is specifically described in the following embodiments. However, the following embodiments are only used to illustrate the present disclosure, and the scope of the present disclosure is not limited thereto.

[0125] Step 1: Use the step signal detection method disclosed in the present invention to detect the step signal of the force and the template deformation variable and estimate the resolution, such as Figure 2 As shown, the stability test is carried out under the same humidity and temperature and the same template material: when the single system (only the force is working, and the structure of the clamping template is self-closed loop) is working, the force is 10N, and the first test data of 3 sections are collected, as shown in FIG. Figure 3 As shown; when multiple systems (structural vibration will affect the large-scale shaking of the structure of the clamping template) work together, add 10N of force and collect 3 sections of the second test data as shown Figure 4 shown.

[0126] The standard deviations of the first test data and the second test data are calculated respectively. The first standard deviations and the second standard deviations are shown in Table 1. Set the first threshold =0.2, the second threshold =0.8, as can be seen from Table 1, the first range is less than the first threshold, and the second range is less than the second threshold, and the test data meets the stability requirement. This is equivalent to the above step S1.

[0127] Table 1

[0128]

[0129] Step 2: Perform data calibration, calculate the mean of the first standard deviation of multiple segments and the second standard deviation of multiple segments, and obtain the first mean , the second mean , calculate the calibration value . This is equivalent to the above step S2.

[0130] Step 3: Under the multi-system collaborative working condition, that is, the second test condition, the initial force is set to 10N, and the step detection is performed under the first working condition, that is, step force is added, the step amount is 1N, the step is 12 times, and the step interval time is consistent. The first detection data is obtained as follows Figure 5 shown.

[0131] According to the relationship between force and deformation, the greater the force, the greater the deformation. The first detection data should show a step-by-step increasing trend. The linear fitting function of the least squares method can be used to characterize the signal change trend. The fitted function expression is shown in formula (3-1). The fitting function curve is shown in Figure 6 This is equivalent to the above step S3.

[0132] (3-1)

[0133] Step 4: Traverse the first test data, calculate the function value corresponding to the horizontal coordinate value in the first test data, and use the function value plus or minus the multiple of the calibration value as the upper and lower thresholds for data screening. If the signal data corresponding to the horizontal coordinate value is greater than the upper threshold or less than the lower threshold, it is determined to be abnormal data. , known , when the signal data corresponding to the corresponding horizontal coordinate value satisfies any one of the formulas (4-1), the data is considered to be abnormal data. The abnormal data screened out are as follows Figure 7 This is equivalent to the above step S4.

[0134] (4-1)

[0135] Step 5: Perform regression estimation on the screened abnormal data, calculate the correlation coefficients of the three stability test data segments, and average the three correlation coefficients to obtain , take the fitting function value of the coordinate corresponding to the current abnormal data as the average value of the estimated quantity in the regression method, calculate the regression value as shown in formula (5-1), and replace the abnormal data with the regression value to obtain the second detection data (regression data) as shown in Figure 8 This is equivalent to the above step S5.

[0136]

[0137] (5-1)

[0138] Step 6: After denoising and reconstructing the second detection data using a wavelet threshold-based method, a step waveform with more obvious step features can be obtained. Since the test signal should originally be in a step shape, that is, a square wave type, after repeated attempts, this embodiment uses the Haar wavelet basis to perform a 7-level wavelet decomposition on the regression data, uses a soft threshold to reduce the noise, and uses DWT and reconstructed data to obtain the third detection data, which is step extraction, as shown in Figure 9. In contrast, the result of directly using the original signal for step extraction is as follows: Fig. 9B This is equivalent to the above step S6.

[0139] from Fig. 9B It can be seen from ① that when the original signal is directly used for step extraction, there will be a situation where the steps are backward, that is, the test signal that does not match the actual situation is extracted.

[0140] Step 7: This embodiment uses the histogram line method to count the number of steps, such as Fig. 10A In contrast, the result of directly using the original signal for step extraction is shown in Fig. 10B As shown. The step height data obtained by subtracting the previous step amount from the latter step amount is shown in Table 2, and the mean of the step height is calculated. Under the current working condition, under a force of 10N, the estimated resolution of the deformation amount corresponding to each increase of 1N is 3.365nm. The steps using the method disclosed in the present invention are clear and clean, and there is no problem of step staggering. For signals with unknown step number and step width, the method disclosed in the present invention has better extraction results. Equivalent to the above-mentioned step S7.

[0141] Table 2

[0142]

[0143] At the same time, step detection is also performed under the other three working conditions, with a step size of 1N, 12 steps, and a consistent step interval, and the obtained step data is tested and compared. Under the second working condition, the step extraction results of step detection are as follows: Fig.11A The step amount statistics are shown in FIG14A. In contrast, the result of directly using the original signal to extract the step amount is shown in FIG14A. Fig. 11B The step amount statistics are shown in FIG14B; the step height data of step extraction using the original signal directly and the method disclosed in the present invention are shown in Table 3. Under the third working condition, the step extraction results of step detection are shown in FIG14B. Fig. 12A The statistical results of step volume are shown in Fig.15A As shown; in contrast, the result of directly using the original signal for step extraction is as follows Fig. 12B The statistical results of step volume are shown in Fig. 15BThe step height data of step extraction using the original signal directly and the method disclosed in the present invention are shown in Table 4. Under the fourth working condition, the step extraction result of step detection is shown in Table 4. Fig.13A The statistical results of step volume are shown in Fig.16A As shown; in contrast, the result of directly using the original signal for step extraction is as follows Fig. 13B The statistical results of step volume are shown in Fig. 16B As shown; the step height data obtained by directly using the original signal and using the method disclosed in the present invention for step extraction are shown in Table 5.

[0144] Table 3

[0145]

[0146] Table 4

[0147]

[0148] Table 5

[0149]

[0150] It can be seen from the above multiple test results that the method disclosed in the present invention has the advantages of good universal adaptability, obvious steps, few noise points, and no staggered layers.

[0151] Fig.17 A block diagram of a step signal detection system according to an embodiment of the present disclosure is schematically shown.

[0152] The present disclosure also provides a step signal detection system, comprising:

[0153] The stability test module is used to perform stability tests on the first test condition and the second test condition respectively, and proceed to the next step if the test passes; wherein the noise of the first test condition is smaller than the noise of the second test condition;

[0154] A first calculation module, used to calculate and obtain a calibration value according to the result of the stability test;

[0155] A step detection module, used for performing step detection under a second test condition to obtain first detection data and then obtain a fitting function;

[0156] A data screening module is used to screen data based on calibration values ​​and fitting functions to determine abnormal data;

[0157] A regression estimation module, used for performing regression estimation on the abnormal data to obtain a regression value, and replacing the abnormal data with the regression value to obtain second detection data;

[0158] A feature extraction module, used for performing step feature extraction on the second detection data to obtain third detection data;

[0159] The second calculation module is used to calculate a plurality of step heights according to the third detection data, and then obtain a target step signal.

[0160] like Fig.17 As shown, the step signal detection system 1700 includes: a stability testing module 1710, a first calculation module 1720, a step detection module 1730, a data screening module 1740, a regression estimation module 1750, a feature extraction module 1760 and a second calculation module 1770.

[0161] The stability test module 1710 is used to perform stability tests on the first test condition and the second test condition respectively, and proceed to the next step if the test passes; wherein the noise of the first test condition is less than the noise of the second test condition. According to an embodiment of the present disclosure, the stability test module 1710 can be used to perform the above reference Figure 1 The S1 step described is not repeated here.

[0162] The first calculation module 1720 is used to calculate and obtain the calibration value according to the result of the stability test. According to an embodiment of the present disclosure, the first calculation module 1720 can be used to perform the above reference Figure 1 The described S2 step will not be repeated here.

[0163] The step detection module 1730 is used to perform step detection under the second test condition to obtain the first detection data and then obtain the fitting function. According to the embodiment of the present disclosure, the step detection module 1730 can be used to perform the above reference Figure 1 The S3 step described is not repeated here.

[0164] The data screening module 1740 is used to screen the data according to the calibration value and the fitting function to determine the abnormal data. According to the embodiment of the present disclosure, the data screening module 1740 can be used to perform the above reference Figure 1 The described S4 step will not be repeated here.

[0165] The regression estimation module 1750 is used to perform regression estimation on the abnormal data to obtain a regression value, and replace the abnormal data with the regression value to obtain the second detection data. According to an embodiment of the present disclosure, the regression estimation module 1750 can be used to perform the above reference Figure 1 The described step S5 will not be repeated here.

[0166] The feature extraction module 1760 is used to extract step features from the second detection data to obtain third detection data. According to an embodiment of the present disclosure, the feature extraction module 1760 can be used to perform the above reference Figure 1The described S6 step will not be repeated here.

[0167] The second calculation module 1770 is used to calculate multiple step heights according to the third detection data, and then obtain the target step signal. According to an embodiment of the present disclosure, the second calculation module 1770 can be used to perform the above reference Figure 1 The described step S7 will not be repeated here.

[0168] It should be noted that any one or more of the modules, sub-modules, units, and sub-units according to the embodiments of the present disclosure, or at least part of the functions of any one of them, can be implemented in one module. Any one or more of the modules, sub-modules, units, and sub-units according to the embodiments of the present disclosure can be split into multiple modules for implementation. Any one or more of the modules, sub-modules, units, and sub-units according to the embodiments of the present disclosure can be at least partially implemented as a hardware circuit, such as a field programmable gate array (FPGA), a programmable logic array (PLA), a system on a chip, a system on a substrate, a system on a package, an application-specific integrated circuit (ASIC), or can be implemented by hardware or firmware in any other reasonable way of integrating or packaging the circuit, or in any one of the three implementation methods of software, hardware, and firmware, or in any appropriate combination of any of them. Alternatively, one or more of the modules, sub-modules, units, and sub-units according to the embodiments of the present disclosure can be at least partially implemented as a computer program module, and when the computer program module is run, the corresponding function can be performed.

[0169] For example, any of the stability test module 1710, the first calculation module 1720, the step detection module 1730, the data screening module 1740, the regression estimation module 1750, the feature extraction module 1760, and the second calculation module 1770 can be combined into one module for implementation, or any of the modules can be split into multiple modules. Alternatively, at least part of the functions of one or more of these modules can be combined with at least part of the functions of other modules and implemented in one module. According to an embodiment of the present disclosure, at least one of the stability test module 1710, the first calculation module 1720, the step detection module 1730, the data screening module 1740, the regression estimation module 1750, the feature extraction module 1760 and the second calculation module 1770 can be at least partially implemented as a hardware circuit, such as a field programmable gate array (FPGA), a programmable logic array (PLA), a system on a chip, a system on a substrate, a system on a package, an application-specific integrated circuit (ASIC), or can be implemented by hardware or firmware such as any other reasonable way of integrating or packaging the circuit, or implemented in any one of the three implementation methods of software, hardware and firmware or in any appropriate combination of any of them. Alternatively, at least one of the stability test module 1710, the first calculation module 1720, the step detection module 1730, the data screening module 1740, the regression estimation module 1750, the feature extraction module 1760 and the second calculation module 1770 can be at least partially implemented as a computer program module, and when the computer program module is run, the corresponding function can be performed.

[0170] It will be appreciated by those skilled in the art that the features described in the various embodiments and / or claims of the present disclosure may be combined and / or combined in a variety of ways, even if such combinations and / or combinations are not explicitly described in the present disclosure. In particular, the features described in the various embodiments and / or claims of the present disclosure may be combined and / or combined in a variety of ways without departing from the spirit and teachings of the present disclosure. All of these combinations and / or combinations fall within the scope of the present disclosure.

[0171] Although the present disclosure has been shown and described with reference to specific exemplary embodiments of the present disclosure, it should be understood by those skilled in the art that various changes in form and details may be made to the present disclosure without departing from the spirit and scope of the present disclosure as defined by the appended claims and their equivalents. Therefore, the scope of the present disclosure should not be limited to the above-mentioned embodiments, but should be determined not only by the appended claims, but also by the equivalents of the appended claims.

Claims

1. A step signal detection method, characterized in that: include: S1, respectively perform stability tests on the first test condition and the second test condition, and proceed to the next step if the tests pass; wherein the noise of the first test condition is less than the noise of the second test condition; S2, calculating and obtaining a calibration value according to the result of the stability test; S3, performing step detection under the second test condition to obtain first detection data, and then obtaining a fitting function; S4, screening data according to the calibration value and the fitting function to determine abnormal data; S5, performing regression estimation on the abnormal data to obtain a regression value, and replacing the abnormal data with the regression value to obtain second detection data; S6, extracting step features from the second detection data to obtain third detection data; S7, calculating and obtaining a plurality of step heights according to the third detection data, and then obtaining a target step signal.

2. The step signal detection method according to claim 1, characterized in that: S1 includes: S11, performing stability tests on the first test condition and the second test condition respectively to obtain multiple sections of first test data and multiple sections of second test data; S12, calculating standard deviations according to the multiple sections of first test data and the multiple sections of second test data, respectively, to obtain multiple sections of first standard deviations and multiple sections of second standard deviations; S13, respectively calculating the range according to the multiple first standard deviations and the multiple second standard deviations to obtain a first range and a second range; S14, respectively judging whether the first range and the second range meet the preset conditions, if so, the test passes and proceeds to the next step.

3. The step signal detection method according to claim 2, characterized in that: S14 includes: S141, determining whether the first range is less than a first threshold; S142, determining whether the second range is less than a second threshold; S143, when the judgment results of S141 and S142 are all yes, it is judged that the stability test has passed and proceeds to the next step.

4. The step signal detection method according to claim 2, characterized in that S2 include: S21, calculating the mean according to the first standard deviations of the multiple segments to obtain a first mean; S22, calculating the mean according to the second standard deviations of the multiple segments to obtain a second mean; S23, calculating the ratio of the second mean value to the first mean value to obtain the calibration value.

5. The step signal detection method according to claim 1, characterized in that S3 include: S31, performing step detection under the second test condition to obtain first detection data; S32, performing function fitting on the overall data according to the trend of the first detection data to obtain a fitting function; wherein the fitting function includes one of a straight line function, a curve function, a step function, and a piecewise function.

6. The step signal detection method according to claim 1, characterized in that: S4 include: S41, traversing the horizontal coordinate values ​​in the first detection data, and calculating the corresponding function value according to the fitting function; S42, determining a data screening range according to the function value and the calibration value; S43, determining whether the signal data corresponding to the horizontal coordinate value in the first detection data is within the data screening range, if not, it is abnormal data.

7. The step signal detection method according to claim 2, characterized in that S5 include: S51, calculating correlation coefficients of multiple segments of data respectively according to the multiple segments of first test data and the multiple segments of second test data; S52, calculating a mean value according to the correlation coefficients of the plurality of data segments; S53, calculating a corresponding regression value according to the function value corresponding to the abnormal data and the mean value; S54, replacing the abnormal data with the regression value to obtain second detection data.

8. The step signal detection method according to claim 1, characterized in that S6 include: S61, performing discrete wavelet transform on the second detection data to obtain denoised data; S62, reconstructing the denoised data to obtain third detection data.

9. The step signal detection method according to claim 1, characterized in that: S7 includes: S71, calculating the third detection data using a unilateral method or a histogram method to obtain a plurality of step amounts; S72, subtracting the adjacent step amounts to obtain a step height, and then obtaining a target step signal.

10. The step signal detection method according to claim 1, characterized in that: S7 and later also include: S8, calculating an average value of the plurality of step heights, where the average value is used to evaluate the estimated resolution of the step signal.

11. A step signal detection method system, characterized in that: include: A stability test module, used to perform stability tests on the first test condition and the second test condition respectively, and proceed to the next step if the tests pass; wherein the noise of the first test condition is less than the noise of the second test condition; A first calculation module, used for calculating and obtaining a calibration value according to the result of the stability test; A step detection module, used for performing step detection under a second test condition to obtain first detection data and then obtain a fitting function; A data screening module, used to screen data according to the calibration value and the fitting function to determine abnormal data; A regression estimation module, used for performing regression estimation on the abnormal data to obtain a regression value, and replacing the abnormal data with the regression value to obtain second detection data; A feature extraction module, used for performing step feature extraction on the second detection data to obtain third detection data; The second calculation module is used to calculate a plurality of step heights according to the third detection data, and then obtain a target step signal.

Images