Radar system for pipeline inner wall modeling and pipeline inner wall modeling method

Abstract

The invention provides a radar system for pipeline inner wall modeling and a pipeline inner wall modeling method, and relates to the technical field of radar application, the system comprises a millimeter wave radar and a laser radar; the millimeter wave radar is configured to emit electromagnetic waves to the inner wall of a to-be-measured pipeline and receive reflected waves of the electromagnetic waves. Based on the reflected waves, using Doppler effect and echo analysis operation to generate structural data of the inner wall of the to-be-detected pipeline; the laser radar is configured to emit a laser beam to the inner wall of a to-be-detected pipeline and receive a reflection signal of the laser beam; according to the reflection signal, calculating the flight time or phase difference of the laser beam, and generating 3D point cloud data of the inner wall of the to-be-detected pipeline; the modeling module is configured to fuse the modeling data by using an MLS algorithm to obtain modeling fusion data; and on the basis of the modeling fusion data, utilizing a Poisson reconstruction model to generate an inner wall model of the to-be-measured pipeline, so as to solve the problem that the existing single radar system is difficult to perform high-precision accurate modeling on the inner wall of the pipeline.

Description

Technical Field

[0001] This application relates to the technical field of radar applications, and in particular, to a radar system for pipeline inner wall modeling and a pipeline inner wall modeling method. Background Art

[0002] In the fields of energy transportation such as oil and natural gas, pipelines, as key transportation carriers, their safety and reliability are of crucial importance. Due to long-term operation, the inner wall of the pipeline may have defects such as deformation and cracks due to corrosion, wear, sediment accumulation, etc. If these defects are not discovered and processed in a timely manner, they will seriously threaten the safe operation of the pipeline. Therefore, regular inspection of pipelines has become an important link in ensuring the safety of energy transportation. As a device that can operate autonomously inside the pipeline and collect data, the in-pipe detector is widely used in the field of pipeline inspection because it can achieve real-time monitoring of the internal condition of the pipeline.

[0003] In response to pipeline inspection requirements, various radar system solutions have been tried currently. Among them, lidar, with its wide scanning range and high detection accuracy, is partially applied to the field of pipeline inspection. Lidar obtains the three-dimensional point cloud data of the pipeline inner wall by emitting laser beams and receiving the reflected signals, and then realizes the preliminary modeling of the pipeline inner wall. On the other hand, millimeter-wave radar has attracted attention due to its strong penetration ability. It can penetrate oil and gas particles and sediments to a certain extent to detect the inside of the pipeline.

[0004] However, in practical applications, the in-pipe detector is only equipped with one radar system, and both of these radar systems have limitations in the application process. Although lidar has high accuracy, its penetration ability is poor, and it is difficult to work stably in the internal environment of the pipeline full of oil and gas particles and sediments; while millimeter-wave radar, although it has strong penetration ability, its detection accuracy is relatively low, and it is difficult to meet the requirements of high-precision modeling of the pipe wall. Summary of the Invention

[0005] This application provides a radar system for pipeline inner wall modeling and a pipeline inner wall modeling method to solve the technical problem that the existing single radar system is difficult to accurately model the pipeline inner wall with high precision.

[0006] The first aspect of this application provides a radar system for pipeline inner wall modeling, including: A millimeter-wave radar and a lidar disposed on the in-pipe detector; the in-pipe detector is configured to travel from the inlet end of the pipeline to be measured to the outlet end of the pipeline to be measured; The millimeter-wave radar is configured to: Transmit electromagnetic waves to the inner wall of the pipeline to be measured and receive the reflected waves of the electromagnetic waves; Based on the reflected wave, using the Doppler effect and echo analysis operations, generate the structural data of the inner wall of the pipeline to be measured; the structural data includes: the distance between the millimeter-wave radar and the inner wall of the pipeline to be measured, the speed of the in-pipe detector, and the material of the pipeline to be measured; The lidar is configured to: Emit a laser beam onto the inner wall of the pipeline to be measured and receive the reflected signal of the laser beam; According to the reflected signal, calculate the flight time or phase difference of the laser beam to generate 3D point cloud data of the inner wall of the pipeline to be measured; A modeling module, the modeling module is communicatively connected to the millimeter-wave radar and the lidar, and the modeling module is configured to: Based on the modeling data, using multi-source spatio-temporal registration operations, align the modeling data in time and space; the modeling data includes: the structural data and the 3D point cloud data; Using the MLS algorithm, fuse the modeling data to obtain modeling fusion data; Based on the modeling fusion data, use the Poisson reconstruction model to generate the inner wall model of the pipeline to be measured.

[0007] In some embodiments, the lidar includes: A first housing, a receiver is disposed inside the first housing; A second housing, the second housing is communicated with the first housing, and a light-transmitting hole is disposed on the surface of the second housing; a bevel mirror is disposed inside the second housing; A laser source, the laser source is configured to: Emit a laser beam to the bevel mirror, and the bevel mirror reflects the laser beam through the light-transmitting hole onto the inner wall of the pipeline to be measured; the reflected signal from the inner wall of the pipeline to be measured is reflected by the bevel mirror to the receiver; The receiver is configured to: According to the reflected signal, calculate the flight time or phase difference of the laser beam to generate 3D point cloud data of the inner wall of the pipeline to be measured.

[0008] In some embodiments, an optical rotary encoder is disposed between the first housing and the second housing, and the optical rotary encoder is communicated with the first housing and the second housing; the laser source is disposed inside the optical rotary encoder; the optical rotary encoder is configured to rotate the second housing at a first preset angle; A servo motor is further disposed inside the second housing, the servo motor is connected to the bevel mirror, and the servo motor is configured to rotate the bevel mirror at a second preset angle.

[0009] In some embodiments, the number of lidars in the horizontal direction of the in-pipe detector is 1, and the horizontal detection range of the lidar is 360°; the number of millimeter-wave radars in the horizontal direction of the in-pipe detector is 4, and the horizontal detection range of the millimeter-wave radar is 90°.

[0010] In some embodiments, the modeling module is further configured to: Obtain each data point in the 3D point cloud data and the neighborhood points of a preset number of the data points; Calculate the average distance and standard deviation between the data point and the neighborhood points; Calculate a retention distance according to the average distance and the standard deviation; the retention distance is: ; wherein, is the average distance; is the standard deviation; Remove the neighborhood points outside the retention distance among the data points.

[0011] In some embodiments, the modeling module is further configured to: Based on the structure data, establish a sliding window; Obtain the data points in the sliding window for a continuous preset number of frames. If there are target data points that do not appear repeatedly among the data points, remove the target data points.

[0012] In some embodiments, the modeling module is further configured to: Use the encoder pulses to align the sampling frequency of the modeling data by linear interpolation method, so that the modeling data is aligned in time; Obtain the odometer wheel data of the in-pipe detector; Use the odometer wheel data to perform a spatial registration operation on the modeling data, so that the modeling data is aligned in space.

[0013] In some embodiments, the modeling module is further configured to: Define an MLS model; the MLS model is configured as: According to the data points in the modeling data and the neighborhood points of the data points, use the weighted least squares method to obtain the fitting surface of the modeling data; the fitting surface is: ; wherein, is the data point; is the neighborhood point; is a polynomial function; is the weight function; is the coefficient of the polynomial function; is the actual value of the height of each neighborhood point; According to the data points in the modeling data, use a KD tree to obtain the neighborhood points of the data points; According to the data points and the neighborhood points, determine a linear equation; the linear equation is: ; In the formula, is the design matrix, determined by the neighborhood points; is the diagonal weight matrix, determined by the data points and the neighborhood points; Input the data points, neighborhood points, and linear equation into the MLS model to obtain the fitted surface of the modeling data; Project the data points onto the fitted surface to obtain the modeled fusion data.

[0014] In some embodiments, the modeling module is further configured to: Based on the modeled fusion data, use a Poisson reconstruction model to obtain the implicit equation of the modeled fusion data; Perform an isosurface extraction operation on the implicit equation to generate the inner wall model of the pipeline to be measured.

[0015] The second aspect of the present application provides a method for modeling the inner wall of a pipeline, which is applied to a radar system for modeling the inner wall of a pipeline described in any of the above embodiments, including: Based on the modeling data, use multi-source spatio-temporal registration operations to align the modeling data in time and space; the modeling data includes: structural data and 3D point cloud data; the structural data includes: the distance between the millimeter-wave radar and the inner wall of the pipeline to be measured, the speed of the in-pipe detector, and the material of the pipeline to be measured; Use the MLS algorithm to fuse the modeling data to obtain the modeled fusion data; Based on the modeled fusion data, use a Poisson reconstruction model to generate the inner wall model of the pipeline to be measured.

[0016] The present application provides a radar system and a method for inner wall modeling of a pipeline. The system includes: a millimeter-wave radar and a lidar disposed on an in-pipe detector; the in-pipe detector is configured to travel from the inlet end of a pipeline to be measured to the outlet end of the pipeline to be measured; the millimeter-wave radar is configured to: emit electromagnetic waves to the inner wall of the pipeline to be measured and receive the reflected waves of the electromagnetic waves; based on the reflected waves, use the Doppler effect and echo analysis operations to generate structure data of the inner wall of the pipeline to be measured; the structure data includes: the distance between the millimeter-wave radar and the inner wall of the pipeline to be measured, the speed of the in-pipe detector, and the material of the pipeline to be measured; the lidar is configured to: emit laser beams to the inner wall of the pipeline to be measured and receive the reflected signals of the laser beams; calculate the flight time or phase difference of the laser beams according to the reflected signals to generate 3D point cloud data of the inner wall of the pipeline to be measured; a modeling module, the modeling module is communicatively connected to the millimeter-wave radar and the lidar, and the modeling module is configured to: based on modeling data, use multi-source spatio-temporal registration operations to align the modeling data in time and space; the modeling data includes: the structure data and the 3D point cloud data; use the MLS algorithm to fuse the modeling data to obtain modeling fusion data; based on the modeling fusion data, use a Poisson reconstruction model to generate a model of the inner wall of the pipeline to be measured, so as to realize that the in-pipe detector is combined with a millimeter-wave radar and a lidar, by combining the advantages of the two radar systems, and performing multi-point cloud data fusion and high-precision three-dimensional modeling of the inner wall of the pipeline through a data fusion method. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions of the present application, the drawings required for use in the embodiments will be briefly introduced below. Obviously, for those of ordinary skill in the art, other drawings can also be obtained based on these drawings without creative efforts.

[0018] Figure 1 It is a flow chart for inner wall modeling of a pipeline in the present application; Figure 2 It is a schematic structural diagram of an in-pipe detector in the present application; Figure 3 It is a schematic structural diagram of a lidar in the present application; Figure 4 It is a schematic diagram of a detection cross-section of a millimeter-wave radar and a lidar in the present application.

[0019] Description of the reference numerals: 1 - In - pipeline detector; 2 - Millimeter - wave radar; 3 - LiDAR; 31 - First housing; 32 - Receiver; 33 - Second housing; 331 - Light - transmitting hole; 34 - Inclined mirror; 35 - Laser source; 36 - Optical rotary encoder; 37 - Servo motor; 4 - Modeling module. Detailed implementation mode

[0020] In order to enable those skilled in the art of the present technology to better understand the technical solutions in this application, the following will clearly and completely describe the technical solutions in the embodiments of this application with reference to the accompanying drawings in the embodiments of this application. Obviously, the described embodiments are only a part of the embodiments of this application, rather than all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of this application.

[0021] Exemplarily, in - pipeline detection technology refers to the technology of detecting a pipeline by an in - pipeline detector running inside the pipeline and using sensors such as a magnetic flux leakage sensor, an eddy current sensor, and an ultrasonic sensor mounted thereon. Traditional in - pipeline detectors rely on contact sensors (such as magnetic flux leakage sensors and ultrasonic sensors), which are vulnerable to oil stains and rust, and cannot obtain high - precision three - dimensional data of the inner wall of the pipeline. In addition, a single radar sensor (such as LiDAR or millimeter - wave radar) cannot balance penetration and accuracy, resulting in detection blind spots (such as oil - stain - covered areas or complex geometric features).

[0022] The following are the problems existing in the current radar system for modeling the inner wall of the pipeline: 1. Contact sensors are vulnerable to working conditions: Most of the sensors mounted on the current in - pipeline detectors are contact sensors, and their accuracy is vulnerable to the working conditions inside the pipeline. The sensors mounted on the current in - pipeline detectors include a caliper sensor, a magnetic flux leakage sensor, an ultrasonic sensor, etc. Among them, the caliper sensor depends on the contact angle between the caliper probe and the inner wall of the pipeline to judge whether the pipeline is deformed; the magnetic flux leakage sensor needs to use a magnet and a steel brush to contact the pipe wall to magnetize the pipe wall and receive the leaked magnetic field through the magnetic flux leakage sensor. When there are sediments inside the pipeline, the caliper and magnetic flux leakage sensors will not be able to effectively give the inner diameter of the pipeline; although the ultrasonic sensor does not need to directly contact the pipe wall, a medium is required between the sensor and the pipe wall to conduct ultrasonic waves. When there are oil stains and rust inside the pipeline, the detection effect of the ultrasonic sensor will be affected.

[0023] 2. Limited effect of single radar: It is difficult for a single radar system to accurately model the inner wall of a pipeline with high precision. LiDAR has poor penetration, and millimeter-wave radar has low precision. Although LiDAR has a wide scanning range and high detection accuracy, its penetration is poor. At present, the internal conditions of pipelines are complex, usually filled with oil and gas particles, and the pipe walls are usually covered with oil stains, sediments, etc., which will affect the detection effect of LiDAR. Although millimeter-wave radar has strong penetration ability and can penetrate oil and gas particles and sediments, its precision is low and it is difficult to be used alone for high-precision modeling of the pipe wall.

[0024] 3. Difficulties in multi-source data fusion: It is impossible to synergistically utilize the advantages of LiDAR (high precision, wide range) and millimeter-wave radar (strong penetration), thus making it impossible to construct a high-precision complete three-dimensional model of the inner wall of the pipeline.

[0025] Since in some technologies, it is difficult for a single radar system to accurately model the inner wall of a pipeline with high precision, to solve this technical problem, the present application provides a radar system for inner wall modeling of a pipeline and a method for inner wall modeling of a pipeline. The following is an explanation of the radar system for inner wall modeling of a pipeline and the method for inner wall modeling of a pipeline: As Figure 2 shown, it is a schematic structural diagram of the in-pipe detector 1 in the present application.

[0026] The first aspect of the present application provides a radar system for inner wall modeling of a pipeline, including: A millimeter-wave radar 2 and a LiDAR 3 disposed on the in-pipe detector 1; the in-pipe detector 1 is configured to travel from the inlet end of the pipeline to be measured to the outlet end of the pipeline to be measured; during detection, the in-pipe detector 1 is placed in the pipeline to be measured, and pressure is applied to make the in-pipe detector 1 move forward along the pipeline to be measured. During the movement process, the millimeter-wave radar 2 and the LiDAR 3 carried on the in-pipe detector 1 will collect the inner wall information of the pipeline to be measured. The millimeter-wave radar 2 and the LiDAR 3 collect pipeline inner wall data during the operation of the in-pipe detector 1 and are used for inner wall modeling.

[0027] The millimeter-wave radar 2 is configured to: Emit electromagnetic waves to the inner wall of the pipeline to be measured and receive the reflected waves of the electromagnetic waves; based on the reflected waves, use the Doppler effect and echo analysis operations to generate the structure data of the inner wall of the pipeline to be measured; the structure data includes: the distance between the millimeter-wave radar 2 and the inner wall of the pipeline to be measured, the speed of the in-pipe detector 1, and the material of the pipeline to be measured; the millimeter-wave radar emits high-frequency electromagnetic waves of 77 GHz or 60 GHz, and detects the distance, speed, and material characteristics (i.e., structure data) of an object through the Doppler effect and echo analysis. The advantages of the millimeter-wave radar 2 in pipeline applications include: strong penetration ability: it can penetrate oil stains and sediments; it is insensitive to liquids and fog, and is suitable for humid environments. However, its limitation is that the resolution is relatively low (centimeter level), and it needs to be combined with a lidar to supplement details.

[0028] Exemplarily, the structure of the millimeter-wave radar is relatively simple and compact. The emission and reception of electromagnetic waves are realized through an antenna and are usually integrated into an antenna board. The structure is fixed and cannot rotate. Therefore, it is impossible to achieve 360° coverage of the horizontal detection direction, and usually, the maximum coverage can only be 120°. Vertically, it can usually cover 20°.

[0029] The lidar 3 is configured to: Emit a laser beam to the inner wall of the pipeline to be measured and receive the reflected signal of the laser beam; according to the reflected signal, calculate the flight time or phase difference of the laser beam to generate the 3D point cloud data of the inner wall of the pipeline to be measured; the lidar 3 generates high-precision 3D point cloud data by emitting a laser beam and receiving the reflected signal, and calculating the flight time (ToF) or phase difference of the laser beam. The advantages of the lidar 3 in pipeline applications include: sub-millimeter resolution and high precision; wide scanning range, the horizontal FOV can usually cover 360°, and the vertical field of view (FOV) can generally reach 25°. However, its limitation is that it is easily interfered by the scattering of fog and oil stains in the pipeline; it cannot penetrate sediments to detect the pipe wall.

[0030] This application provides a radar system for pipeline inner wall modeling. Based on two types of radars, namely the millimeter-wave radar 2 and the lidar 3, a radar system for high-precision modeling of the pipeline inner wall is constructed. The advantage of the lidar 3 lies in its high precision and wide scanning range, but its penetration ability is poor. The millimeter-wave radar 2 has strong penetration ability and can penetrate oil stains, but its precision is poor and the scanning range is narrow. By combining the advantages of the two radar systems and performing multi-point cloud fusion and high-precision three-dimensional modeling of the pipeline inner wall through a data fusion method, the disadvantages that the sensors carried on the current in-pipe detector are easily affected by working conditions and a single radar is difficult to balance penetration ability and precision are overcome.

[0031] As Figure 3 shown, it is the structural schematic diagram of the lidar 3 in this application.

[0032] In this embodiment, the lidar 3 includes: A first housing 31, in which a receiver 32 is disposed; a second housing 33, which is communicated with the first housing 31, and a light-transmitting hole 331 is disposed on the surface of the second housing 33; a bevel mirror 34 is disposed in the second housing 33; a laser source 35, which is configured to: Emit a laser beam to the bevel mirror 34, and the bevel mirror 34 reflects the laser beam through the light-transmitting hole 331 to the inner wall of the pipeline to be measured; the reflected signal from the inner wall of the pipeline to be measured is reflected by the bevel mirror 34 to the receiver 32.

[0033] The receiver 32 is configured to: Calculate the flight time or phase difference of the laser beam according to the reflected signal, and generate 3D point cloud data of the inner wall of the pipeline to be measured.

[0034] In this embodiment, an optical rotary encoder 36 is disposed between the first housing 31 and the second housing 33, and the optical rotary encoder 36 is communicated with the first housing 31 and the second housing 33; the laser source 35 is disposed in the optical rotary encoder 36; the optical rotary encoder 36 is configured to rotate the second housing 33 at a first preset angle; a servo motor 37 is further disposed in the second housing 33, the servo motor 37 is connected to the bevel mirror 34, and the servo motor 37 is configured to rotate the bevel mirror 34 at a second preset angle. The laser beam is emitted by the laser source 35, reflected by the bevel mirror 34 and then emitted, and after irradiating the inner wall of the pipeline to be measured, the reflected laser beam is reflected by the bevel mirror 34 to the receiver 32. The lidar 3 can rotate freely in the horizontal direction to achieve 360° coverage, and the angle of the bevel mirror 34 is controlled by the servo motor 37 in the vertical direction to achieve vertical coverage. Wherein, the first preset angle A is 360°; the second preset angle B is generally 90°, and the second preset angle B is set based on the complete coverage of the vertical detection direction, which is not limited herein.

[0035] As [[ID=□]] Figure 4 shown, it is a schematic diagram of the detection cross-section of the millimeter-wave radar and the lidar in this application.

[0036] In this embodiment, the number of lidars 3 in the horizontal direction of the in-pipeline detector 1 is 1, and the horizontal detection range of the lidar 3 is 360°; the number of millimeter-wave radars 2 in the horizontal direction of the in-pipeline detector 1 is ¼, and the horizontal detection range of the millimeter-wave radar 2 is 90°.

[0037] It should be noted that there may be an inaccuracy in the description "the number of millimeter-wave radars 2 in the horizontal direction of the in-pipeline detector 1 is ¼" in the original text. It is recommended to check and correct it according to the actual situation. The above translation is based on the existing text.Specifically, generally 1 to 2 lidars 3 are provided in the pipeline internal detector 1; in an axial installation manner, the vertical axis of the lidar 3 is located on the horizontal plane, and at the same time, it is ensured that the axis coincides with the central axis of the pipeline internal detector 1, and the laser emission direction is along the radial direction of the pipeline internal detector 1. At this time, the "horizontal plane" detected by the lidar 3 is the longitudinal circular section of the pipeline. The installation positions are at the front end and / or the end of the pipeline internal detector 1, and the longitudinal section (high-precision 3D point cloud data) of the pipeline to be measured is scanned.

[0038] Specifically, generally 4 to 8 millimeter-wave radars 2 are provided in the pipeline internal detector 1; in a circumferentially uniformly distributed setting manner. In order to detect the longitudinal section of the pipeline to be measured, the placement method of the millimeter-wave radar 2 needs to be similar to that of the lidar 3. However, since the millimeter-wave radar 2 generally cannot cover 360° in the horizontal direction, according to the coverage angle of the millimeter-wave radar 2, multiple millimeter-wave radars 2 are uniformly placed along the circumference of the pipeline internal detector 1, and the electromagnetic wave emission direction is distributed along the radial direction of the pipeline internal detector 1. Each millimeter-wave radar 2 detects a part of the fan-shaped area (90°) of the longitudinal section of the pipeline to be measured, ensuring that the detection results of all millimeter-wave radars 2 cover the complete longitudinal section of the pipeline to be measured. The installation position is in the middle of the pipeline internal detector 1, and the underlying structure data of the pipeline to be measured is obtained by penetrating the oil stain.

[0039] Exemplarily, when 1 lidar 3 is installed at the front end of the pipeline internal detector 1 and 4 millimeter-wave radars 2 are installed in the middle of the pipeline internal detector 1, a schematic diagram of the detection ranges of the millimeter-wave radar 2 and the lidar 3 at a certain moment is as Figure 4 shown. When the lidar 3 is installed at the front end of the pipeline internal detector 1, since the lidar 3 can cover a 360° range, the detection range is the complete circular section of the pipeline to be measured at the corresponding position. A single millimeter-wave radar 2 cannot cover the complete section. Therefore, 4 millimeter-wave radars 2 are evenly arranged along the circumference, and each detects a fan-shaped area within a 90° range, jointly covering the complete section of the pipeline to be measured.

[0040] As Figure 1 shown, it is a flow chart of pipeline inner wall modeling in this application.

[0041] First, preprocessing operations are performed on the modeling data, and the specific operation process is as follows: The modeling module 4 is further configured to: Obtain each data point in the 3D point cloud data and the neighborhood points of a preset number of the data points; calculate the average distance and standard deviation between the data point and the neighborhood points; calculate the retention distance according to the average distance and the standard deviation; the retention distance is: ; In the formula, is the average distance; is the standard deviation; Remove the neighborhood points outside the retention distance among the said data points.

[0042] Specifically, for 3D point cloud data denoising and filtering, statistical outlier rejection is adopted, which specifically includes the following steps: Calculate the average distance and standard deviation between each data point and k neighborhood points (k = 50), and then remove the points whose distance is outside. Through the denoising and filtering operation of the said 3D point cloud data, the computational complexity of the said 3D point cloud data is reduced.

[0043] The modeling module 4 is further configured to: Based on the said structural data, establish a sliding window; Obtain the data points in the sliding window of continuously preset number of frames. If there are target data points that do not appear repeatedly among the data points, then remove the target data points.

[0044] Specifically, for the millimeter-wave radar 2, time-domain consistency check is adopted. First, establish a sliding window (5-frame historical data), and then remove the points that do not appear repeatedly in 3 consecutive frames (transient artifacts) to achieve denoising, thereby reducing the computational complexity of the said structural data.

[0045] Perform multi-source spatio-temporal registration operation on the said modeling data. The specific operation process is as follows: Modeling module 4, the modeling module 4 is communicatively connected to the millimeter-wave radar 2 and the lidar 3. The modeling module 4 is configured to: Based on the modeling data, use multi-source spatio-temporal registration operation to align the said modeling data in time and space; The modeling data includes: the said structural data and the 3D point cloud data.

[0046] The modeling module 4 is further configured to: Use the encoder pulse to align the sampling frequency of the said modeling data by linear interpolation method, so that the said modeling data is aligned in time; Obtain the odometer data of the in-pipe detector 1; Use the odometer data to perform spatial registration operation on the said modeling data, so that the said modeling data is aligned in space.

[0047] Specifically, the multi-source spatio-temporal registration operation is used to align the modeled data obtained after filtering and denoising in terms of time and space. For time alignment, the pulses of the encoder can be utilized to align the sampling frequencies of the point cloud data of the two radar systems, namely the structural data and the 3D point cloud data, through linear interpolation, so as to align the modeled data in time. For space alignment, the odometer wheel data carried on the in-pipe detector 1 can be used for alignment. According to the odometer wheel data, the positions of the in-pipe detector 1 at each moment can be obtained. And since the installation positions of the millimeter-wave radar 2 and the lidar 3 on the in-pipe detector 1 are fixed, the positions of the modeled data at each moment can be obtained, and the modeled data can be spatially registered to achieve spatial alignment of the modeled data. By aligning the millimeter-wave radar 2 and the lidar 3 in time and space, it lays a foundation for subsequent fusion of the modeled data.

[0048] Perform the MLS fusion operation on the modeled data, and the specific operation process is as follows: Utilize the MLS algorithm to fuse the modeled data to obtain modeled fusion data; the MLS algorithm is used to fuse the point cloud data from multiple perspectives, namely the modeled data. The basic principle is to perform polynomial fitting on the local modeled data to find a function curve suitable for local use of the point cloud. Then, the coordinates of each data point are recalculated through this curve to achieve smoothing of the modeled data.

[0049] The modeling module 4 is further configured as: Define the MLS model; the MLS model is configured as: According to the data points in the modeled data and the neighborhood points of the data points, use the weighted least squares method to obtain the fitting surface of the modeled data; the core of the MLS model is to construct a local polynomial surface (usually linear) for the neighborhood of each data point. For any data point in the modeled data, its neighborhood point 's fitting surface is obtained by minimizing the weighted least squares error, and the fitting surface is: ; In the formula, is the data point; is the neighborhood point; is a polynomial function, which is an estimated value of a certain attribute, such as the estimated height of each neighborhood point; is the weight function, the closer to the central data point, the greater the weight, and the Gaussian kernel function is adopted; are the coefficients of the polynomial function; is the actual value of the height of each neighborhood point. Among them, the goal of the MLS model is to optimize the coefficient to minimize the value of the fitting surface, and then with The polynomial function with as the coefficient is the approximate surface around the data points.

[0050] According to the data points in the modeling data, using a KD tree, obtain the neighborhood points of the data points; through the given data points , find all neighborhood points within its radius r (or the nearest k points).

[0051] According to the data points and the neighborhood points, determine a linear equation; the linear equation is: ; In the formula, is the design matrix, determined by the neighborhood points, is the value of the polynomial function at ; is the diagonal weight matrix, determined by the data points and the neighborhood points, is composed of ; is the height value or attribute value of the neighborhood points.

[0052] Input the data points, neighborhood points, and linear equation into the MLS model to obtain the fitted surface of the modeling data; by inputting the data points, neighborhood points, and linear equation into the MLS model for solution, obtain the fitted surface of the modeling data.

[0053] Project the data points onto the fitted surface to obtain the modeled fusion data. Project the original points onto the fitted local surface to obtain the smoothed positions . Through 's set, obtain the modeled fusion data. [[ID= / / ]]

[0054] Perform a 3D modeling operation on the modeled fusion data, and the specific operation process is as follows: Based on the modeled fusion data, use the Poisson reconstruction model to generate the inner wall model of the pipeline to be measured.

[0055] The modeling module 4 is further configured to: Based on the modeled fusion data, use the Poisson reconstruction model to obtain the implicit equation of the modeled fusion data; perform an isosurface extraction operation on the implicit equation to generate the inner wall model of the pipeline to be measured.

[0056] Specifically, Poisson surface reconstruction is a triangulation reconstruction algorithm based on implicit functions. By solving the Poisson equation, an implicit equation describing the surface information of the modeled fusion data is obtained, and then an isosurface extraction is performed on this equation to finally generate a surface mesh model with geometric entity information, that is, the three-dimensional model of the inner wall of the pipeline to be measured. Then, through professional software such as CloudCompare, the mesh data is converted into a standard STL file for users to use.

[0057] The second aspect of the present application provides a method for modeling the inner wall of a pipeline, which is applied to a radar system for modeling the inner wall of a pipeline described in any one of the above first aspects, and includes: Based on the modeling data, using multi-source spatio-temporal registration operations, align the modeling data in time and space; the modeling data includes: structural data and 3D point cloud data; the structural data includes: the distance between the millimeter-wave radar and the inner wall of the pipeline to be measured, the speed of the in-pipe detector, and the material of the pipeline to be measured. Using the MLS algorithm, fuse the modeling data to obtain modeling fusion data; Based on the modeling fusion data, use the Poisson reconstruction model to generate the inner wall model of the pipeline to be measured.

[0058] It should be noted that the effects of the above method embodiments can be referred to the effects of the above system embodiments, which will not be elaborated here.

[0059] The above specific implementation manners have further elaborated the purpose, technical solutions, and beneficial effects of the embodiments of the present application. It should be understood that the above are only the specific implementation manners of the embodiments of the present application, and are not used to limit the protection scope of the embodiments of the present application. Any modifications, equivalent replacements, improvements, etc. made on the basis of the technical solutions of the embodiments of the present application should be included in the protection scope of the embodiments of the present application.

Claims

1. A radar system for inner wall modeling of a pipeline, characterized in that, Including: A millimeter-wave radar (2) and a lidar (3) provided on an in-pipe detector (1); the in-pipe detector (1) is configured to travel from the inlet end of a pipeline to be measured to the outlet end of the pipeline to be measured; The millimeter-wave radar (2) is configured to: Transmit electromagnetic waves to the inner wall of the pipeline to be measured and receive the reflected waves of the electromagnetic waves; Based on the reflected waves, using the Doppler effect and echo analysis operations, generate structure data of the inner wall of the pipeline to be measured; the structure data includes: the distance between the millimeter-wave radar (2) and the inner wall of the pipeline to be measured, the speed of the in-pipe detector (1), and the material of the pipeline to be measured; The lidar (3) is configured to: Transmit a laser beam to the inner wall of the pipeline to be measured and receive the reflected signal of the laser beam; According to the reflected signal, calculate the flight time or phase difference of the laser beam to generate 3D point cloud data of the inner wall of the pipeline to be measured; A modeling module (4), the modeling module (4) is communicatively connected to the millimeter-wave radar (2) and the lidar (3), and the modeling module (4) is configured to: Based on the modeling data, using multi-source spatio-temporal registration operations, align the modeling data in time and space; the modeling data includes: the structure data and the 3D point cloud data; Using the MLS algorithm, fuse the modeling data to obtain modeling fusion data; Based on the modeling fusion data, using the Poisson reconstruction model, generate a model of the inner wall of the pipeline to be measured.

2. The radar system for pipeline inner wall modeling according to claim 1, wherein, The lidar (3) includes: A first housing (31), a receiver (32) is provided inside the first housing (31); A second housing (33), the second housing (33) is communicated with the first housing (31), and a light-transmitting hole (331) is provided on the surface of the second housing (33); a bevel mirror (34) is provided inside the second housing (33); A laser source (35), the laser source (35) is configured to: Transmit a laser beam to the bevel mirror (34), and the bevel mirror (34) reflects the laser beam through the light-transmitting hole (331) to the inner wall of the pipeline to be measured; the reflected signal from the inner wall of the pipeline to be measured is reflected by the bevel mirror (34) to the receiver (32); The receiver (32) is configured to: According to the reflected signal, calculate the flight time or phase difference of the laser beam to generate 3D point cloud data of the inner wall of the pipeline to be measured.

3. A radar system for pipeline inner wall modeling according to claim 2, characterized in that, An optical rotary encoder (36) is provided between the first housing (31) and the second housing (33), and the optical rotary encoder (36) is communicated with the first housing (31) and the second housing (33); the laser source (35) is provided inside the optical rotary encoder (36); the optical rotary encoder (36) is configured to rotate the second housing (33) at a first preset angle; A servo motor (37) is further provided inside the second housing (33), the servo motor (37) is connected to the bevel mirror (34), and the servo motor (37) is configured to rotate the bevel mirror (34) at a second preset angle.

4. A radar system for inner wall modeling of a pipeline according to claim 1, characterized in that, The number of the lidars (3) located in the horizontal direction of the in-pipe detector (1) is 1, and the horizontal detection range of the lidar (3) is 360°; the number of the millimeter-wave radars (2) located in the horizontal direction of the in-pipe detector (1) is 4, and the horizontal detection range of the millimeter-wave radar (2) is 90°.

5. A radar system for pipeline inner wall modeling according to claim 1, characterized in that, The modeling module (4) is further configured to: Obtain each data point in the 3D point cloud data and the neighborhood points of a preset number of the data points; Calculate the average distance and standard deviation between the data point and the neighborhood points; Calculate a retention distance according to the average distance and the standard deviation; the retention distance is: ； In the formula, is the average distance; is the standard deviation; Remove the neighborhood points outside the retention distance among the data points.

6. A radar system for pipeline inner wall modeling according to claim 1, characterized in that, The modeling module (4) is further configured to: Establish a sliding window based on the structure data; Obtain the data points in the sliding window for a continuous preset number of frames, and if there are target data points that do not appear repeatedly among the data points, remove the target data points.

7. A radar system for inner wall modeling of a pipeline according to claim 1, characterized in that, The modeling module (4) is further configured to: Use an encoder pulse to align the sampling frequency of the modeling data by linear interpolation method, so that the modeling data is aligned in time; Obtain the odometer wheel data of the in-pipe detector (1); Use the odometer wheel data to perform a spatial registration operation on the modeling data, so that the modeling data is aligned in space.

8. A radar system for pipeline inner wall modeling according to claim 1, characterized in that, The modeling module (4) is further configured to: Define an MLS model; the MLS model is configured to: According to the data points in the modeling data and the neighborhood points of the data points, use the weighted least squares method to obtain a fitting surface of the modeling data; the fitting surface is: ； Wherein, is a data point; is a neighborhood point; is a polynomial function; is a weight function; are the coefficients of the polynomial function; is the actual value of the height of each neighborhood point; According to the data points in the modeling data, use a KD tree to obtain the neighborhood points of the data points; Determine a linear equation according to the data point and the neighborhood points; the linear equation is: ； wherein, is a design matrix determined by the neighborhood points; is a diagonal weight matrix determined by the data points and the neighborhood points; Input the data point, the neighborhood points, and the linear equation into the MLS model to obtain a fitting surface of the modeling data; Project the data points onto the fitting surface to obtain modeling fusion data.

9. A radar system for inner wall modeling of a pipeline according to claim 1, characterized in that, The modeling module (4) is further configured to: Based on the modeling fusion data, use a Poisson reconstruction model to obtain an implicit equation of the modeling fusion data; Perform an isosurface extraction operation on the implicit equation to generate the inner wall model of the pipeline to be measured.

10. A method for modeling the inner wall of a pipeline, which is applied to a radar system for modeling the inner wall of a pipeline according to any one of claims 1 to 9 above, characterized in that, Including: Based on the modeling data, use a multi-source spatio-temporal registration operation to align the modeling data in time and space; The modeling data includes: structure data and 3D point cloud data; the structure data includes: the distance between the millimeter-wave radar and the inner wall of the pipeline to be measured, the speed of the in-pipe detector, and the material of the pipeline to be measured; Use the MLS algorithm to fuse the modeling data to obtain modeling fusion data; Based on the modeling fusion data, use a Poisson reconstruction model to generate the inner wall model of the pipeline to be measured.

Images