Method for automatically processing shock wave pressure signal

Abstract

The invention discloses a method for automatically processing a shock wave pressure signal. The method comprises the following steps: step 1, reading shock wave pressure data; 2, searching a local maximum value in the shock wave pressure data; 3, calculating the characteristic quantity of a local maximum value region; 4, determining a shock wave peak value according to each related characteristic quantity; 5, performing zero line drift processing based on the shock wave peak value; 6, based on the shock wave pressure data and the time corresponding to the shock wave peak value, the shock wave arrival time and the shock wave end time are estimated; corresponding shock wave pressure data are obtained; 7, waveform abnormal region processing is carried out on the shock wave pressure data; and step 8, automatically extracting the characteristic parameters of the shock waves based on the processed shock wave pressure data. The method for automatically processing the shock wave pressure signals is high in efficiency and accuracy, can process the pressure signals from a plurality of sensors in batches, and is easy to standardize.

Description

Technical Field

[0001] This invention belongs to the field of pattern recognition and signal processing technology, particularly the field of information processing technology for explosion shock wave pressure signals, and specifically relates to a method for automatically processing shock wave pressure signals. Background Art

[0002] In weapon and ammunition power testing, characteristic parameters such as the overpressure peak value, impulse, and barometric pressure duration of shock waves are important means of evaluating weapon and ammunition power. Rapidly and accurately extracting shock wave characteristic parameters is of great significance for ammunition damage assessment.

[0003] Due to the harsh testing environment, the measured shock wave signals are affected by noise, fragment impacts, etc., making the shock wave signals difficult to identify and the estimation of shock wave characteristic parameters inaccurate, seriously affecting the correct assessment of weapon and ammunition power. Currently, the mainstream method for shock wave data processing is manual extraction of characteristic parameters. The disadvantages of this method are, firstly, low efficiency, failing to achieve real-time data analysis and transmission to provide rapid and effective support for decision-makers; and secondly, manual extraction of characteristic parameters relies heavily on personal experience, making it susceptible to the influence of individual subjective factors, leading to discrepancies in data processing results and affecting the reliability of ammunition power assessment. Therefore, current explosion shock wave processing technology lacks an automated method that can identify shock wave pressure signals, process abnormal data appearing in the signals, and finally output shock wave characteristic parameters. Summary of the Invention

[0004] To address the problems of low efficiency, high dependence on manual extraction of shock wave characteristic parameters, susceptibility to errors, and difficulty in standardization in existing methods, the present invention aims to provide an automatic shock wave pressure signal method, including automatic identification of shock wave signals, correction of abnormal data in the signal caused by fragments or noise, and automatic extraction of shock wave characteristic parameters.

[0005] To achieve the above objectives, the present invention employs the following technical solution:

[0006] A method for automatically processing shock wave pressure signals specifically includes the following steps:

[0007] Step 1: Read the shock wave pressure data. The shock wave pressure data includes time data and its corresponding shock wave pressure data. There are n different times, t1, t2, ..., tn. n The corresponding shock wave pressure data are p1, p2, ..., p n ;

[0008] Step 2: Locate local maxima in the shock wave pressure data;

[0009] Step 3: Calculate the characteristic quantities of the local maxima region;

[0010] Step 4: Determine the peak value of the shock wave based on the relevant characteristic quantities;

[0011] Step 5: Based on the peak value of the shock wave, perform zero-line drift processing to obtain the shock wave pressure data after zero-line drift processing;

[0012] Step 6: Based on the shock wave pressure data obtained in Step 5 and the time corresponding to the peak value of the shock wave, estimate the arrival time and end time of the shock wave; based on the obtained arrival time and end time of the shock wave, obtain the corresponding shock wave pressure data;

[0013] Step 7: Perform waveform anomaly processing on the shock wave pressure data obtained in Step 6 to obtain the processed shock wave pressure data;

[0014] Step 8: Based on the shock wave pressure data processed in Step 7, automatically extract the characteristic parameters of the shock wave.

[0015] Compared with the prior art, the method for automatically processing shock wave pressure signals of the present invention has the following advantages in the following two aspects:

[0016] (1) The automatic processing method for shock wave pressure signals of the present invention can automatically identify shock wave pressure signals and automatically correct zero line drift and abnormal data caused by fragments or noise in the signals.

[0017] (2) The automatic processing method for shock wave pressure signals of the present invention can automatically process shock wave pressure data from multiple sensors in batches.

[0018] In summary, compared with the prior art, the method of the present invention improves efficiency, reduces dependence, and the automated operation is less prone to errors and easier to standardize. Attached Figure Description

[0019] Figure 1 This is a flowchart of the automatic processing method for shock wave pressure signals according to the present invention.

[0020] Figure 2 It is a time history curve of shock wave pressure that includes local maxima.

[0021] Figure 3 It is a time history curve of shock wave pressure with abnormal waveform.

[0022] Figure 4 This is a flowchart for handling abnormal shock wave waveforms.

[0023] Figure 5 This is a pressure time history curve after correction of the abnormal region of the shock wave waveform. Detailed Implementation

[0024] The present invention will be further described in detail below with reference to the embodiments. It should be noted that the present invention is not limited to the following specific embodiments. Those skilled in the art can make various other specific modifications and combinations based on the technical teachings disclosed in the present invention without departing from the essence of the present invention. These modifications and combinations are still within the protection scope of the present invention.

[0025] like Figure 1 As shown, the method for automatically processing shock wave pressure signals provided by this invention specifically includes the following steps:

[0026] Step 1: Read the shock wave pressure data. The shock wave pressure data includes time data and its corresponding shock wave pressure data. There are n different times, t1, t2, ..., tn. n The corresponding shock wave pressure data are p1, p2, ..., p n .

[0027] Step 2: Find local maxima in the shock wave pressure data.

[0028] In this embodiment, step 2 specifically includes the following sub-steps:

[0029] Step 2-1: Traverse the shock wave pressure data to find all local maximum pressure values ​​of the shock wave; specifically, a local maximum pressure value p i It refers to p i It is greater than the pressure data of the two directly adjacent shock waves, i.e., p i-1 <p i <p i+1 ;

[0030] Step 2-2: Given a height threshold Th1, filter out local maximum pressure values ​​that are less than this threshold;

[0031] Steps 2-3: If the distance between two local maxima is less than a given threshold Th2, filter out the smaller of the two local maxima. Specifically, the distance between two local maxima refers to the time difference between these two values, for example, p. i With p j Let be two local maxima, and the distance between them be |t|. i -t j |

[0032] Steps 2-1, 2-2, and 2-3 above are merely one embodiment of the present invention, and the scope of protection of the present invention is not limited to the definition and search method of such local maxima.

[0033] Step 3: Calculate the characteristic quantities of the local maxima region. Assume that step 2 found m local maxima: Calculate the relevant features within each local maximum region, including:

[0034] (1) Local maxima: k = 1, 2, ..., m;

[0035] (2) α% width: Given a percentage α% (the default value in this embodiment is 5%), calculate the threshold. For the k-th local maximum From its corresponding time Find the moment corresponding to the first pressure value less than Th3 to the left. From its corresponding time Search to the right for the moment corresponding to the first pressure value less than Th3. α% width

[0036] (3) Ascent time: corresponding time The time corresponding to the first pressure value less than Th3 on its left The time difference between them;

[0037] (4) Rate of ascent: from time t: The corresponding pressure value rises to The average speed;

[0038] (5) Downward trend: From time point At the time The correlation coefficient between pressure value data and time.

[0039] by Figure 2 For example, in Figure 2 There are 5 local maxima in the data. The feature quantities of each local maxima region calculated under a 5% threshold are shown in Table 1.

[0040] Table 1 Local maxima-related features

[0041] Maximum value number Local maxima 5% width Rise time speed of ascent Downtrend 1 0.07 32 6 0.009 -0.69 2 0.10 168 19 0.005 -0.86 3 0.45 3012 7 0.059 -0.94 4 0.52 3012 122 0.004 -0.95 5 0.29 3012 372 0.001 -0.89

[0042] The feature quantity of the local maxima in step 3 above is only one example of the present invention, and the scope of protection of the present invention is not limited to the above feature quantity.

[0043] Step 4: Determine the peak value of the shock wave based on the relevant characteristic quantities.

[0044] Step 4 specifically includes the following sub-steps:

[0045] Step 4-1: Assign scores to the m local maxima regions using various relevant feature quantities. The higher the score, the more likely the local maxima are shock wave peaks.

[0046] (1) Assign points to the m local maxima according to their local maxima. Specifically: the smallest local maximum is worth 1 point, the second smallest is worth 2 points, and so on, with the largest being worth m points.

[0047] (2) Assign points to the m local maxima according to their width. Specifically: the smallest width is worth 1 point, the second smallest width is worth 2 points, and so on, with the largest width being worth m points.

[0048] (3) Assign points to the m local maxima according to their rise time. Specifically: the one with the longest rise time is worth 1 point, the one with the second longest rise time is worth 2 points, and so on, with the one with the shortest rise time being worth m points.

[0049] (4) Assign points to the m local maxima according to their rate of ascent. Specifically: the slowest rate is 1 point, the second slowest rate is 2 points, and so on, with the fastest rate being m points.

[0050] (5) Assign points to the m local maxima according to their downward trend. Specifically: the largest downward trend value is worth 1 point, the second largest downward trend value is worth 2 points, and so on, with the smallest downward trend value being worth m points.

[0051] Step 4-2: Calculate the weighted average score for each local maximum.

[0052] Specifically, let m be the number of local maxima and d be the number of relevant features. kj The score w is the score obtained for the j-th eigenvalue of the k-th local maximum. j Let the weights corresponding to j features satisfy the following condition: The weights are set by the user, and the default value is [value missing]. The weighted average score of the k-th local maximum is then calculated as follows:

[0053]

[0054] Step 4-3: Calculate the probability of the shock wave peak based on the weighted average score of each local maximum, and determine the shock wave peak.

[0055] Specifically, the probability of the shock wave peak is calculated using the softmax function as follows:

[0056]

[0057] The local maximum value with the highest probability of a shock wave peak is the shock wave peak value.

[0058] Based on Table 1, the scores, weighted scores, and shock wave peak probabilities of each characteristic parameter for each local maximum are shown in Table 2. The third local maximum has the highest probability, so this maximum is determined as the shock wave peak value.

[0059] Table 2. Fractions of local maxima and probability of shock wave peaks

[0060]

[0061] Step 4 above is an embodiment of the present invention. The scope of protection of the present invention is not limited to the above-described method for assigning local maximum value regions and the method for calculating shock wave probability.

[0062] Step 5: Based on the peak value of the shock wave, perform zero-line drift processing to obtain the shock wave pressure data after zero-line drift processing.

[0063] Specifically, in complex environments, sensors and testing systems are easily affected by external environmental factors such as temperature, vibration, and electromagnetic interference, causing zero-line drift. This ultimately affects the calculation results of the peak value and impulse of the shock wave characteristic signal, thus requiring zero-line drift correction. The correction method is as follows: select a period of time before the peak time of the shock wave and calculate the average pressure during this period; then subtract the calculated average pressure from the shock wave pressure data in step 1 to obtain the shock wave pressure data after zero-line drift processing. For the sake of simplicity, p1, p2, ..., p are still used here. n This represents the n different times t1, t2, ..., t after zero-line drift processing. n The corresponding shock wave pressure data.

[0064] Step 6: Based on the shock wave pressure data obtained in Step 5 and the time corresponding to the peak value of the shock wave, estimate the arrival time and end time of the shock wave; based on the obtained arrival time and end time of the shock wave, obtain the corresponding shock wave pressure data.

[0065] Specifically, let t be the time corresponding to the peak value of the shock wave after the zero-line drift treatment in step 5. k , will be from t k Search for the first satisfaction on the left and The moment As an estimate of the arrival time of the shock wave; from t k Search for the first satisfaction on the right and The moment As an estimate of the end time of the shock wave. Time interval. This is called the barotropic region of the shock wave.

[0066] Step 7: Perform waveform anomaly processing on the shock wave pressure data obtained in Step 6 to obtain the processed shock wave pressure data.

[0067] The shock wave from an explosion can vary due to the characteristics of the explosion source, the properties of the propagation medium, and the influence of fragments during the explosion. The most common anomaly is an upward bulge at the falling edge of the shock wave pressure time history curve, such as... Figure 3 Regions 1 and 2 are shown in the diagram. These waveform anomalies affect the calculation of the shock wave impulse; therefore, appropriate processing of these anomalies is required. The steps are as follows: Figure 4 As shown, the details are as follows:

[0068] Step 7-1: Read the shock wave pressure data from the peak time to the end time from the processed shock wave data obtained in Step 6.

[0069] Let the pressure data of the shock wave in the positive pressure region after step 6 be: The corresponding peak time is t = t k The end time is The following steps are for time zones Shock wave pressure value p t Process it.

[0070] Step 7-2: Fit the read shock wave pressure data using an exponential function to obtain the fitted data.

[0071] Specifically, assuming the shock wave pressure data starts from time t k At that time It conforms to the following exponential decay relationship:

[0072]

[0073] Here, parameters a and b are unknown parameters.

[0074] shock wave pressure data p t Perform exponential function fitting. Specifically, let the estimates of a and b be respectively... and p t Fitted values

[0075] Step 7-3: Detect abnormal areas.

[0076] Based on the shock wave pressure data p obtained in step 7-1 t and the fitted data from step 7-2 Calculate residuals If a time region t exists a <t<t b If all residuals in a region are greater than zero, and there exists a residual greater than a given threshold α, then this region is an outlier region.

[0077] Step 7-4: Refit the exponential function to obtain new fitted values.

[0078] Specifically, using the data obtained in step 7-1 excluding the data in the outlier regions, an exponential function fitting is performed again to obtain estimates of the unknown parameters a and b. and Then, substitute these values ​​into the equation from step 7-2 to obtain the shock wave pressure values ​​p for all regions, including the abnormal region. t Fitted values: And calculate the residuals

[0079] Step 7-5: Re-detect and correct abnormal areas.

[0080] Perform the same abnormal region detection process as in step 7-3. If an abnormal region exists, then within this region, use the fitted value obtained in step 7-4. Replace the corresponding p t If no abnormal area exists, the shock wave pressure data is p from step 7-1. t constant.

[0081] Figure 5 This is a pressure-time history curve after processing for the abnormal regions, where regions 1 and 2 are respectively... Figure 3 The curves after processing abnormal region 1 and abnormal region 2.

[0082] The application of the exponential function formula in steps 7-2 and 7-4 above is merely one embodiment of the present invention. The scope of protection of the present invention is not limited to the exponential fitting method described in this embodiment; fitting of other functions still falls within the scope of protection of the present invention.

[0083] Step 8: Based on the shock wave pressure data processed in Step 7, automatically extract the characteristic parameters of the shock wave, including:

[0084] Shock wave arrival time:

[0085] Peak value: p k , where k is the peak time;

[0086] Specific impulse: pressure curve p t In the time zone The area;

[0087] Duration:

[0088] This invention automates the identification of the peak value and barometric pressure range of shock waves, automatically corrects zero-line drift and abnormal shock wave regions, and finally automatically extracts characteristic parameters, overcoming the reliance on traditional manual processing and greatly improving the processing efficiency of shock wave data.

[0089] This invention was used to automate batch processing of six shock wave datasets, each containing 500 shock wave data points, with sample sizes ranging from 63,024 to 527,860. Table 1 shows the results of a specific implementation example. As can be seen from Table 1, the data processing time is less than one minute. In contrast, traditional manual processing of a single shock wave data point takes at least three minutes, and processing 500 data points would take several days. Therefore, this invention has a significant advantage over traditional manual shock wave data processing.

[0090] Table 1. Results of automated processing of shock wave data.

[0091] Data Number Number of shock waves sampling volume Data processing time (s) 1 500 63024 32.66 2 500 85775 40.73 3 500 100410 41.12 4 500 113776 31.53 5 500 459434 49.79 6 500 527860 37.2

Claims

1. A method for automatically processing shock wave pressure signals, characterized in that, Specifically, the following steps are included: Step 1: Read the shock wave pressure data. The shock wave pressure data includes time data and its corresponding shock wave pressure data. There are n different times, t1, t2, ..., tn. n The corresponding shock wave pressure data are p1, p2, ..., p n ; Step 2: Locate local maxima in the shock wave pressure data; Step 3: Calculate the characteristic quantities of the local maxima region; Step 4: Determine the peak value of the shock wave based on the relevant characteristic quantities; Step 5: Based on the peak value of the shock wave, perform zero-line drift processing to obtain the shock wave pressure data after zero-line drift processing; Step 6: Based on the shock wave pressure data obtained in Step 5 and the time corresponding to the peak value of the shock wave, estimate the arrival time and end time of the shock wave; based on the obtained arrival time and end time of the shock wave, obtain the corresponding shock wave pressure data; Step 7: Perform waveform anomaly processing on the shock wave pressure data obtained in Step 6 to obtain the processed shock wave pressure data; Step 8: Based on the shock wave pressure data processed in Step 7, automatically extract the characteristic parameters of the shock wave.

2. The method for automatically processing shock wave pressure signals as described in claim 1, characterized in that, Step 2 specifically includes the following sub-steps: Step 2-1: Traverse the shock wave pressure data to find all local maximum pressure values ​​of the shock wave; specifically, a local maximum pressure value p i It refers to p i It is greater than the pressure data of the two directly adjacent shock waves, i.e., p i-1 <p i <p i+1 ; Step 2-2: Given a height threshold Th1, filter out local maximum pressure values ​​that are less than this threshold; Step 2-3: If the distance between two local maximum pressure values ​​is less than the given threshold Th2, filter out the smaller of the two local maximum pressure values; the distance between two local maximum pressure values ​​refers to the time difference between the two values.

3. The method for automatically processing shock wave pressure signals as described in claim 2, characterized in that, In step 3, assume that step 2 found m local maxima: Calculate the relevant features within each local maximum region, including: (1) Local maxima: (2) α% width: Given a percentage α%, calculate the threshold. For the k-th local maximum From its corresponding time Find the moment corresponding to the first pressure value less than Th3 to the left. From its corresponding time Search to the right for the moment corresponding to the first pressure value less than Th3. α% width (3) Ascent time: corresponding time The time corresponding to the first pressure value less than Th3 on its left The time difference between them; (4) Rate of ascent: from time t: The corresponding pressure value rises to The average speed; (5) Downward trend: From time point At the time The correlation coefficient between pressure value data and time.

4. The method for automatically processing shock wave pressure signals as described in claim 3, characterized in that, Step 4 specifically includes the following sub-steps: Step 4-1: Assign scores to the m local maxima regions using the relevant feature quantities; (1) Assign scores to the m local maxima according to the local maxima; (2) Assign scores to the m local maxima according to their width; (3) Assign scores to the m local maxima according to their rise time; (4) Assign scores to the m local maxima according to their rate of ascent; (5) Assign scores to the m local maxima according to their descending trend; Step 4-2: Calculate the weighted average score for each local maximum; Specifically, let m be the number of local maxima and d be the number of relevant features. kj The score w is the score obtained for the j-th eigenvalue of the k-th local maximum. j Let the weights corresponding to j features satisfy the following condition: The weights are set by the user, and the default value is [value missing]. The weighted average score of the k-th local maximum is then calculated as follows: Step 4-3: Calculate the probability of the shock wave peak based on the weighted average score of each local maximum, and determine the shock wave peak; Specifically, the probability of the shock wave peak is calculated using the softmax function as follows: The local maximum value with the highest probability of a shock wave peak is the shock wave peak value.

5. The method for automatically processing shock wave pressure signals as described in claim 4, characterized in that, In step 6, let t be the time corresponding to the peak value of the shock wave after the zero-line drift treatment in step 5. k , will be from t k Search for the first satisfaction on the left and The moment As an estimate of the arrival time of the shock wave; from t k Search for the first satisfaction on the right and The moment As an estimate of the end time of the shock wave; time interval This is called the barotropic region of the shock wave.

6. The method for automatically processing shock wave pressure signals as described in claim 4, characterized in that, Step 6, step 7 includes the following sub-steps: Step 7-1: Read the shock wave pressure data from the peak time to the end time from the processed shock wave data obtained in Step 6; Step 7-2: Fit the read shock wave pressure data using an exponential function to obtain the fitted data; Step 7-3: Detect abnormal areas; Step 7-4: Refit the exponential function to obtain new fitted values; Step 7-5: Re-detect and correct abnormal areas; if abnormal areas exist, replace the corresponding shock wave pressure value with the fitted value obtained in step 7-4 within these areas; if no abnormal areas exist, the shock wave pressure value remains unchanged.

Images