A mixing station material volume detection system and method based on dynamic space compensation and multiple echo analysis

Abstract

The application discloses a kind of based on dynamic space compensation and multiple echo analysis's mixing station material volume detection system and method, through the echo data of laser measuring line dynamic adjustment cleaning device's working state, can clean laser scanning device, to actively remove optical interference, environmental adaptability is strong;At the same time, by self-calibration reference datum provides static space pose reference, carries out pose compensation repair to material point cloud data, can adaptively correct mechanical pose drift, solves the pain point of sensor pose drift under the strong vibration environment of mixing station, without frequent calibration manually;Further, in combination with the reference curved surface of hopper bottom surface under the empty load state and the constructed material surface parameterized continuous curved surface equation to carry out volume calculation, can capture material partial heap and other irregular geometric characteristics, improve detection stability reliability.

Description

Technical Field

[0001] This invention relates to the field of civil engineering technology, specifically to a material volume detection system and method for mixing plants based on dynamic spatial compensation and multiple echo analysis. Background Technology

[0002] The detection of material volume in the batching hoppers of concrete mixing plants is a crucial step in ensuring the accuracy of concrete mix proportions and the continuity of production. Currently, the detection of material volume in the batching hoppers of mixing plants mainly relies on traditional manual inspection methods. Operators need to judge the material level by visual inspection or simple tools. This method is not only time-consuming and labor-intensive, but also prone to significant subjective errors. Existing technologies also use devices such as cameras or lidar for material volume detection, but these are costly, involve complex data processing, and have low detection efficiency.

[0003] Chinese Patent No. CN 120740443 A discloses a method and apparatus for detecting the volume of material in a hopper, comprising the following steps: S1 using a 3D camera to take a 3D picture of the material in the hopper at an angle and generating a 3D point cloud map; S2 in the generated point cloud map, detecting the areas with voids one by one, and then filling each void area to calibrate the 3D point cloud map; S3 starting to detect the volume of the material, obtaining the height difference between the surface height of the material and the bottom of the hopper, and then calculating the volume of the corresponding material in the hopper by integrating the bottom surface information of the calibrated 3D point cloud map.

[0004] However, existing detection technologies still have the following insurmountable shortcomings in the extreme industrial environment of mixing plants:

[0005] (1) Weak resistance to dust interference and frequent optical failures: The high concentration of dust generated during the operation of the mixing plant will not only block the laser or 3D imaging optical path, causing noise and depth misjudgment, but will also adhere and deposit on the optical lens. Existing detection devices will fail due to light intensity decay after running for several hours, requiring manual intervention and making continuous operation impossible.

[0006] (2) Geometric reference drift under dynamic working conditions: The strong vibration generated by the unloading and weighing of the batching machine can easily cause the sensor bracket to shift or deflect. Existing technology relies on the fixed calibration parameters of the initial installation. Changes in position will cause cumulative deviation of coordinate mapping, resulting in a rapid decrease in measurement accuracy over time.

[0007] (3) Low accuracy of modeling irregular stacking: Existing algorithms mostly assume that the material surface is regular and smooth, which makes it difficult to adapt to the irregular shapes such as collapse, hanging and stacking caused by moisture content, aggregate particle size and impact of falling material under actual working conditions. They cannot accurately capture nonlinear geometric features, and there are large systematic errors in volume calculation.

[0008] Therefore, providing a material volume detection system for mixing plants that can actively eliminate optical interference, adaptively correct mechanical pose drift, and perform volume detection on complex collapsing materials to improve detection stability and reliability has become an urgent problem to be solved in this field. Summary of the Invention

[0009] To address the shortcomings of existing technologies, the purpose of this invention is to provide a stable and reliable material volume detection system and method for mixing plants based on dynamic spatial compensation and multiple echo analysis, which can achieve optical interference removal and spatial coordinate system self-repair.

[0010] To achieve the above objectives, the present invention provides a material volume detection system for a mixing plant based on dynamic spatial compensation and multiple echo analysis, comprising a dual-axis laser scanning unit, a self-calibrating reference standard, and a multi-dimensional data processing unit.

[0011] The dual-axis laser scanning unit is positioned above the hopper and includes a laser scanning device and a cleaning device. The laser scanning device emits laser lines into the hopper and adjusts the emission direction of the laser lines to scan the initial point cloud data of the hopper when it is empty and the material point cloud data when it is loaded. The cleaning device is used to clean the laser scanning device.

[0012] The self-calibrating reference datum is located at the top of the hopper, reflects the laser measurement line, and provides a static spatial pose reference.

[0013] The multi-dimensional data processing unit can dynamically adjust the working state of the cleaning device according to the echo data of the laser measurement line. It can also remove dust noise and discrete noise in the material point cloud data, and perform pose compensation and repair on the material point cloud data based on the static spatial pose reference to construct the parameterized continuous surface equation of the material surface. Then, it can calculate the material volume by combining the reference surface of the hopper bottom obtained based on the initial point cloud data.

[0014] Furthermore, the laser scanning device includes a laser emitter and a reflecting mirror. The laser emitter can emit laser measurement lines towards the reflecting mirror, and the reflecting mirror can reflect the laser measurement lines to the hopper. The laser emitter and the reflecting mirror can rotate relative to each other.

[0015] Furthermore, the cleaning device includes a blowing module, a vibration cleaning module, and an elastic flexible scraping module, wherein the vibration cleaning module is disposed on the back side of the reflective mirror.

[0016] Furthermore, the self-calibrating reference reference is disposed in the non-loading area at the top of the hopper.

[0017] Furthermore, the dual-axis laser scanning unit scans the self-calibration reference reference component when the hopper is empty to obtain standard spatial coordinates, and converts the polar coordinates when scanning the material point cloud data when the hopper is loaded into rectangular coordinates in the hopper space to obtain the current measured coordinates.

[0018] Furthermore, the multi-dimensional data processing unit, based on the singular value decomposition algorithm, compares the standard spatial coordinates with the current measured coordinates to solve for the optimal rotation matrix and translation vector, and performs pose compensation and repair on the material point cloud data.

[0019] Furthermore, the multi-dimensional data processing unit, based on time-of-flight multiple echo analysis technology, collects the echo sequence generated by the dual-axis laser scanning unit after emitting laser measurement lines in the hopper loading state, and identifies valid echoes by using preset pulse width thresholds and echo intensity thresholds, eliminating dust noise points, and then uses a radius filter to eliminate discrete noise points where the number of neighboring points within a preset radius is less than a preset threshold.

[0020] Furthermore, the multi-dimensional data processing unit uses principal component analysis to extract the feature values ​​of the material point cloud data, determines the material accumulation morphology, and constructs the parameterized continuous surface equation of the material surface based on the non-uniform rational B-spline algorithm.

[0021] Furthermore, the multi-dimensional data processing unit projects and closes the parameterized continuous surface equation of the material surface onto the reference surface of the hopper bottom, and uses a discretized micro-element mesh to perform spatial triple integral calculation to obtain the material volume.

[0022] To achieve the above objectives, the present invention provides a method for detecting the volume of materials in a mixing plant based on dynamic spatial compensation and multiple echo analysis. This method is based on a material volume detection system for a mixing plant constructed using the aforementioned scheme and includes:

[0023] Static benchmark modeling: In the empty state of the hopper, the dual-axis laser scanning unit scans the initial point cloud data of the inner wall of the hopper to obtain the reference surface of the bottom surface of the hopper, and scans the self-calibrating reference benchmark to obtain the standard spatial coordinates.

[0024] Dynamic point cloud acquisition and noise reduction: When the hopper is loaded, the dual-axis laser scanning unit scans the material in the hopper to acquire the material point cloud data. The multi-dimensional data processing unit removes dust noise and discrete noise in the material point cloud data and dynamically adjusts the cleaning device according to the echo data of the laser measurement line to clean the laser scanning device.

[0025] Self-healing spatial coordinate system: The multi-dimensional data processing unit performs pose compensation and repair on the material point cloud data based on the static spatial pose reference.

[0026] Material feature identification and surface fitting: The multi-dimensional data processing unit extracts feature values ​​of the repaired material point cloud data and constructs the parameterized continuous surface equation of the material surface.

[0027] Spatial volume integral calculation: The multi-dimensional data processing unit combines the parameterized continuous surface equation of the material surface with the reference surface of the hopper bottom to calculate the material volume.

[0028] The material volume detection system and method for mixing plants based on dynamic spatial compensation and multiple echo analysis provided by this invention can dynamically adjust the working state of the cleaning device through the echo data of the laser measuring line, and can clean the laser scanning device to actively remove optical interference, with strong environmental adaptability.

[0029] Meanwhile, by providing a static spatial pose reference through a self-calibrating reference reference, pose compensation and repair of material point cloud data can be performed, which can adaptively correct mechanical pose drift, solve the pain point of sensor pose drift under strong vibration environment of mixing plant, and eliminate the need for frequent manual calibration.

[0030] Furthermore, by combining the reference surface of the hopper bottom under the empty state with the constructed parameterized continuous surface equation of the material surface for volume calculation, irregular geometric features such as material piling can be captured, thereby improving the stability and reliability of detection. Attached Figure Description

[0031] The present invention will be further described below with reference to the accompanying drawings and specific embodiments.

[0032] Figure 1 A block diagram of the material volume detection system for a mixing plant based on dynamic spatial compensation and multiple echo analysis provided by the present invention.

[0033] Figure 2 This is a schematic diagram of the material volume detection system for a mixing plant based on dynamic spatial compensation and multiple echo analysis provided by the present invention.

[0034] Figure 3 This is a schematic diagram of the laser scanning device in this invention;

[0035] Figure 4 This is a schematic diagram of the vibration cleaning module in this invention;

[0036] Figure 5 This is a schematic diagram of the structure of the elastic flexible scraping module in this invention.

[0037] Figure label:

[0038] 1. Dual-axis laser scanning unit; 11. Laser scanning device; 111. Mounting frame; 112. Laser emitter; 113. Reflecting mirror; 114. Rotary mounting frame; 1141. Connecting support rod; 1142. Mounting base; 115. Rotary motor; 116. Offset motor; 117. Absolute encoder; 12. Cleaning device; 121. Blowing module; 122. Vibration cleaning module; 123. Flexible scraping module; 1231. Flexible scraper; 1232. Sweeping motor;

[0039] 2. Self-calibrating reference standard; 3. Multi-dimensional data processing unit; 4. Hopper. Detailed Implementation

[0040] To make the technical means, creative features, objectives and effects of this invention easier to understand, the invention will be further described below with reference to specific illustrations.

[0041] See Figure 1 and Figure 2 The image shows an example of a material volume detection system for a mixing plant based on dynamic spatial compensation and multiple echo analysis provided by the present invention.

[0042] The material volume detection system for mixing plants based on dynamic spatial compensation and multiple echo analysis in this example mainly includes a dual-axis laser scanning unit 1, a self-calibrating reference datum 2, and a multi-dimensional data processing unit 3.

[0043] The dual-axis laser scanning unit 1 is set above the hopper 4 and includes a laser scanning device 11 and a cleaning device 12. The laser scanning device 11 can emit laser measurement lines to the hopper 4 and adjust the emission direction of the laser measurement lines to scan the initial point cloud data of the hopper 4 in the empty state and the material point cloud data in the hopper in the loaded state. The cleaning device 12 is used to clean the laser scanning device 11.

[0044] The self-calibrating reference reference 2 is set at the top of the hopper 4, which can reflect the laser measurement line and provide a static spatial pose reference;

[0045] The multi-dimensional data processing unit 3 can dynamically adjust the working state of the cleaning device 12 according to the echo data of the laser measurement line. It can also remove dust noise and discrete noise in the material point cloud data, and perform pose compensation and repair on the material point cloud data based on the static spatial pose reference. It constructs the parameterized continuous surface equation of the material surface, and calculates the material volume by combining the reference surface of the hopper bottom obtained based on the initial point cloud data. It can actively remove optical interference, adaptively correct mechanical pose drift, and perform volume detection on complex collapsed materials, thereby improving the stability and reliability of detection.

[0046] Combination Figure 2The dual-axis laser scanning unit 1 is located above the hopper 4 and includes a laser scanning device 11 and a cleaning device 12. The laser scanning device 11 can emit laser measurement lines to the hopper 4, and the cleaning device 12 can clean the laser scanning device 11 to ensure that the laser measurement lines are not affected by dust and actively remove optical interference.

[0047] Combination Figure 3 Specifically, the laser scanning device 11 includes a mounting frame 111, a laser emitter 112, and a reflecting mirror 113. The laser emitter 112 and the reflecting mirror 113 are respectively disposed on both sides of the mounting frame 111. The laser emitter 112 can emit laser measurement lines to the reflecting mirror 113, so that the reflecting mirror 113 reflects the laser measurement lines to the hopper 4. The reflecting mirror 113 can rotate relative to the laser emitter 112 to adjust the reflection direction of the laser measurement lines, thereby covering the inner wall of the hopper 4 and obtaining comprehensive and accurate point cloud data.

[0048] Furthermore, the reflector 113 is connected to the mounting frame 111 via a rotating mounting bracket 114. The rotating mounting bracket 114 includes a connecting support rod 1141 and a mounting base 1142. One end of the connecting support rod 1142 is connected to the mounting base 1142, and the other end is hinged to the reflector 113. The mounting base 1142 is hinged to the side plate of the mounting frame 111 and connected to a rotary motor 115, so that the rotary motor 115 can drive the rotating mounting base 114 to rotate around the transverse axis of the mounting frame 111, thereby causing the reflector 113 to rotate synchronously facing the top or bottom surface of the mounting frame 111 to adjust the radial reflection direction of the laser measurement line.

[0049] The reflector 113 is connected to the mounting frame 111 via a rotating mounting bracket 114. The rotating mounting bracket 114 consists of a connecting support rod 1141 and a mounting base 1142. One end of the connecting support rod 1141 is connected to the mounting base 1142, and the other end is hinged to the reflector 113. The mounting base 1142 is hinged to the side plate of the mounting frame 111 and connected to a rotary motor 115. Therefore, the rotary motor 115 can drive the rotating mounting bracket 114 to rotate around the transverse axis of the mounting frame 111, thereby causing the reflector 113 to rotate synchronously, so that the reflector 113 rotates toward the top or bottom surface of the mounting frame 111, and synchronously adjusts the radial reflection direction of the laser measuring line (i.e., the laser measuring line is adjusted up and down along the height of the hopper).

[0050] Meanwhile, the reflector 113 is hinged to the connecting support rod 1141 and can rotate independently relative to the rotating mounting frame 114. The reflector 113 is also connected to the deflection motor 116. Therefore, the deflection motor 116 can drive the reflector 113 to rotate toward both sides of the mounting frame 111 and simultaneously adjust the axial reflection direction of the laser measuring line (i.e., the laser measuring line is adjusted left and right along the width of the hopper).

[0051] Based on the above structure, the reflector 113 can rotate on two axes relative to the laser emitter 112, flexibly adjusting the reflection direction of the laser measuring line to ensure that the laser measuring line can fully cover the entire inner wall of the hopper 4, thereby obtaining comprehensive and accurate point cloud data of the inner wall of the hopper and the material.

[0052] Combination Figure 3 Preferably, the rotary motor 115 and the deflection motor 116 each integrate an absolute encoder 117. The absolute encoder 117 can measure the lateral axis rotation angle of the rotary motor 115 and the independent rotation angle of the deflection motor 116 in real time, and feed it back to the multi-dimensional data processing unit 3. This allows the multi-dimensional data processing unit 3 to combine the angle data fed back by the absolute encoder 117 to accurately control the rotation direction and amplitude of the rotary motor 115 and the deflection motor 116, thereby precisely adjusting the dual-axis rotation angle of the reflector 113. This ensures that the scanning trajectory of the laser measuring line covers the entire space of the hopper, avoiding incomplete point cloud data acquisition and insufficient accuracy due to angle deviation.

[0053] As a preferred configuration, the multi-dimensional data processing unit 3 coordinates the control of the rotary motor 115 and the deflection motor 11, so that the laser measuring line scans the inner cavity of the hopper 4 in an equiangular spiral path, ensuring scanning accuracy and stability.

[0054] As an example, the multi-dimensional data processing unit 3 decomposes the equiangular spiral scanning trajectory into two dimensions: radial and axial motion. The rotary motor 115 drives the reflector 113 to rotate around the transverse axis of the mounting frame 111, controlling the laser measuring line to advance radially layer by layer along the height direction of the hopper 4. The deflection motor 116 drives the reflector 113 to rotate toward both sides of the mounting frame 111, controlling the laser measuring line to continuously scan axially along the width direction of the hopper 4. At the same time, the multi-dimensional data processing unit 3 dynamically adjusts the speed and angle of the rotary motor 115 and the deflection motor 116 according to the mathematical model of the equiangular spiral, ensuring that the laser measuring line advances radially with a constant pitch increment every time it rotates through a fixed angle, ultimately realizing that the laser measuring line scans the inner cavity of the hopper with an equiangular spiral path.

[0055] The laser scanning device 11 thus constitutes a laser measuring line that can emit laser measuring lines into the hopper 4, and the scanning path covers the inner wall of the hopper 4 or the surface of the material, thus acquiring comprehensive and accurate point cloud data.

[0056] To reduce the interference of dust on the laser measurement line, the cleaning device 12 can clean the reflective mirror 113, actively remove optical interference, and improve the accuracy of point cloud data.

[0057] Specifically, the cleaning device 12 includes a blowing module 121, a vibration cleaning module 122, and an elastic flexible scraping module 123, wherein, for example... Figure 3As shown, the purging module 121 is installed inside the mounting frame 111 and positioned above the reflector 113. It can spray high-pressure air onto the reflector 113 to remove dust from the reflector 113.

[0058] Combination Figure 4 Furthermore, the vibration cleaning module 122 is preferably composed of ultrasonic transducers, such as piezoelectric ceramic transducers, and the four ultrasonic transducers are arranged in a cross-shaped symmetrical distribution, which can synchronously vibrate and shake off the dust attached to the reflective mirror 113.

[0059] Combination Figure 5 In conjunction with this, the flexible scraping module 123 is mounted on the mounting frame 111 and includes a flexible scraper 1231 and a cleaning motor 1232. Preferably, the flexible scraper 1231 is hinged to the deflection motor 116 via the cleaning motor 1232, and its installation position avoids the dual-axis rotation trajectory driven by the rotating motor 115 and the deflection motor 116 around the reflector 113, so as not to occupy the rotation space of the reflector 113. During cleaning, the cleaning motor 1232 can drive the flexible scraper 1231 to move and adhere to the surface of the reflector 113, and achieve the scraping and cleaning of the dust deposited on the reflector 113 under the synchronous vibration of the vibration cleaning module 122. The scraped dust can also be shaken off synchronously under the vibration.

[0060] After cleaning is completed, the flexible scraper 1231 is reset to the initial avoidance position under the drive of the sweeping motor 1232, and the reflector 113 resumes normal dual-axis rotation, realizing the coordinated operation of the cleaning function and the rotation function, without interfering with the radial rotation of the reflector 113 around the transverse axis of the mounting frame 111 and the axial rotation of the relative rotating mounting frame 114.

[0061] Therefore, the cleaning device 12 can achieve multiple cleaning of the reflective mirror 113 by working together with the blowing module 121, the vibration cleaning module 122 and the elastic flexible scraping module 123. It can effectively remove the dust deposited on the reflective mirror 113, thereby actively removing optical interference, reducing the interference of dust on the laser measurement line and improving the accuracy of point cloud data.

[0062] It should be noted that after the laser measuring line irradiates the surface of the material (or dust), it generates a reflection and reflects the echo sequence back to the dual-axis laser scanning unit 1.

[0063] To ensure cleaning effectiveness, the multi-dimensional data processing unit 3 can continuously collect echo data of the laser measurement line in real time when the laser scanning device 11 scans the material in the hopper 4 and emits laser measurement lines to the surface of the material, based on time-of-flight multiple echo analysis technology. It can also dynamically adjust the working status of the blowing module 121, the vibration cleaning module 122 and the elastic flexible scraping module 123 according to the laser echo signal-to-noise ratio.

[0064] Specifically, the multi-dimensional data processing unit 3 compares the laser echo signal-to-noise ratio (SNR) with a preset SNR benchmark and a SNR threshold. If the laser echo SNR drops from the SNR benchmark to the SNR threshold, the multi-dimensional data processing unit 3 first controls the blowing module 121 to spray high-pressure air onto the reflector surface 113 to remove dust from the reflector surface 113. If the laser echo SNR does not recover to the SNR benchmark after blowing the reflector surface 113, the vibration cleaning module 122 and the elastic flexible scraping module 123 are controlled to work synchronously to clean the dust deposited on the reflector surface 113. This avoids over-cleaning and damaging components while ensuring cleaning efficiency, maximizing the optical cleanliness of the reflector surface 113, and actively eliminating optical interference.

[0065] The dual-axis laser scanning unit 1 thus constitutes a laser scanning device 11 and a cleaning device 12 working together to emit laser measurement lines into the hopper 4, and adjust the emission direction of the laser measurement lines to cover the inner wall of the hopper 4 or the surface of the material, and actively remove optical interference to improve the accuracy of point cloud data.

[0066] In practical applications, the dual-axis laser scanning unit 1 first scans the inner wall of the hopper 4 to obtain initial point cloud data when the hopper is empty, and then scans the material in the hopper 4 to obtain material point cloud data when the hopper is loaded.

[0067] It should be noted that when the dual-axis laser scanning unit 1 emits laser measurement lines, the multi-dimensional data processing unit 3 can process the echo sequence corresponding to each laser measurement line, extract the distance, angle and other information carried by the echo, and convert this information into three-dimensional spatial coordinate points. A large number of three-dimensional spatial coordinate points are collected to form point cloud data. This processing is a conventional technical means in this field and will not be described in detail here.

[0068] To ensure the accuracy and reliability of material point cloud data and reduce dust interference, the multi-dimensional data processing unit 3 can remove dust noise and discrete noise from the material point cloud data.

[0069] Specifically, when the multi-dimensional data processing unit 3 acquires the echo sequence generated by the dual-axis laser scanning unit 1 after emitting laser measurement lines in the hopper loading state, it can pass the preset pulse width threshold. and echo intensity threshold Identify valid echoes; where the pulse width threshold is... and echo intensity threshold The settings are configured to meet the characteristics of the echo signal reflected after the laser measuring line irradiates the dust.

[0070] If in the echo sequence of material point cloud data, the pulse width of a certain echo signal... and echo intensity satisfy > and > If the signal is clear, the echo signal is considered a valid echo; otherwise, it is considered an invalid echo and is discarded. This achieves the identification of valid echoes and the precise filtering of dust noise.

[0071] Furthermore, the multi-dimensional data processing unit 3 can also use a radius filter to remove sampling points within a preset radius. Number of inner neighbor points Discrete noise points less than a preset threshold are used to ensure the accuracy of material point cloud data.

[0072] Therefore, the multi-dimensional data processing unit 3 removes dust noise and discrete noise from the material point cloud data, ensuring the accuracy and reliability of the material point cloud data.

[0073] The strong vibrations generated by the hopper 4 during loading and operation are transmitted to the mounting frame 111 through the mixing plant frame, mounting foundation and other structures, causing slight displacement or angular deflection of the dual-axis laser scanning unit 1. At the same time, the dual-axis laser scanning unit 1 needs to maintain a fixed relative spatial position with the hopper 4 as a scanning reference. Under long-term vibration and impact, the relative mounting reference between the dual-axis laser scanning unit 1 and the hopper 4 will also show slight offset, resulting in a pose deviation in the material point cloud data acquired by the dual-axis laser scanning unit 1.

[0074] To this end, the system sets a self-calibration reference reference 2 at the top of the hopper 4. Before starting the mixing plant, the system scans the self-calibration reference reference 2 with a dual-axis laser scanning unit 1 to obtain the spatial coordinates of the self-calibration reference reference 2. The spatial coordinates are used as a static spatial pose reference to supplement and repair the pose of the material point cloud data, so as to adaptively correct mechanical pose drift, eliminate deviations, and ensure accurate and reliable detection.

[0075] Specifically, the self-calibration reference reference 2 is made of retroreflective material, enabling it to reflect the laser measurement line. The reflected echo intensity is much stronger than that of materials and dust, creating a significant abrupt change in echo intensity. Based on this abrupt change in echo intensity, the multi-dimensional data processing unit 3 can determine that the echo signal originates from the self-calibration reference reference 2, accurately pinpoint its spatial coordinates, and set these coordinates as the standard spatial coordinates. It serves as a static spatial pose reference, providing a basis for subsequent pose compensation and repair of material point cloud data.

[0076] The self-calibrating reference reference 2 is preferably placed in the non-loading area at the top of the hopper to avoid interference from the loading operation and to ensure its reflection performance and coordinate stability.

[0077] After acquiring the static spatial pose reference, the dual-axis laser scanning unit 1 first scans the inner wall of the hopper 4 to obtain initial point cloud data when the hopper is empty, and then scans the material in the hopper 4 to obtain material point cloud data when the hopper is loaded.

[0078] Furthermore, the multi-dimensional data processing unit 3 processes the polar coordinates of the material point cloud data scanned by the dual-axis laser scanning unit 1 when the hopper is loaded. Real-time conversion to rectangular coordinates of hopper space To obtain the current measured coordinates .

[0079] As an example, the specific calculation is as follows: .

[0080] Furthermore, the multi-dimensional data processing unit 3, based on the singular value decomposition (SVD) algorithm, compares the standard spatial coordinates. Compared with the current measured coordinates Solve for the optimal rotation matrix Translation vector And perform pose compensation and repair on the material point cloud data.

[0081] As an example, the multi-dimensional data processing unit 3 first constructs a decentralized covariance matrix. ,in, and These are the geometric centroids of the point cloud data obtained by the dual-axis laser scanning unit 1 scanning the self-calibration reference datum 2, and the geometric centroids of the material point cloud data (point cloud data to be repaired by pose compensation after noise reduction) obtained by the dual-axis laser scanning unit 1 scanning the material in the hopper when the hopper is loaded.

[0082] Next, the multi-dimensional data processing unit decomposes the covariance matrix H into singular value decomposition. Solving for the given information yields the following results. as well as Thus, the optimal rotation matrix is ​​obtained. Translation vector .

[0083] Furthermore, the multi-dimensional data processing unit 3 processes the material point cloud data. Pose compensation and repair were performed to obtain the corrected material point cloud data. The specific calculation is as follows: This is to achieve pose compensation and repair of material point cloud data.

[0084] Therefore, the multi-dimensional data processing unit 3 can perform pose compensation and repair on the material point cloud data based on the static spatial pose reference of the self-calibrated reference reference component, adaptively correct mechanical pose drift, solve the pain point of pose drift in the strong vibration environment of the mixing plant, eliminate the need for frequent manual calibration, and ensure that the measurement accuracy does not drift over time, making the measurement accurate and reliable.

[0085] Furthermore, the multi-dimensional data processing unit 3 can also construct the parameterized continuous surface equation of the material surface based on the denoised and pose-restored material point cloud data.

[0086] Specifically, the multi-dimensional data processing unit 3 uses principal component analysis (PCA) to extract the feature values ​​of the material point cloud data, determines the material packing morphology, and constructs the parameterized continuous surface equation of the material surface based on the non-uniform rational B-spline algorithm (NURBS).

[0087] As an example, the multi-dimensional data processing unit 3 first centralizes the denoised and pose-compensated material point cloud data to construct a point cloud covariance matrix. Then, it performs eigenvalue decomposition on the covariance matrix to obtain three eigenvalues ​​(corresponding to the variance of the material point cloud data in the principal directions of three-dimensional space), sorted by magnitude as λ1≥λ2≥λ3. By analyzing the proportional relationship of the eigenvalues, the material accumulation pattern can be determined. For example, if λ1 is much larger than λ2 and λ3, it indicates that the material extends along a single direction, indicating a partial accumulation pattern; if λ1 and λ2 are close and both are much larger than λ3, it indicates that the material extends in the plane and converges in the height direction, indicating a collapse or flattened pattern; if the three eigenvalues ​​are close in value, it is determined to be an approximately uniform cone-shaped or mound-shaped accumulation, thereby capturing irregular geometric features such as partial material accumulation and improving the stability and reliability of detection.

[0088] Furthermore, the multi-dimensional data processing unit 3 employs the Non-Uniform Rational B-Spline (NURBS) algorithm, combined with... Rank and A quasi-uniform basis function (preferably order 3) is used, with a 32×32 control point lattice selected, and the basis function orders p=3 and q=3 are set to ensure the parameterized continuous surface equation of the generated material surface. At the edges, a boundary constraint algorithm automatically conforms to the geometry of the hopper's inner wall, preventing volumetric voids in dead corners of the hopper, and establishing the parameterized continuous surface equations of the material surface. Specifically:

[0089]

[0090] in, As control points, As a weighting factor, and respectively along Xianghe towards Rank and Quasi-uniform basis functions.

[0091] Therefore, the multi-dimensional data processing unit 3 uses the NURBS surface fitting algorithm, which can accurately capture irregular geometric features such as material piling and wall adhesion, and generate accurate parameterized continuous surface equations for the material surface.

[0092] Furthermore, the multi-dimensional data processing unit 3 can also combine the parameterized continuous surface equation of the material surface with the reference surface of the bottom surface of the hopper obtained based on the initial point cloud data under the empty state of the hopper to calculate the material volume.

[0093] As an example, after the multi-dimensional data processing unit 3 performs noise reduction and pose compensation on the initial point cloud data, it selects a subset of the point cloud in the bottom region of the hopper 4, fits a reference surface (such as a plane or cone) that conforms to the geometric characteristics of the bottom of the hopper using the least squares method, and uses the fitted surface as the reference surface B(x, y) of the bottom of the hopper for subsequent material volume calculation.

[0094] Next, the multi-dimensional data processing unit 3 calculates the parameterized continuous surface equation of the material surface. The material volume is obtained by projecting the material onto the reference surface B(x, y) at the bottom of the hopper and performing a spatial triple integral using a discretized infinitesimal mesh. The specific calculation is as follows:

[0095]

[0096] Where Δx and Δy are the areas of the discrete infinitesimal mesh.

[0097] In some embodiments, the calculated material volume is also combined with the hopper sidewall inclination angle for edge constraint correction to ensure the accuracy of the results.

[0098] As an example, firstly, based on the preset inclination angle of the hopper sidewall, the edge region where the material point cloud contacts the hopper sidewall is identified. Then, based on this inclination angle, the parametric surface of the material surface in the edge region is geometrically constrained to correct the integral boundary that exceeds the actual volume of the hopper due to point cloud sampling error or surface fitting deviation. This eliminates the volume calculation deviation at the edge, making the final volume result more consistent with the actual loading conditions of the hopper and ensuring calculation accuracy.

[0099] This constitutes the material volume detection system for mixing plants based on dynamic spatial compensation and multiple echo analysis provided by the present invention.

[0100] This invention also provides a method for detecting the volume of materials in a mixing plant based on dynamic spatial compensation and multiple echo analysis. Based on a detection system constructed using the above-mentioned scheme, this detection method includes:

[0101] First, static benchmark modeling is performed. Under the empty state of the hopper, the dual-axis laser scanning unit 1 scans the initial point cloud data of the inner wall of the hopper 4 to obtain the benchmark surface of the bottom of the hopper, and scans the self-calibration reference benchmark 2 to obtain the standard spatial coordinates.

[0102] Then, dynamic point cloud acquisition and noise reduction are performed. In the hopper loading state, the dual-axis laser scanning unit 1 scans the material in the hopper 4 to acquire material point cloud data. The multi-dimensional data processing unit 3 removes dust noise and discrete noise in the material point cloud data and dynamically adjusts the cleaning device 12 according to the echo data of the laser measurement line to clean the laser scanning device 11.

[0103] Furthermore, spatial coordinate system self-repair is performed, and the multi-dimensional data processing unit 3 performs pose compensation and repair on the material point cloud data based on the static spatial pose reference.

[0104] Next, material feature identification and surface fitting are performed. The multi-dimensional data processing unit 3 extracts the feature values ​​of the repaired material point cloud data and constructs the parameterized continuous surface equation of the material surface.

[0105] Finally, spatial volume integration is performed. The multi-dimensional data processing unit 3 combines the parameterized continuous surface equation of the material surface with the reference surface of the hopper bottom to calculate the material volume.

[0106] The material volume detection system and method for mixing plants based on dynamic spatial compensation and multiple echo analysis provided by this invention can dynamically adjust the working state of the cleaning device through the echo data of the laser measuring line, and can clean the laser scanning device to actively remove optical interference, with strong environmental adaptability.

[0107] Meanwhile, by providing a static spatial pose reference through a self-calibrating reference reference, pose compensation and repair of material point cloud data can be performed, which can adaptively correct mechanical pose drift, solve the pain point of sensor pose drift under strong vibration environment of mixing plant, and eliminate the need for frequent manual calibration.

[0108] Furthermore, by combining the reference surface of the hopper bottom under the empty state with the constructed parameterized continuous surface equation of the material surface for volume calculation, irregular geometric features such as material piling can be captured, thereby improving the stability and reliability of detection.

[0109] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of this invention is defined by the appended claims and their equivalents.

Claims

1. A material volume detection system for a mixing plant based on dynamic spatial compensation and multiple echo analysis, characterized in that, It includes a dual-axis laser scanning unit, a self-calibrating reference standard, and a multi-dimensional data processing unit. The dual-axis laser scanning unit is positioned above the hopper and includes a laser scanning device and a cleaning device. The laser scanning device emits laser lines into the hopper and adjusts the emission direction of the laser lines to scan the initial point cloud data of the hopper when it is empty and the material point cloud data when it is loaded. The cleaning device is used to clean the laser scanning device. The self-calibrating reference datum is located at the top of the hopper, reflects the laser measurement line, and provides a static spatial pose reference. The multi-dimensional data processing unit can dynamically adjust the working state of the cleaning device according to the echo data of the laser measurement line. It can also remove dust noise and discrete noise in the material point cloud data, and perform pose compensation and repair on the material point cloud data based on the static spatial pose reference to construct the parameterized continuous surface equation of the material surface. Then, it can calculate the material volume by combining the reference surface of the hopper bottom obtained based on the initial point cloud data.

2. The material volume detection system for a mixing plant based on dynamic spatial compensation and multiple echo analysis according to claim 1, characterized in that, The laser scanning device includes a laser emitter and a reflective mirror. The laser emitter can emit laser measurement lines towards the reflective mirror, and the reflective mirror can reflect the laser measurement lines to the hopper. The reflective mirror can rotate relative to the laser emitter.

3. The material volume detection system for a mixing plant based on dynamic spatial compensation and multiple echo analysis according to claim 2, characterized in that, The cleaning device includes a blowing module, a vibration cleaning module, and an elastic flexible scraping module, with the vibration cleaning module located on the back of the reflective mirror.

4. The material volume detection system for a mixing plant based on dynamic spatial compensation and multiple echo analysis according to claim 1, characterized in that, The self-calibrating reference datum is located in the non-loading area at the top of the hopper.

5. The material volume detection system for a mixing plant based on dynamic spatial compensation and multiple echo analysis according to claim 1, characterized in that, The dual-axis laser scanning unit scans the self-calibration reference reference component when the hopper is empty to obtain standard spatial coordinates, and converts the polar coordinates of the material point cloud data when scanning the hopper in the loaded state into rectangular coordinates in the hopper space to obtain the current measured coordinates.

6. The material volume detection system for a mixing plant based on dynamic spatial compensation and multiple echo analysis according to claim 5, characterized in that, The multi-dimensional data processing unit, based on the singular value decomposition algorithm, compares the standard spatial coordinates with the current measured coordinates to solve for the optimal rotation matrix and translation vector, and performs pose compensation and repair on the material point cloud data.

7. The material volume detection system for a mixing plant based on dynamic spatial compensation and multiple echo analysis according to claim 1, characterized in that, The multi-dimensional data processing unit, based on time-of-flight multiple echo analysis technology, collects the echo sequence generated by the dual-axis laser scanning unit after emitting laser measurement lines in the hopper loading state. It identifies valid echoes by using preset pulse width thresholds and echo intensity thresholds, removes dust noise, and then uses a radius filter to remove discrete noise points where the number of neighboring points within a preset radius is less than a preset threshold.

8. The material volume detection system for a mixing plant based on dynamic spatial compensation and multiple echo analysis according to claim 1, characterized in that, The multi-dimensional data processing unit uses principal component analysis to extract the feature values ​​of the material point cloud data, determines the material accumulation morphology, and constructs the parameterized continuous surface equation of the material surface based on the non-uniform rational B-spline algorithm.

9. The material volume detection system for a mixing plant based on dynamic spatial compensation and multiple echo analysis according to claim 1, characterized in that, The multi-dimensional data processing unit projects and closes the parameterized continuous surface equation of the material surface onto the reference surface of the hopper bottom, and uses a discretized micro-element mesh to perform spatial triple integral calculation to obtain the material volume.

10. A method for detecting the volume of materials in a mixing plant based on dynamic spatial compensation and multiple echo analysis, characterized in that, Based on the material volume detection system for mixing plants constructed using the above scheme and based on dynamic spatial compensation and multiple echo analysis, this detection method includes: Static benchmark modeling: In the empty state of the hopper, the dual-axis laser scanning unit scans the initial point cloud data of the inner wall of the hopper to obtain the reference surface of the bottom surface of the hopper, and scans the self-calibrating reference benchmark to obtain the standard spatial coordinates. Dynamic point cloud acquisition and noise reduction: When the hopper is loaded, the dual-axis laser scanning unit scans the material in the hopper to acquire the material point cloud data. The multi-dimensional data processing unit removes dust noise and discrete noise in the material point cloud data and dynamically adjusts the cleaning device according to the echo data of the laser measurement line to clean the laser scanning device. Self-healing spatial coordinate system: The multi-dimensional data processing unit performs pose compensation and repair on the material point cloud data based on the static spatial pose reference. Material feature identification and surface fitting: The multi-dimensional data processing unit extracts feature values ​​of the repaired material point cloud data and constructs the parameterized continuous surface equation of the material surface. Spatial volume integral calculation: The multi-dimensional data processing unit combines the parameterized continuous surface equation of the material surface with the reference surface of the hopper bottom to calculate the material volume.

Images