Pipeline sand flow monitoring method and system based on ultrasonic signal layered optimization

Abstract

The present invention relates to the technical field of oil and gas production, and specifically to a pipeline sand flow monitoring method and system based on ultrasonic signal layered optimization, wherein the pipeline sand flow monitoring method mainly comprises: firstly, using a piezoelectric ultrasonic sensor to collect ultrasonic signals produced by gravel hitting the pipe wall in the pipeline, and performing adaptive nonlinear smoothing filtering on the ultrasonic signals, and then performing noise suppression based on frequency domain sparsity analysis, and finally constructing a mathematical model of the relationship between the sampling signal and the sand flow, and using this to monitor the sand flow. The present invention can solve the problem of signal susceptibility to interference and confusion in traditional ultrasonic monitoring methods to the greatest extent, and realize high-performance real-time measurement of sand flow in oil and gas pipelines and calculation of cumulative sand output. When the real-time sand flow or cumulative sand output exceeds the warning value, an alarm will be issued to indicate possible serious sand events, providing data support for relevant decisions.

Description

Technical Field

[0001] The invention relates to the technical field of oil and gas exploitation, and in particular to a pipeline sand flow monitoring method and system based on ultrasonic signal layered optimization. Background Art

[0002] As oil and gas fields are developed, sand particles in the underground rock formations may be carried into the wellbore and pipelines. This phenomenon is called sand production. Sand production has become an important factor affecting oil and gas well production and pipeline safety. Sand production not only reduces oil well production, but may also cause pipeline wear, blockage and equipment failure. In severe cases, it may even cause environmental pollution and safety accidents.

[0003] Traditional sand production monitoring methods mostly rely on physical filtration and regular inspections, which are not only costly but also have poor real-time performance. In recent years, ultrasonic technology has gradually been used in the field of industrial detection due to its advantages of non-contact and real-time monitoring. Therefore, an ultrasonic monitoring method based on acoustic emission sensors is proposed for sand production monitoring in oil and gas wells. This method is based on the principle of piezoelectric effect. The piezoelectric ultrasonic sensor captures the ultrasonic signal generated by the collision or friction of sand and gravel on the pipe wall and converts it into an electrical signal. By analyzing the characteristics of the ultrasonic signal, important information such as sand intensity and flow rate can be inferred.

[0004] However, for the ultrasonic monitoring method, the collected ultrasonic signal is easily interfered by the spike signal caused by external impact or mechanical vibration, resulting in false alarm of sand output; it is also easily affected by the large amount of background noise in the oil and gas pipeline system, resulting in interference and signal confusion. The existing digital signal processing methods such as median filtering and wavelet transform cannot effectively solve these problems; and because the propagation of ultrasonic signals in the pipeline will be affected by the material and structure, the signal will attenuate as the propagation distance increases, so the location of the sensor installation needs to be strictly selected. At the same time, because the ultrasonic frequency is high, there are high requirements for the data transmission rate during data transmission. Summary of the invention

[0005] Based on the above problems, the present invention provides a pipeline sand flow monitoring method and system based on ultrasonic signal layered optimization, which aims to solve the problem of false alarm of sand production and signal interference confusion caused by peak signal interference and background noise in the received signal when using a piezoelectric ultrasonic sensor to perform ultrasonic monitoring of the pipeline sand flow.

[0006] The technical solution adopted by the present invention is: a pipeline sand flow monitoring method based on ultrasonic signal layered optimization, comprising the following steps:

[0007] Step 1: Use a piezoelectric ultrasonic sensor to collect ultrasonic signals at n discrete sampling points produced by gravel hitting the pipe wall in the pipeline, and obtain sampling signals of corresponding amplitudes at n discrete sampling points. ,in .

[0008] Step 2: Sample the signal Perform adaptive nonlinear smoothing filtering to obtain the filtered sampling signal , specifically including the following steps:

[0009] Step 2.1: Calculate the sampled signal The previous point And the latter The average , as shown below:

[0010] ;

[0011] Step 2.2: Calculate the sampled signal With the average The difference , as shown below:

[0012] ;

[0013] Step 2.3: Use the adaptive smoothing formula to sample the signal Replace the outliers in the filter to obtain the filtered sampling signal , specifically including the following steps:

[0014] Step 2.3.1: Set two sliding windows of different sizes, where the size of the first sliding window is , the sliding step length is , the total number of sampling points within the first sliding window is , the size of the second sliding window is , the sliding step length is , the total number of sampling points within the second sliding window ,and > > , > ;

[0015] Step 2.3.2: Calculate the replacement threshold , as shown below:

[0016] ;

[0017] in: The sampling signal of N sampling points within the second sliding window range The average value of The sampling signal The standard deviation within the second sliding window is calculated as follows:

[0018] ;

[0019] in: express The sampling signal value of the amplitude corresponding to the zth discrete sampling point in the past;

[0020] Step 2.3.3: Design a replacement function based on the Sigmoid function , as shown below:

[0021] ;

[0022] Among them: Parameter Deciding on a replacement function The overall smoothing strength of the signal is given by the following formula:

[0023] ;

[0024] In the above formula: is the Reynolds number of the fluid, reflecting the dynamic characteristics of the fluid, is the density of the fluid in the pipe, is the flow rate of the fluid in the pipe, is the pipe diameter, is the fluid viscosity;

[0025] parameter Deciding on a replacement function The overall sensitivity to the signal is given by:

[0026] ;

[0027] in: It is a dynamic reference standard deviation, which is calculated based on the first sliding window over a longer period of time and reflects the overall fluctuation of the signal over a long period of time. It is calculated by the following formula:

[0028] ;

[0029] in: is the sampling signal within the first sliding window The mean of

[0030] Step 2.3.4: When > When using replace , otherwise unchanged.

[0031] Step 3: Sample the filtered signal Perform a noise suppression procedure based on frequency domain sparsity analysis to obtain an enhanced sampled signal , specifically including the following steps:

[0032] Step 3.1: Transform the filtered sampled signal through discrete Fourier transform Convert to frequency domain signal ,in: Represents the sampled signal after filtering at the i-th discrete sampling point The corresponding frequency domain signal The frequency in ;

[0033] Step 3.2: Calculate the frequency domain that causes the piezoelectric ultrasonic sensor response In its neighborhood arrive Sparsity coefficient in the frequency range , where the sparsity coefficient The frequency band close to 1 represents the target signal, and the sparsity coefficient The frequency band close to 0 indicates noise, and the sparsity coefficient It is obtained by the following formula:

[0034] ;

[0035] in: is the frequency window width, which is 500Hz here, considering the frequency characteristics of the target signal and the noise signal; Indicates from arrive All frequencies within the range;

[0036] Step 3.3: Obtain the sparsity coefficient of the full frequency domain, and define the adaptive sparsity enhancement function according to the maximum value of the sparsity coefficient , obtained by the following formula:

[0037] ;

[0038] in: and is the sparsity coefficient in the full frequency domain The maximum and minimum values ​​in ;

[0039] Step 3.4: Use adaptive sparsity enhancement function For frequency domain signals Perform enhancement processing to obtain the enhanced frequency domain signal , obtained by the following formula:

[0040] ;

[0041] Step 3.5: Enhance the processed frequency domain signal Through the inverse discrete Fourier transform, it is restored to the enhanced sampling signal in the time domain .

[0042] Step 4: Use linear interpolation to convert the discrete sampled signal Convert to a continuous sampling signal , as shown below:

[0043] ;

[0044] Where: T is the sampling interval, which is 2×10 -6 s.

[0045] Step 5: Construct the enhanced continuous sampling signal Sand flow rate The mathematical relationship model between them is shown in the following formula:

[0046] ;

[0047] in: is the flow rate of the fluid in the pipeline; is the speed of sound wave propagation in the fluid; is the fluid viscosity; is the pipe diameter; is the fluid noise; and is the compensation parameter.

[0048] Step 6: Measure the sand flow rate Monitor and determine whether it is greater than the set sand flow threshold;

[0049] When the sand flow When the sand flow rate is greater than the sand flow rate threshold, an alarm is issued. When the flow rate is less than the sand discharge threshold, return to step 1.

[0050] In addition, the present invention also provides a pipeline sand flow monitoring system using the above-mentioned pipeline sand flow monitoring method, including: a piezoelectric ultrasonic sensor, a sensor fixing module, a sensor power supply module, a data acquisition module, a Gigabit Ethernet module and a host computer; wherein the piezoelectric ultrasonic sensor is used to collect ultrasonic waves generated by the impact of sand in the pipeline on the inner wall of the pipeline and generate electrical signals; the sensor fixing module is used to fix the piezoelectric ultrasonic sensor on the pipeline to be monitored; the sensor power supply module is electrically connected to the piezoelectric ultrasonic sensor through a BNC coaxial cable, and is used to power the piezoelectric ultrasonic sensor; the data acquisition module is used to collect the electrical signals generated by the piezoelectric ultrasonic sensor in real time; the Gigabit Ethernet module is connected to the host computer through an RJ45 network communication cable, and is used to send the electrical signals collected by the data acquisition module to the host computer; the host computer is used to receive the electrical signals, and process and calculate the received electrical signals according to the above-mentioned monitoring method to obtain the sand flow rate, and at the same time, an alarm is issued when the sand flow rate exceeds the warning value.

[0051] The beneficial effects of the present invention are:

[0052] 1. The present invention uses the signals collected by the piezoelectric ultrasonic sensor to provide a pipeline sand flow monitoring method based on ultrasonic signal layered optimization. This method can solve the problem of signal susceptibility to interference and confusion in the traditional ultrasonic monitoring method to the greatest extent, and realize high-performance real-time measurement of the sand flow of the oil and gas pipeline and calculation of the cumulative sand production. When the real-time sand flow or the cumulative sand production exceeds the warning value, an alarm will be issued to indicate possible serious sand production events, providing data support for relevant decisions.

[0053] 2. The monitoring system provided by the present invention based on the above-mentioned sand flow monitoring method can continuously measure the sand flow of oil and gas pipelines for a long time without regular inspection and cleaning, is applicable to a variety of fluid environments, and is not easily affected by external interference; at the same time, the optimal installation position of the sensor is determined through preliminary simple experiments to ensure the sensor collection effect; in addition, the system uses Gigabit Ethernet for data transmission to ensure the real-time nature of data transmission. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 The present invention is a flow chart of a pipeline sand flow monitoring method based on ultrasonic signal layered optimization. DETAILED DESCRIPTION

[0055] The technical solution adopted by the present invention will be described clearly and completely below.

[0056] It should be noted that the embodiments described below are only part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without making any creative work are within the scope of protection of the present invention.

[0057] First, the present invention provides a pipeline sand flow monitoring method based on ultrasonic signal layered optimization, such as Figure 1 As shown, the following steps are included:

[0058] Step 1: Use a piezoelectric ultrasonic sensor to collect ultrasonic signals at n discrete sampling points produced by gravel hitting the pipe wall in the pipeline, and obtain sampling signals of corresponding amplitudes at n discrete sampling points. ,in ;

[0059] Specifically, a piezoelectric ultrasonic sensor is used to collect ultrasonic signals of n discrete sampling points produced by gravel in the pipeline hitting the inner wall of the pipeline at a time t at the bend on the outer surface of the pipeline, and output it as sampling signals of amplitudes corresponding to the n discrete sampling points. ;

[0060] At the same time, in order to avoid excessive high-frequency noise, such as the operating noise of mechanical equipment such as pumps and compressors, pipeline structure resonance, and electromagnetic noise generated by instruments and equipment, which may affect the sampling signal In this embodiment, when setting the sampling rate of the data acquisition card of the piezoelectric ultrasonic sensor, it is necessary to obtain a complete sampling signal at the highest possible sampling rate. , and at the same time, the sampling rate should not be too high to avoid acquiring more high-frequency noise;

[0061] In this embodiment, the sampling rate of the data acquisition card is set to 500ksps, that is, the signal acquisition card collects 50,000 points at a sampling rate of 500K per second and sends data once;

[0062] In addition, for pipes with varying diameters, the piezoelectric ultrasonic sensor is preferably arranged at a bend on the outer surface of the pipe at a position 1.75 times the pipe diameter to ensure that the ultrasonic signal generated by the collision can be captured by the piezoelectric ultrasonic sensor to the greatest extent.

[0063] Step 2: Sample the signal Perform adaptive nonlinear smoothing filtering to obtain the filtered sampling signal By using nonlinear smoothing, the impact of external foreign objects and the transient impact caused by mechanical equipment such as pipeline valve opening and sudden changes in the working state of the pump on the sampling signal are eliminated. The peak impact interference caused specifically includes the following steps:

[0064] Step 2.1: Calculate the sampled signal The previous point And the latter The average , as shown below:

[0065] (1);

[0066] Step 2.2: Calculate the sampled signal With the average The difference , as shown below:

[0067] (2);

[0068] Step 2.3: Use the adaptive smoothing formula to sample the signal Replace the outliers in the filter to obtain the filtered sampling signal , specifically including the following steps:

[0069] Step 2.3.1: Set two sliding windows of different sizes, where the size of the first sliding window is , the sliding step length is , the total number of sampling points within the first sliding window is , the size of the second sliding window is , the sliding step length is , the total number of sampling points within the second sliding window ,and > > , > ;

[0070] In this embodiment, the size of the first sliding window , total number of sampling points , sliding step length The size of the second sliding window is 100 sampling points. , sliding step length There are 10 sampling points, and the total number of sampling points is ;

[0071] Using sliding windows to divide time series data can increase the weight of the data cycle time series. By focusing on sampling points that are close in time, it can reduce noise interference and enhance signal characteristics, thereby improving the accuracy and efficiency of analysis. At the same time, sliding windows can realize real-time data updates, reduce memory requirements and improve system response speed.

[0072] Step 2.3.2: Calculate the replacement threshold , as shown below:

[0073] (3);

[0074] in: Within the second sliding window Sample signal at sampling points The average value of The sampling signal The standard deviation within the second sliding window is calculated as follows:

[0075] (4);

[0076] in: express The sampling signal value of the amplitude corresponding to the zth discrete sampling point in the past;

[0077] Since the impulse interference signal roughly obeys the Gaussian distribution, in this embodiment, the average value of the sampled signal within the second sliding window of a smaller time is used. Three times the standard deviation The sum of , can cover 99.7% of normal data points. Most normal data points will fall within this range, and points outside this range are outliers.

[0078] Step 2.3.3: Design a replacement function based on the Sigmoid function , as shown below:

[0079] (5);

[0080] Among them: Parameter Deciding on a replacement function The overall smoothing strength of the signal is given by the following formula:

[0081] (6);

[0082] In the above formula: is the Reynolds number of the fluid, reflecting the dynamic characteristics of the fluid, is the density of the fluid in the pipe, is the flow rate of the fluid in the pipe, is the pipe diameter, is the fluid viscosity;

[0083] parameter Deciding on a replacement function The overall sensitivity to the signal is given by:

[0084] (7);

[0085] in: It is a dynamic reference standard deviation, which is calculated based on the first sliding window over a longer period of time and reflects the overall fluctuation of the signal over a long period of time. It is calculated by the following formula:

[0086] (8);

[0087] in: is the sampling signal within the first sliding window The mean of

[0088] parameter It is dynamically calculated by the ratio of two standard deviations. In this way, It can adaptively reflect the amplitude change of the current signal relative to a longer period of time, thereby adjusting the smoothing process according to the characteristics of the signal;

[0089] Step 2.3.4: When > When using replace , otherwise unchanged.

[0090] Step 3: Sample the filtered signal Perform a noise suppression procedure based on frequency domain sparsity analysis to obtain an enhanced sampled signal , specifically including the following steps:

[0091] Step 3.1: Transform the filtered sampled signal through discrete Fourier transform Convert to frequency domain signal ,in: Represents the sampled signal after filtering at the i-th discrete sampling point The corresponding frequency domain signal The frequency in ;

[0092] Step 3.2: Calculate the frequency domain that causes the piezoelectric ultrasonic sensor response In its neighborhood arrive Sparsity coefficient in the frequency range , where the sparsity coefficient The frequency band close to 1 represents the target signal, and the sparsity coefficient The frequency band close to 0 indicates noise, and the sparsity coefficient It is obtained by the following formula:

[0093] (9);

[0094] in: is the frequency window width, which is 500Hz here, considering the frequency characteristics of the target signal and the noise signal; Indicates from arrive All frequencies within the range;

[0095] Step 3.3: Obtain the sparsity coefficient of the full frequency domain, and define the adaptive sparsity enhancement function according to the maximum value of the sparsity coefficient , obtained by the following formula:

[0096] (10);

[0097] in: and is the sparsity coefficient in the full frequency domain The maximum and minimum values ​​in ;

[0098] Step 3.4: Use adaptive sparsity enhancement function For frequency domain signals Perform enhancement processing to obtain the enhanced frequency domain signal , obtained by the following formula:

[0099] (11);

[0100] Specifically, the enhancement function Based on the sparsity coefficient Definition, used to filter the sampled signal To enhance, is the local sparsity coefficient, and are the maximum and minimum values ​​of the sparsity coefficient in the frequency domain of the piezoelectric ultrasonic sensor response. Through this relationship, in formula (10), the sparsity coefficient The higher the frequency component (target signal), the corresponding enhancement function The larger the value, the signal of this frequency can be enhanced by formula (11). Similarly, the sparsity coefficient Lower frequency components (noise) will be suppressed.

[0101] Step 3.5: Enhance the processed frequency domain signal Through the inverse discrete Fourier transform, it is restored to the enhanced sampling signal in the time domain .

[0102] Step 4: Use linear interpolation to convert the discrete sampled signal Convert to a continuous sampling signal , as shown below:

[0103] (12);

[0104] Where: T is the sampling interval, which is 2×10 -6 s.

[0105] Step 5: Construct the enhanced continuous sampling signal Sand flow rate The mathematical model of the relationship between is derived as follows:

[0106] According to the basic relationship between the velocity of the sand and the energy E of the ultrasonic wave generated by the sand colliding with the pipe wall, that is, when the sand in the pipe moves in the fluid and hits the pipe wall, the energy E of the ultrasonic wave generated is proportional to the kinetic energy of the sand:

[0107] (13);

[0108] Where: m is the mass of the gravel; v is the velocity of the gravel when it hits;

[0109] The velocity v of the gravel in the fluid is usually the same as the flow rate of the fluid in the pipe. is relevant, so set:

[0110] (14);

[0111] Where: C is a coefficient related to the fluid flow state. According to formulas (13)-(14), it can be obtained that:

[0112] (15);

[0113] Assume that at time t, a total of n collision events occurred, and the total energy of n collision events is is i ( ) The energy generated by the collision event The sum of:

[0114] (16);

[0115] in: is the mass of gravel at the i-th collision;

[0116] Sand flow rate in fluid Total mass of particles Assume that the total mass of particles that collide within time t is ,in is the number of particles, is the particle mass, then the sand flow rate for:

[0117] (17);

[0118] According to formulas (16)-(17), we can get:

[0119] (18);

[0120] Then, consider the following fluid factors and the impact of background noise:

[0121] Speed ​​of sound : reflects the speed of sound wave propagation in the fluid and affects the propagation efficiency of the acoustic emission signal;

[0122] Fluid Viscosity : reflects the resistance of the fluid to the movement of particles, fluid viscosity Affects the velocity attenuation of particles, thus affecting the intensity of the acoustic emission signal;

[0123] Pipe diameter : The larger the cross-sectional area of ​​the pipe, the lower the particle distribution density, resulting in a lower collision frequency per unit time;

[0124] Fluid noise :The fluid noise in the pipeline will cause the measurement signal to be superimposed and amplified;

[0125] Taking into account the impact of the above parameters, we can get:

[0126] (19);

[0127] In addition, considering the influence of factors such as sensor sensitivity, pipe material, and gravel that cannot collide with the pipe wall to generate signals, compensation parameters are introduced. and , and the total energy Converted into the energy accumulation caused by multiple collisions to obtain the sand flow rate The calculation formula is:

[0128] (20);

[0129] Since the output intensity of the signal collected by the piezoelectric ultrasonic sensor is proportional to the energy of the signal, that is, the greater the energy of the ultrasonic wave, the greater the amplitude of the corresponding sampling signal, so in formula (20):

[0130] (twenty one);

[0131] According to formula (20)-(21), the final sand flow rate is obtained With the sampling signal The mathematical relationship model between:

[0132] (twenty two);

[0133] Furthermore, we can integrate formula (22) to obtain Time has come The cumulative sand flow rate during the time , as shown below:

[0134] (twenty three);

[0135] Step 6: Set the sand flow rate threshold according to the sand flow rate. With the sampling signal The mathematical relationship model between them is used to determine the sand flow rate. Is it greater than the sand flow threshold? When the sand flow rate is greater than the sand flow rate threshold, an alarm is issued. When it is less than the sand flow rate threshold, return to step 1;

[0136] Specifically, when it is necessary to monitor the sand flow rate at a certain time t, set the sand flow rate threshold According to formula (22), the sand discharge rate corresponding to time t can be obtained , when the sand flow rate Sand flow threshold when greater than When the sand flow rate at time t is too large, an alarm is sounded, indicating that the sand flow rate at time t is too large, which may cause abnormal problems in the pipeline;

[0137] When you need to monitor Time has come When the cumulative sand flow rate is within a certain time, set the cumulative sand flow rate threshold According to formula (23), we can get Time has come The cumulative sand flow rate during the time , when the cumulative sand output Greater than the cumulative sand flow threshold When the sand output is too large, an alarm is sounded, indicating that the accumulated sand output is too large, which may cause abnormal problems in the pipeline;

[0138] After the abnormal problem is handled, the current alarm can be cleared by resetting it and returning to step 1 to restart monitoring.

[0139] Secondly, according to the above monitoring method, the present invention also provides a monitoring system for pipeline sand flow based on ultrasonic signal layered optimization, including: a piezoelectric ultrasonic sensor, a sensor fixing module, a sensor power supply module, a data acquisition module, a Gigabit Ethernet module, and a host computer;

[0140] Among them, the piezoelectric ultrasonic sensor is used to collect the ultrasonic waves generated by the sand and gravel in the pipeline hitting the inner wall of the pipeline and generate electrical signals;

[0141] The sensor fixing module is used to fix the piezoelectric ultrasonic sensor on the pipeline to be monitored. The main functional component of the sensor fixing module is preferably a strong magnetic adsorption bracket, which can adsorb and fix the sensor on the outer surface of the metal pipeline to collect the acoustic emission signal generated by the impact of gravel on the pipe wall. At the same time, in order to ensure the signal quality, preferably, an ultrasonic coupling agent can be filled between the piezoelectric ultrasonic sensor and the pipeline surface;

[0142] The sensor power supply module is electrically connected to the piezoelectric ultrasonic sensor through a BNC coaxial cable, and is used to supply power to the piezoelectric ultrasonic sensor;

[0143] The data acquisition module is used to collect the electrical signal generated by the piezoelectric ultrasonic sensor in real time;

[0144] The Gigabit Ethernet module is connected to the host computer via an RJ45 network communication cable and is used to send the electrical signals collected by the data acquisition module to the host computer;

[0145] The host computer is used to receive the electrical signal, and process and calculate the received electrical signal according to the above monitoring method to obtain the sand flow rate. At the same time, an alarm is issued when the sand flow rate or the cumulative sand output exceeds the warning value.

Claims

1. A pipeline sand flow monitoring method based on ultrasonic signal layered optimization, characterized in that: The following steps are involved: Step 1: Use a piezoelectric ultrasonic sensor to collect ultrasonic signals at n discrete sampling points produced by gravel hitting the pipe wall in the pipeline, and obtain sampling signals of corresponding amplitudes at n discrete sampling points. ,in ; Step 2: Sample the signal Perform adaptive nonlinear smoothing filtering to obtain the filtered sampling signal ; Step 3: Sample the filtered signal Perform a noise suppression procedure based on frequency domain sparsity analysis to obtain an enhanced sampled signal ; Step 4: Use linear interpolation to convert the discrete sampled signal Convert to a continuous sampling signal ; Step 5: Construct a continuous sampling signal Sand flow rate The mathematical relationship model between them is shown in the following formula: （1） In formula (1): is the flow rate of the fluid in the pipe, is the speed of sound wave propagation in the fluid, is the fluid viscosity, is the pipe diameter, is the fluid noise, and is the compensation parameter; Step 6: Set the sand flow threshold according to the continuous sampling signal Sand flow rate The mathematical relationship model between Monitor the sand flow rate. When the sand flow rate is greater than the sand flow rate threshold, an alarm is issued. When the flow rate is less than the sand discharge threshold, return to step 1 and restart monitoring.

2. The pipeline sand flow monitoring method based on ultrasonic signal layered optimization according to claim 1 is characterized in that: In the step 1, the sampling rate of the signal acquisition card of the piezoelectric ultrasonic sensor is set to 500ksps, and the signal acquisition card collects 50,000 points at a sampling rate of 500K and sends data once.

3. The pipeline sand flow monitoring method based on ultrasonic signal layered optimization according to claim 1 is characterized in that: The step 2 specifically includes: Step 2.1: Calculate the sampled signal The previous point And the latter The average , as shown below: （2） Step 2.2: Calculate the sampled signal With the average The difference , as shown below: （3） Step 2.3: Use the adaptive smoothing formula to sample the signal Replace the outliers in the filter to obtain the filtered sampling signal .

4. The pipeline sand flow monitoring method based on ultrasonic signal layered optimization according to claim 3 is characterized in that: The step 2.3 specifically includes: Step 2.3.1: Set two sliding windows of different sizes, where the size of the first sliding window is , the sliding step length is , the total number of sampling points within the first sliding window is , the size of the second sliding window is , the sliding step length is , the total number of sampling points within the second sliding window ,and > > , > ; Step 2.3.2: Calculate the replacement threshold , as shown below: （4） In formula (4): The sampling signal of N sampling points within the second sliding window range The average value of The sampling signal The standard deviation within the second sliding window is calculated as follows: （5） In formula (5): express The sampling signal value of the amplitude corresponding to the zth discrete sampling point in the past; Step 2.3.3: Design a replacement function based on the Sigmoid function , as shown below: （6） In formula (6): Parameter Deciding on a replacement function The overall smoothing strength of the signal is given by the following formula: （7） In formula (7): is the Reynolds number of the fluid, reflecting the dynamic characteristics of the fluid, is the density of the fluid in the pipe, is the flow rate of the fluid in the pipe, is the pipe diameter, is the fluid viscosity; parameter Deciding on a replacement function The overall sensitivity to the signal is given by: （8） In formula (8): It is a dynamic reference standard deviation, which is calculated based on the first sliding window over a longer period of time and reflects the overall fluctuation of the signal over a long period of time. It is calculated by the following formula: （9） In formula (9): is the sampling signal within the first sliding window The mean of Step 2.3.4: When > When using replace , otherwise unchanged.

5. The pipeline sand flow monitoring method based on ultrasonic signal layered optimization according to claim 4 is characterized in that: In step 2.3.1, the size of the first sliding window is , total number of sampling points , sliding step length The size of the second sliding window is 100 sampling points. , sliding step length There are 10 sampling points, and the total number of sampling points is .

6. The pipeline sand flow monitoring method based on ultrasonic signal layered optimization according to claim 1 is characterized in that: The step 3 specifically includes: Step 3.1: Transform the filtered sampled signal through discrete Fourier transform Convert to frequency domain signal ,in: Represents the sampled signal after filtering at the i-th discrete sampling point The corresponding frequency domain signal The frequency in ; Step 3.2: Calculate the frequency domain that causes the piezoelectric ultrasonic sensor response In its neighborhood arrive Sparsity coefficient in the frequency range , where the sparsity coefficient The frequency band close to 1 represents the target signal, and the sparsity coefficient The frequency band close to 0 indicates noise, and the sparsity coefficient It is obtained by the following formula: （10） In formula (10): is the frequency window width, which is 500Hz here, considering the frequency characteristics of the target signal and the noise signal; Indicates from arrive All frequencies within the range; Step 3.3: Obtain the sparsity coefficient of the full frequency domain, and define the adaptive sparsity enhancement function according to the maximum value of the sparsity coefficient , obtained by the following formula: （11） In formula (11): and is the sparsity coefficient in the full frequency domain The maximum and minimum values ​​in ; Step 3.4: Use adaptive sparsity enhancement function For frequency domain signals Perform enhancement processing to obtain the enhanced frequency domain signal , obtained by the following formula: （12） Step 3.5: Enhance the processed frequency domain signal Through the inverse discrete Fourier transform, it is restored to the enhanced sampling signal in the time domain .

7. The pipeline sand flow monitoring method based on ultrasonic signal layered optimization according to claim 1 is characterized in that: The discrete sampling signal in step 4 With continuous sampling signal Satisfy the following formula: （13） In formula (13), T is the sampling interval, which is 2×10 -6 s.

8. The pipeline sand flow monitoring method based on ultrasonic signal layered optimization according to claim 1 is characterized in that: Calculated by the following formula arrive Cumulative sand output within a certain period of time : （14）。 9. The pipeline sand flow monitoring method based on ultrasonic signal layered optimization according to claim 8 is characterized in that: set up arrive The cumulative sand flow threshold within a certain time period. When the accumulated sand flow rate is greater than the threshold, an alarm is sounded. When it is less than the cumulative sand flow rate threshold, return to step 1 and restart monitoring.

10. A pipeline sand flow monitoring system using the pipeline sand flow monitoring method based on ultrasonic signal layered optimization according to claim 1, characterized in that: include: Piezoelectric ultrasonic sensor, sensor fixing module, sensor power supply module, data acquisition module, Gigabit Ethernet module and host computer; Among them, the piezoelectric ultrasonic sensor is used to collect the ultrasonic waves generated by the sand and gravel in the pipeline hitting the inner wall of the pipeline and generate electrical signals; The sensor fixing module is used to fix the piezoelectric ultrasonic sensor on the pipeline to be monitored; The sensor power supply module is electrically connected to the piezoelectric ultrasonic sensor through a BNC coaxial cable, and is used to supply power to the piezoelectric ultrasonic sensor; The data acquisition module is used to collect the electrical signal generated by the piezoelectric ultrasonic sensor in real time; The Gigabit Ethernet module is connected to the host computer via an RJ45 network communication cable and is used to send the electrical signals collected by the data acquisition module to the host computer; The host computer is used to receive the electrical signal, process and calculate the electrical signal to obtain the sand flow rate, and monitor the sand flow rate at the same time, and alarm when the sand flow rate exceeds the warning value.

Images