Hole bottom sediment thickness self-adaptive direction measurement method for irregular hole shape

Abstract

The invention relates to a hole bottom sediment thickness adaptive direction measurement method for irregular hole shapes. The method comprises the following steps: acquiring hole wall images and vertical distance data in multiple directions; image preprocessing is carried out, hole wall feature points are extracted, three-dimensional coordinates are converted, a dynamic three-dimensional model is fitted, deformation grades are divided, and an inclination angle is determined; preferentially distributing detection points in a serious deformation area, calculating an optimal angle, monitoring by a gyroscope, correcting by an electromagnetic damper, adjusting by a servo motor, and ensuring that an included angle between a probe axis and a hole bottom tangent line is compliant; transmitting multi-frequency ultrasonic waves according to deformation grades, analyzing reflected signals through wavelet transformation, extracting sediment layering characteristics, and calculating the distance from a probe to a layering surface through speed correction; the reference distance is calculated based on the reference coordinates, the sediment total thickness and the dense layer thickness are finally obtained, and the measurement method which is adaptive to hole wall deformation, eliminates interference, dynamically adjusts detection parameters and is accurate in reference positioning can meet the high-precision requirement of engineering for sediment thickness and compactness quantification.

Description

Technical Field

[0001] This invention belongs to the technical field of hole bottom thickness measurement methods in building drilling, and particularly relates to an adaptive directional measurement method for bottom sediment thickness in irregular hole shapes. Background Technology

[0002] In engineering fields such as building pile foundation construction, geological exploration, and oil and gas well drilling, the thickness and density of the sediment at the bottom of the borehole are key indicators affecting the quality and safety of the project. Excessive sediment thickness or insufficient density can lead to problems such as reduced bearing capacity of the pile foundation, distorted exploration data, and insufficient stability of the oil and gas well tubing. Therefore, it is necessary to accurately measure the thickness of the sediment at the bottom of the borehole.

[0003] Current methods for measuring the thickness of sediment at the bottom of boreholes mainly include manual probing, ultrasonic reflection, and laser ranging. However, these methods have significant technical limitations when dealing with irregular borehole shapes, such as borehole walls with protrusions, depressions, or local tilting, which are common in geologically complex areas or long-depth boreholes. Measurement accuracy is greatly affected by the shape of the hole: Traditional methods often assume that the hole is a regular circle and use a single detection direction or fixed frequency. If there is a depression in the hole wall, the detection signal is easily reflected by the hole wall at the depression, and the distance of the hole wall is mistakenly taken as the thickness of the sediment. If the hole wall is tilted, the probe axis is not perpendicular to the sediment surface, and the thickness calculation error will be caused by the deviation of the signal propagation path. The error can reach 10%-30%, which cannot be adapted to the complex shape of irregular holes.

[0004] Interference factors are difficult to eliminate: water accumulation, air bubbles, dust and other interferences in the borehole can cause the detection signal to attenuate or scatter. Traditional ultrasonic methods use a fixed threshold to distinguish between sediment and interference signals, which can easily misjudge bubble reflection waves as sediment surface reflection waves. Laser ranging methods are greatly affected by water vapor in the borehole and have poor signal penetration. They cannot accurately identify the layer boundary between the upper loose zone and the lower dense zone of the sediment, resulting in thickness measurement that can only obtain the total thickness and cannot quantify the density parameters.

[0005] Lack of dynamic adaptation capability: The detection point layout and detection angle adjustment of traditional methods are mostly preset fixed modes, which cannot be adjusted in real time according to the deformation of the borehole wall. For example, if the detection points are not densified in severely deformed areas, the sediment thickness data in that area will be missing. When the borehole wall vibration causes the probe angle to shift, there is no real-time compensation mechanism, which further amplifies the measurement error and makes it difficult to meet the data reliability requirements of high-precision engineering.

[0006] Large reference positioning deviation: The calculation of sediment thickness requires the bottom of the borehole wall as the reference. However, traditional methods often determine the bottom of the borehole by using a depth scale and manual estimation, or by fitting the bottom contour of the borehole based on a two-dimensional image of a single depth. This makes it impossible to construct a three-dimensional model of the entire depth of the borehole wall. Under irregular borehole shapes, the boundary of the bottom area is blurred, and the reference positioning deviation can reach 5-10mm, which directly leads to unreliable sediment thickness calculation results.

[0007] In summary, existing measurement methods cannot effectively solve the problem of accurate measurement of sediment thickness under irregular hole shapes. There is an urgent need for a measurement method that can adapt to hole wall deformation, eliminate interference, dynamically adjust detection parameters, and accurately position the reference, so as to meet the high-precision requirements of engineering for sediment thickness and density measurement. Summary of the Invention

[0008] The purpose of this invention is to provide an adaptive direction measurement method for the thickness of sediment at the bottom of irregular holes, in order to solve the technical problem in the prior art that it is impossible to accurately measure the thickness of sediment under irregular hole shapes.

[0009] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is as follows: An adaptive direction measurement method for bottom sediment thickness in irregularly shaped boreholes includes the following steps: S1: Acquire multiple sets of hole wall image data from multiple different orientations using an image acquisition device, and simultaneously acquire the vertical distance data between the image acquisition device and the hole wall during the acquisition of each set of data. S2: Extract the coordinates of feature points on the hole wall, correct the feature point coordinates to three-dimensional coordinates with the axis of the carrier unit as the origin, obtain three-dimensional feature points, use the three-dimensional feature points to fit and generate a dynamic three-dimensional cross-sectional contour model of the irregular hole, classify the hole wall deformation level, and determine the hole wall tilt angle in different directions inside the hole. S3: Prioritize setting up detection points in severely deformed areas and calculate the optimal detection angle for each detection point; set up multiple sets of sediment detection components at the bottom of the carrier unit, monitor the probe angular displacement in real time through a gyroscope, correct the angular deviation caused by the vibration of the hole wall, and drive the servo motor to adjust the orientation of the detection components so that the angle between the probe axis and the tangent direction of the bottom of the hole in the corresponding orientation is kept within the specified range. S4: Based on the deformation level of the borehole wall, control the sediment detection component to start multi-frequency detection and emit ultrasonic signals of the corresponding frequency, receive the signals reflected by the sediment surface, use wavelet transform algorithm to decompose the reflected signals into multiple scales, extract the features of the upper loose area and the lower dense area of ​​the sediment, correct the signal propagation speed, and calculate the detection distance from the probe to the upper and lower surfaces of the sediment respectively. S5: Based on the dynamic three-dimensional cross-sectional profile model, extract the reference coordinates of the bottom of the hole wall corresponding to each detection point and calculate the reference distance; subtract the reference distance of the same detection point from the detection distance of the upper surface of the sediment to obtain the total thickness of the sediment, and subtract the detection distances of the upper and lower surfaces of the sediment to obtain the thickness of the dense layer.

[0010] Preferably, the specific process for extracting the coordinates of feature points on the hole wall in step S2 is as follows: S211: Gaussian filtering algorithm is used on the hole wall image data to smooth the high-frequency interference signals generated by uneven light inside the hole and camera sensor noise in the image, while retaining key contour information including the hole wall edge and the water / bubble boundary. S212: Calculate the grayscale value of each pixel using a weighted average formula, and convert the denoised color image into a grayscale image; S213: To address the grayscale differences between water accumulation inside the hole, air bubbles, and the hole wall, the Otsu adaptive thresholding algorithm is employed. Traverse the grayscale values ​​from 0 to 255, calculate the inter-class variance between the hole wall region and the non-hole wall region under different thresholds, and select the grayscale value with the largest inter-class variance as the segmentation threshold; perform binarization processing on the grayscale image according to the determined threshold. S214: An edge detection algorithm with strong noise resistance is used to perform edge detection on the binarized image to extract the hole wall contour, and finally form a continuous and complete hole wall edge contour line to eliminate isolated water accumulation and bubble edges; S215: Establish a pixel coordinate system with the top left corner of the image as the origin, and extract the two-dimensional coordinates of all pixels on the edge contour of the hole wall. u , v ), forming an initial feature point set; calculating the feature point density, processing according to the feature point density to ensure that the feature point density of the entire hole wall meets the requirements, and sorting them in the circumferential direction to form an ordered feature point sequence {( u 1, v 1),( u 2, v 2),...,( u n , v n )}.

[0011] Preferably, in step S2, the feature point coordinates are corrected to three-dimensional coordinates with the carrier unit axis as the origin. The specific process for obtaining the three-dimensional feature points is as follows: S221: With the axis of the measuring device carrier unit as the Z-axis and the center of the cross-section of the carrier unit as the origin O, establish the X-axis and Y-axis within the cross-section to form a right-hand rule three-dimensional coordinate system O-XYZ; S222: Obtain information including focal length f Pixel size s Principal point coordinates ( u 0, v 0); S223: Coordinate transformation calculation: Transform each feature point ( uᵢ,vᵢ Convert to three-dimensional coordinates Xᵢ,Yᵢ,Zᵢ ); S224: Convert the imaging plane coordinates to three-dimensional spatial coordinates.

[0012] Preferably, the specific process of step S2 is as follows: S231: Group 3D feature points according to Z-axis coordinate: Set the layer interval Δ Z The Z-axis is divided into [...] Z 1, Z 1+Δ Z ]、[ Z 1+Δ Z , Z 1+2Δ Z The interval of ] will Zᵢ Feature points belonging to the same interval are grouped together to form a depth layer feature point group, and each group corresponds to the cross-section of a hole at a certain depth. S232: The B-spline curve fitting algorithm is adopted, and the fitting results are optimized by combining the hole wall features to finally obtain a smooth and accurate single-depth layer profile. S233: Arrange the optimized contours of all depth layers in Z-axis order, and supplement the contour details between adjacent depth layers through linear interpolation to form a dynamic three-dimensional cross-sectional contour model. S234: Classify the hole wall deformation level based on the degree of deviation between the hole wall profile in the model and the "standard circular profile"; S235: Calculate the angle between the tangent of the hole wall and the horizontal plane. The angle between the tangent of the hole wall and the horizontal plane is the inclination angle of the hole wall. S236: Determine the position and range of the hole bottom based on the contour shape changes in the Z-axis direction of the model.

[0013] Preferably, the specific process of step S3 is as follows: S31: Based on the hole wall deformation level output by the dynamic three-dimensional cross-sectional profile model, divide the detection area and determine the number of detection points in each area to ensure that the severely deformed area obtains a higher detection density. S32: Based on the inclination angle of the borehole wall and combined with the borehole wall morphology at the location of the detection point, calculate the optimal detection angle for each detection point; S33: Real-time angle compensation is achieved based on dynamic correction of gyroscope and electromagnetic damper to correct probe angular displacement deviation caused by vibration inside the hole; S34: Based on real-time angle compensation, the servo motor drives the sediment detection component to adjust its orientation so that the angle between the probe axis and the tangent direction at the bottom of the hole reaches the optimal detection angle.

[0014] Preferably, the specific process for extracting the characteristics of the upper loose layer and the lower dense layer of sediment in step S4 is as follows: S41: Select decomposition parameters to perform wavelet transform multi-scale decomposition, analyze the decomposition results. The first and second level detail components correspond to near-surface reflection, the third and fourth level detail components correspond to intermediate reflection, and the fifth level detail component corresponds to deep reflection. S42: Peak detection and feature determination of layered reflected waves: Peak detection is performed on the detailed components of each layer, and the upper loose area and the lower dense area of ​​the sediment are distinguished according to the peak voltage Vp.

[0015] Preferably, the specific process for calculating the detection distance from the probe to the surface of the upper loose zone and the lower dense zone of the sediment in step S4 is as follows: S43: Calculate the one-way distance based on the path of the detection signal from the probe to the sediment surface and back to the receiver; S44: Calculate the distance from the probe to the surface of the loose zone above the sediment. d 1; S45: Calculate the distance from the probe to the surface of the dense zone below the sediment. d 2 .

[0016] Preferably, the specific process of extracting the reference coordinates of the bottom of the borehole wall corresponding to each detection point and calculating the reference distance in step S5 based on the dynamic three-dimensional cross-sectional profile model is as follows: S51: From the dynamic 3D cross-sectional profile model, select the Z-axis coordinate and... Z p A completely consistent cross-sectional profile layer, i.e., the bottom depth layer of the borehole wall, is extracted from the profile layer with Z=505mm in the model to ensure that the subsequent coordinate extraction matches the depth of the detection point; S52: Perform probe point coordinate mapping: Given the plane coordinates of the probe point in the O-XY plane ( X p ,Y p );by( X p ,Y p Starting from the axis of the carrier unit, draw a ray in a radial direction away from the axis of the carrier unit. The first intersection of the ray with the contour curve of the depth layer at the bottom of the hole wall is defined as the reference point at the bottom of the hole wall. P 0 ; S53: Extract reference coordinates: The contour of each depth layer of the dynamic 3D model is generated by fitting the corrected 3D feature points, and each contour point stores accurate coordinates. X,Y,Z Spatial coordinates; read P 0 The coordinate data in the model are the reference coordinates of the bottom of the hole wall; S54: Calculate the reference distance to the bottom of the borehole wall: First, obtain the initial Z-axis coordinate of the probe, then correct it by combining the angle sensor data to obtain the corrected Z-axis coordinate. Based on the corrected coordinate of the probe and the reference coordinate of the bottom of the borehole wall, obtain the reference distance.

[0017] The beneficial effects of this invention include: 1. Adapting to Irregular Hole Shapes, Significantly Improving Measurement Accuracy: Dynamic 3D Model Construction Enables Precise Baseline Positioning: The constructed dynamic 3D cross-sectional contour model fully presents the irregular shape (protrusions, depressions, tilts) of the hole wall along the depth direction. Based on the base coordinates of the bottom of the hole wall extracted from the model, the baseline accuracy is effectively improved compared to traditional 2D baseline positioning. Simultaneously, the model supports real-time updates, dynamically adapting to changes in hole wall shape to ensure consistent baseline positioning across the entire depth range. Adaptive detection direction adjustment eliminates angular deviations, and closed-loop control through real-time gyroscope monitoring, electromagnetic damper compensation, and servo motor adjustment ensures the probe axis maintains a stable angle of 80°-100° with the tangent direction at the bottom of the hole. This completely solves the signal offset problem caused by hole wall tilt in traditional fixed-direction detection, effectively reducing thickness errors caused by angular deviations and significantly improving measurement accuracy under irregular hole shapes.

[0018] 2. Multi-technology integration eliminates interference and achieves quantification of sediment stratification: Multi-frequency detection + wavelet transform accurately distinguishes stratification boundaries: Step S4 adapts the detection frequency according to the deformation level of the borehole wall. High-frequency signals have high resolution and can identify thin, dense sediment layers; low-frequency signals have strong penetration and can penetrate thick, loose sediment layers; combined with the wavelet transform algorithm for multi-scale decomposition of the reflected signal, the characteristics of the upper loose zone and the lower dense zone of the sediment can be accurately extracted. Compared with traditional methods that can only measure the total thickness, this method achieves quantification of sediment stratification thickness and provides data support for density evaluation.

[0019] 3. Strong dynamic adaptability, comprehensive and reliable data coverage: The detection parameters are dynamically adapted to the deformation of the borehole wall. The detection frequency, detection point density, and detection angle are all adjusted in real time based on the actual deformation of the borehole wall. High-frequency signals are used to improve resolution in areas with slight deformation, while low-frequency signals are used to enhance penetration in areas with severe deformation. The angle deviation is compensated in real time when the borehole wall vibrates. Compared with traditional fixed parameter detection, the data adaptability is improved, ensuring that the measurement data in different deformation areas can reflect the true state of sediment.

[0020] 4. Data redundancy and anomaly removal ensure reliability: By removing outliers and calculating the average thickness using a weighted average, random errors are effectively eliminated; if there is insufficient valid data, re-detection is triggered to ensure data coverage; at the same time, the density level is divided by the ratio of the total sediment thickness to the dense layer thickness, providing a two-dimensional evaluation index of thickness and density for the project. Compared with the traditional method that only provides the total thickness, the amount of data information is increased, providing a precise basis for engineering treatment suggestions. Attached Figure Description

[0021] Figure 1 This is a flowchart illustrating the adaptive direction measurement method for bottom sediment thickness of irregularly shaped holes according to the present invention.

[0022] Figure 2This is a schematic diagram of the process for extracting the characteristics of the upper loose layer and the lower dense layer of sediment according to the present invention. Detailed Implementation

[0023] The following is in conjunction with the appendix Figure 1~Figure 2 The present invention will be further described in detail below: See appendix Figure 1 As shown, the adaptive direction measurement method for the bottom sediment thickness of irregularly shaped holes includes the following steps: S1: Acquire multiple sets of hole wall image data from multiple different orientations using an image acquisition device, and simultaneously acquire the vertical distance data between the image acquisition device and the hole wall during the acquisition of each set of data.

[0024] S2: Fusion processing of borehole wall image data and vertical distance data: An edge detection algorithm combined with grayscale thresholding is used to eliminate interference from water accumulation and air bubbles inside the borehole, extracting the coordinates of borehole wall feature points with a density ≥ 2 points / square centimeter. Based on the vertical distance data, a spatial coordinate transformation algorithm is used to correct the two-dimensional feature point coordinates to three-dimensional coordinates with the carrier unit axis as the origin, obtaining three-dimensional feature points. The corrected three-dimensional feature points are used to fit and generate a dynamic three-dimensional cross-sectional contour model of the irregular borehole. Based on this model, the borehole wall deformation level is classified into slight, moderate, and severe, and the borehole wall tilt angle and the preliminary range of the borehole bottom area in different directions are determined.

[0025] S3: Based on the inclination angle and deformation level of the borehole wall, perform adaptive adjustment of the detection direction: prioritize the deployment of detection points in severely deformed areas, accounting for ≥50% of the total detection points, and calculate the optimal detection angle for each detection point; deploy at least 4 sets of sediment detection components at the bottom of the carrier unit, monitor the probe angular displacement in real time through a gyroscope, and combine the output compensation torque of the electromagnetic damper to correct the angular deviation caused by borehole wall vibration, drive the servo motor to adjust the orientation of the detection components, and keep the angle between the probe axis and the corresponding orientation of the bottom tangent of the borehole stably maintained at 80°-100°.

[0026] S4: According to the defined hole wall deformation levels, control the sediment detection component to start multi-frequency detection: 500kHz-1MHz high-frequency ultrasonic detection signals are emitted in slightly deformed areas, and 100kHz-300kHz low-frequency detection signals are emitted in severely deformed areas; in moderately deformed areas: a high-low frequency mixed detection strategy is adopted, that is, 300kHz-500kHz medium-frequency ultrasonic detection signals are emitted first for the same detection point, and the signal reflection intensity is dynamically adjusted. If the reflected signal is clear, i.e., the signal-to-noise ratio is ≥25dB, the medium frequency is maintained; if the signal attenuation is severe, i.e., the signal-to-noise ratio is <25dB, the 100kHz-300kHz low-frequency ultrasonic detection signal is switched; after receiving the signal reflected from the sediment surface, the wavelet transform algorithm is used to decompose the reflected signal into multiple scales, extract the characteristics of the upper loose area (reflection wave peak value <0.3V) and the lower dense area (reflection wave peak value >0.5V) of the sediment, and combine the real-time temperature inside the hole (temperature sensor acquisition frequency ≥1Hz) to correct the signal propagation speed, and calculate the detection distance from the probe to the upper and lower surfaces of the sediment respectively.

[0027] S5: Based on the dynamic three-dimensional cross-sectional profile model, extract the reference coordinates of the bottom of the hole wall corresponding to each detection point and calculate the reference distance; subtract the reference distance of the same detection point from the detection distance of the upper surface of the sediment to obtain the total thickness of the sediment, and subtract the detection distances of the upper and lower surfaces of the sediment to obtain the thickness of the dense layer.

[0028] Example 2 Based on Example 1, the specific process of extracting the coordinates of feature points on the hole wall in step S2 using an edge detection algorithm combined with grayscale thresholding to eliminate interference from water accumulation and air bubbles inside the hole is as follows: S211: Noise Reduction Processing: Gaussian filtering algorithm is used on the hole wall image data to smooth the high-frequency interference signals generated by uneven light inside the hole and camera sensor noise in the image, and retain key contour information such as hole wall edge and water / bubble boundary to avoid subsequent algorithms misidentifying noise as feature points.

[0029] S212: Image Grayscale Conversion: Convert the denoised color image (RGB format) into an 8-bit grayscale image (pixel grayscale value range 0-255): Calculate the grayscale value of each pixel using the weighted average formula Gray=0.299R+0.587G+0.114B to eliminate the interference of color information on edge detection and unify the processing dimensions of subsequent algorithms.

[0030] S213: To address the grayscale differences between water accumulation inside the hole (grayscale value typically 50-100, appearing light gray due to reflected light), air bubbles (grayscale value 150-200, appearing bright white), and the hole wall (grayscale value 20-60, appearing dark gray due to material differences), the Otsu adaptive thresholding algorithm is employed. Traverse the grayscale values ​​from 0 to 255, calculate the inter-class variance between the hole wall region and the non-hole wall region (water accumulation, air bubbles, air) under different thresholds, and select the grayscale value with the largest inter-class variance as the segmentation threshold, which is usually 70-90, and can be dynamically adjusted according to the actual light intensity inside the hole.

[0031] The grayscale image is binarized according to the segmentation threshold: if the pixel grayscale value is less than or equal to the segmentation threshold, it is determined to be a candidate region for the hole wall and is assigned a value of 0 (black); if the pixel grayscale value is greater than the segmentation threshold, it is determined to be an interference / background region, including water accumulation, bubbles, and air, and is assigned a value of 255 (white). At this point, the image is initially separated into a black and white binary image, and the hole wall region is presented in the form of black blocks, which narrows the processing range for edge detection.

[0032] S214: An edge detection algorithm is used to perform edge detection on the binarized image to extract the hole wall contour. Step 1: Calculate the gradient: Calculate the gradient magnitude and direction of each pixel using the Sobel operator (x and y directions), capture pixels with abrupt changes in grayscale value within the hole wall region, i.e., potential edge points. The gradient direction is perpendicular to the edge tangent, providing a directional basis for subsequent edge connections.

[0033] Step 2: Non-maximum suppression: Scan the gradient image. If the gradient magnitude of a pixel is not a local maximum in its gradient direction, such as if the gradient magnitude of an adjacent pixel is larger, then the pixel is determined to be a non-edge point and suppressed, and its value is set to 0. Only the local maximum points in the gradient direction are retained to achieve edge refinement and avoid the feature point redundancy caused by thick edges.

[0034] Step 3: Dual-threshold detection and edge connection: Set a high threshold (usually 70% of the maximum gradient) and a low threshold (usually 30% of the maximum gradient): Pixels with gradient magnitudes greater than the high threshold are directly identified as strong edge points and thus determined to be hole wall edges. Pixels with gradient magnitudes between the low and high thresholds are identified as weak edge points if they are connected to strong edge points, and are added as hole wall edges. Pixels with gradient magnitude less than the low threshold are identified as non-edge points and removed. This process ultimately forms a continuous and complete contour line of the hole wall edge, eliminating isolated water accumulation and bubble edges (because water accumulation and bubble edges lack continuous strong edge point support, they will be judged as non-edge points).

[0035] S215: Extraction of feature point coordinates on the hole wall: Coordinate extraction: Establish a pixel coordinate system with the top left corner of the image as the origin, and extract the two-dimensional coordinates of all pixels on the contour of the hole wall. u , v ), u In the horizontal direction,v The initial feature point set is formed in the vertical direction; Density screening and supplementation: Calculate the feature point density (number of feature points per unit area). If the density in a certain area is <2 points / square centimeter, supplement feature points according to the uniform interpolation principle, such as inserting 1-2 points between adjacent points; if the density is too high, >5 points / square centimeter, eliminate redundant points by sampling at equal intervals, finally ensuring that the feature point density of the entire borehole wall meets the requirements, and sort them in the circumferential direction to form an ordered feature point sequence {( u 1, v 1),( u 2, v 2),...,( u n , v n )}.

[0036] In step S2, the feature point coordinates are corrected to three-dimensional coordinates with the carrier unit axis as the origin. The specific process of obtaining the three-dimensional feature points is based on the vertical distance data, mapping the two-dimensional pixel coordinates to three-dimensional coordinates with the carrier unit axis as the origin, thereby realizing the fusion of image features and spatial position. The specific process is as follows: S221: Define a three-dimensional coordinate system: Using the axis of the measuring device carrier unit (the central symmetry axis of the cylindrical structure) as the Z-axis (parallel to the hole axis, downwards is positive), and the center of the carrier unit's cross-section (perpendicular to the Z-axis) as the origin O, establish an X-axis (horizontally to the right) and a Y-axis (vertically forward) within the cross-section, forming a right-hand rule three-dimensional coordinate system O-XYZ. Ensure that all coordinates are based on the carrier unit and reflect the spatial position of the hole wall relative to the carrier. S222: Input key parameters: Perform internal parameter calibration on the camera in advance to obtain the following parameters: focal length f : The unit is mm, representing the distance from the optical center of the camera to the imaging plane, such as f =8mm; Pixel size s The unit is mm / pixel. It refers to the physical size of a single pixel on the imaging plane, such as s = 0.003 mm / pixel. Principal point coordinates ( u 0, v 0): The intersection of the camera's optical axis and the imaging plane, usually the center of the image, such as the center of a 1920×1080 image. u 0, v 0) = (960, 540); S223: Coordinate transformation calculation: For each feature point ( uᵢ,vᵢ ), combined with its corresponding vertical distance dᵢ Convert to three-dimensional coordinates using the following steps ( Xᵢ,Yᵢ,Zᵢ ): First, convert the pixel coordinates to physical coordinates of the imaging plane: The physical coordinates (o-xy) of the imaging plane take the principal point as the origin, and the x and y axes are parallel to the u and v axes of the pixel coordinate system, respectively. The transformation formula is: x ᵢ=( u ᵢ- u 0)× s ; y ᵢ=( v ᵢ- v 0)× s ; If the pixel coordinates are in the opposite direction to the physical coordinates, a negative sign needs to be added for adjustment, such as... y ᵢ=-( v ᵢ- v 0)× s ; S224: Convert the physical coordinates of the imaging plane to three-dimensional spatial coordinates: Laser ranging data dᵢ The radial distance from the feature point to the axis of the carrier unit is the distance from the feature point projected onto point O in the O-XY plane. The camera optical axis extends radially outward (at an angle with the axis). i =0°), the conversion formula is: X ᵢ= d ᵢ×(xᵢ / ( x ᵢ²+ y ᵢ²+ f ²) 1 / 2 ),Right now X Axial components reflect horizontal position; Y ᵢ= d ᵢ×(yᵢ / ( x ᵢ²+ y ᵢ²+ f ²) 1 / 2 ),Right now Y The axis component reflects the position in the vertical forward direction; Z ᵢ= Z 0+( v ᵢ- v 0)× k ,Right now Z Axial components reflect depth position; Z 0 represents the hole depth corresponding to the main point. k The vertical-to-depth conversion factor for pixels is calibrated by the camera mounting angle, such as... k =0.5mm / pixel.

[0037] After conversion, the correction error is ≤0.1mm, ensuring that the three-dimensional coordinates are consistent with the actual position of the hole wall.

[0038] In step S2, the corrected three-dimensional feature points are used to fit and generate a dynamic three-dimensional cross-sectional profile model of the irregular hole. Based on this model, the hole wall deformation level is divided into slight, moderate, and severe, and the hole wall inclination angle and the preliminary range of the hole bottom area in different orientations are determined. The specific process is as follows: S231: Three-dimensional feature point layering: Grouping 3D feature points by Z-axis coordinate (depth): Set the layer interval Δ Z Δ Z The range is 1-5mm, which can be adjusted according to accuracy requirements. It can be set to ΔZ=2mm, dividing the Z-axis into [ Z 1, Z 1+Δ Z ]、[ Z 1+Δ Z , Z 1+2Δ Z ] and other intervals; Will Zᵢ Feature points belonging to the same interval are grouped together to form a depth layer feature point group. Each group corresponds to the cross-section of a hole at a certain depth. For example, the feature point group with a depth of Z=500-502mm corresponds to the cross-sectional profile at that depth.

[0039] S232: Perform single-depth-layer contour fitting: The B-spline curve fitting algorithm was used, and the fitting results were optimized by combining the hole wall features: Initial Fitting: Based on the feature points (X,Y) within the depth layer, a 3rd-order B-spline curve is defined. The curve control points are calculated using the least squares method to minimize the sum of squared distances from the feature points to the curve, thus generating the initial cross-sectional profile curve. C k ( X,Y ), k This is the depth layer number.

[0040] Optimization and adjustment: If feature points are dense in a certain area, such as the protrusion of the hole wall, assign a higher weight to the feature points in that area, which can be 1.5-2.0, and refit to ensure that the curve fits the shape of the protrusion; remove abnormal feature points with fitting error >0.2mm, recalculate the control points, and finally obtain a smooth and accurate single-depth layer profile.

[0041] S233: Model Integration and Dynamic Feature Implementation: Model integration: The optimized contours of all depth layers are arranged in Z-axis order, and the contour details between adjacent depth layers are supplemented by linear interpolation, such as the transition shape between the contours of depth Z=500-502mm and Z=502-504mm, forming a complete dynamic three-dimensional cross-sectional contour model, which can present features such as hole wall tilt, protrusion, and depression through three-dimensional visualization.

[0042] Dynamic updates: If new feature points are subsequently collected (such as the discovery of unidentified depressions) or distance data is updated (due to slight movement of the carrier), only the feature point group of the corresponding depth layer is refitted, without the need to reconstruct the entire model. The update response time is ≤100ms, ensuring that the model always matches the actual shape of the hole.

[0043] S234: Classification of Hole Wall Deformation Levels (Slight / Moderate / Severe): Classification based on the degree of deviation between the hole wall profile in the model and the standard circular profile: Standard circular profile fitting: For the cross-sectional profile of each depth layer, fit an ideal circle with the same area and radius. R 0 = ( S / π ) 1 / 2 S is the area enclosed by the outline; Deviation calculation: Calculate the radial deviation Δ of each point on the profile from the ideal circle. rᵢ =| rᵢ-R 0 |, rᵢ Given the distance from the contour point to the origin O, calculate the maximum deviation Δ. r max With average deviation Δ r avg ; Level determination: Slight deformation: Δr max ≤1mm and Δ r avg ≤0.5mm, hole wall is nearly circular, with minor local protrusions / depressions; Moderate deformation: 1mm < Δ r max ≤3mm and 0.5mm<Δ r avg ≤1.5mm, the hole wall has obvious protrusions / depressions, but does not affect the overall shape; Severe deformation: Δ r max >3mm or Δ r avg >1.5mm, the hole wall has significant protrusions / depressions, such as localized narrowing or widening.

[0044] S235: Calculate the angle between the tangent of the hole wall and the horizontal plane. The angle between the tangent of the hole wall and the horizontal plane is the inclination angle of the hole wall. The inclination angle of the borehole wall refers to the angle between the tangent of the borehole wall and the horizontal plane (O-XY plane). The calculation process is as follows: Feature points on the contour of each depth layer are selected in the model. One point can be selected every 30°. For each feature point, the tangent direction vector of the contour at that point is calculated by the derivative of the B-spline curve. Calculate the angle between the tangent direction vector and the horizontal plane (X-axis direction); this is the inclination angle of the hole wall in that orientation. oh , oh =15° indicates that the borehole wall is tilted upwards by 15° in that azimuth direction; the tilt angles of each azimuth direction are counted in a circular direction to form a distribution table of borehole wall tilt angles, which provides a basis for subsequent adjustment of the detection direction.

[0045] S236: Preliminary determination of the area at the bottom of the borehole: The location and extent of the hole bottom are determined based on the contour morphology changes along the Z-axis in the model. Bottom feature identification: The contour features of the bottom region are that the depth in the Z-axis direction no longer increases, and the contour area tends to be stable. The surface of the sediment is usually relatively flat, and the contour area change rate is <5%. Depth range determination: Find the depth where the rate of change of the contour area in the Z-axis direction is first less than 5%. Z bottom , let Z∈[ Z bottom , Z bottom The area within +10mm is defined as the bottom area of ​​the hole. Determining the horizontal range: Extract the contour of the depth layer at the bottom of the borehole, and the horizontal area enclosed by it (within the O-XY plane) is the preliminary range of the bottom of the borehole. The detection points of the subsequent sediment detection components are preferentially deployed within this range.

[0046] Example 3 Based on Example 1 or Example 2, the specific process of step S3 is as follows: S31: Zonal layout of detection points (priority allocation based on borehole wall deformation level): First, based on the hole wall deformation level (slight / moderate / severe) output by the dynamic 3D cross-sectional profile model, the detection area is divided and the number of detection points in each area is determined to ensure a higher detection density in severely deformed areas. The specific operation is as follows: Deformation region division and priority ranking: Region division: Centered on the axis of the carrier unit (Z axis of the O-XYZ coordinate system), the initial range of the bottom area of ​​the hole is divided into multiple fan-shaped regions in the circumferential direction, such as 8 regions, each region at 45°. Based on the deformation level judgment results of each depth layer in step S2, the deformation level of each fan-shaped region is marked, such as region 1 being severely deformed, region 2 being moderately deformed, and region 3 being slightly deformed.

[0047] Priority setting: Set the detection priority according to the deformation level. The priority from high to low is: severe deformation area > moderate deformation area > slight deformation area. Due to the complex morphology of the borehole wall in the severely deformed area, such as large bulges / depressions, the thickness of the sediment may vary significantly. Detection points should be set up in the severely deformed area first to ensure the representativeness of the measurement.

[0048] Allocation and location determination of detection points: Quantity allocation rules: Let the total number of detection points be N, where N≥4, consistent with the number of sediment detection components, and allocate them according to the following ratio: Severely deformed areas: The number of detection points N1 ≥ 50% N, such as when N=4, N1 ≥ 2; when N=6, N1 ≥ 3; Moderately deformed areas: the number of detection points N2 = 30%N~40%N, supplementing and covering areas that are not severely deformed but have complex shapes; For areas with slight deformation: the number of detection points N3 = 10%N~20%N, only a small number of detection points are needed to verify the thickness uniformity; If a certain deformation level area does not exist, such as a severely deformed area, the detection point quota for that area will be allocated to the next highest priority area, such as transferring the N1 quota to N2.

[0049] Location determination method: Severely deformed areas: Detection points are placed at equal angular intervals within the area. For example, two detection points are placed in a 45° fan-shaped area with an interval of 22.5° to ensure coverage of all key protrusions / depressions within the area. Moderate / slightly deformed areas: One detection point is set up at the center and one at the edge of the area to measure both the overall thickness of the area and the local thickness. Finally, the planar coordinates of all detection points are formed. X p ,Y p (based on the O-XY coordinate system), and corresponds one-to-one with the sediment detection components; S32: Calculation of the optimal detection angle for each detection point (based on the borehole wall inclination angle): Based on the borehole wall inclination angle and combined with the borehole wall morphology at the detection point location, the optimal detection angle for each detection point is calculated, i.e., the angle at which the probe axis should be adjusted to ensure an angle of 80°-100° with the tangent direction at the bottom of the borehole. The specific calculation process is as follows: Hole wall tilt angle extraction and mapping: Extract each detection point from the table of borehole wall tilt angle distribution. X p ,Y p The inclination angle of the borehole wall in the corresponding orientation α p , α p This refers to the angle between the tangent of the borehole wall at the detection point and the O-XY plane (horizontal plane), such as... α p = 20° means the hole wall is inclined upwards at a 20° angle. α p =-15° indicates that the hole wall is tilted downwards by 15°.

[0050] The plane coordinates of the detection point ( X p ,Y p Mapping the probe point to the depth direction (Z-axis) of the borehole wall determines the borehole bottom depth Z corresponding to that probe point. p , and the depth range of the bottom region of the hole [ Z bottom , Z bottom [+10mm] consistent, ensuring extraction α p This indicates the true tilt angle at the bottom of the hole, avoiding misreading of the angle due to depth deviation.

[0051] The optimal detection angle is calculated as follows: Optimal detection angle θ p The calculation objective is to make the angle θ between the probe axis of the sediment detection component and the tangent direction at the bottom of the hole at the detection point. p ∈[80°,100°], specifically calculated in two scenarios: Scenario 1: The borehole wall is inclined towards the center of the borehole (inward, such as in a narrowing area): If α p When the value is positive, the borehole wall slopes inwards and upwards, and the tangent at the bottom of the borehole slopes upwards at α. p At this point, the probe axis needs to be adjusted downwards to match the tangential direction: Optimal detection angle i p =90°- α p +Δ i Δ i The compensation coefficient is set at 5°-10° and adjusted according to the flatness of the hole wall: Δθ is set to 10° when the flatness is poor and 5° when the flatness is good. for example: α p=10°, Δθ=5°, then i p =90°-10°+5°=85°, which falls within the range of 80°-100°.

[0052] Scenario 2: The borehole wall is inclined in a direction away from the borehole center (outward inclination, such as in the enlarged diameter area): like α p A negative value (hole wall slopes downwards and outwards), the tangent at the bottom of the hole slopes downwards |α p At this point, the probe axis needs to be adjusted upwards: Optimal detection angle i p =90°+| α p |-Δ i ; For example: α p =-15°, Δθ=8°, then θ p =90°+15°-8°=97°, which falls within the range of 80°-100°.

[0053] Angular validity verification: Calculated i p Afterwards, it needs to be verified whether it is within the mechanical adjustment range. Due to the structural limitations of the sediment detection component, the adjustment range is usually 0°-180°. like i p ∈[80°,100°] and within the mechanical range, it is directly taken as the optimal angle for the detection point; like i p Exceeding 80°-100°, such as α p =25°, θ is calculated. p =70°, then adjust Δθ, for example, increase Δθ from 5° to 15°, and recalculate until θ is equal to 70°. p ∈[80°,100°]; If Δ is adjusted i If the above calculation is still not satisfied, then a backup location near the detection point is selected again, and the above calculation is repeated to ensure that all detection points have an effective optimal angle. S33: Real-time angle compensation (based on dynamic correction using gyroscope and electromagnetic damper): Before adjusting the angle of the servo motor-driven detection component, a dynamic compensation system consisting of a gyroscope and an electromagnetic damper is needed to correct the probe angular displacement deviation caused by vibrations within the hole (such as slight shaking of the carrier unit or dust impact on the hole wall) to ensure adjustment accuracy. The specific process is as follows: Real-time monitoring of angular displacement: A miniature gyroscope is installed on the probe bracket of each sediment detection component, with a sampling frequency ≥100Hz, to collect real-time angular displacement data of the probe along the X-axis (horizontal direction) and Y-axis (vertical forward direction), denoted as Δθ. x Δθᵧ), angular displacement refers to the deviation between the current angle of the probe and the initial calibration angle, such as Δθ x =0.5° indicates that the probe deviates from the initial angle by 0.5° in the X-axis direction.

[0054] The angular displacement data acquired by the gyroscope is filtered to remove high-frequency noise, including spurious displacement data caused by instantaneous vibration, to obtain the true angular displacement deviation value (Δθ). x '、Δθᵧ').

[0055] Compensation torque calculation and output: The control chip calculates the required compensation torque M based on the actual angular displacement deviation using the following formula: M = k ×(Δ i x '² +Δ θᵧ'² ) 1 / 2 ; in k This is the torque coefficient, determined based on the probe mass and support stiffness, typically... k =0.1 N·m / °, (Δ i x '² +Δ i ᵧ'² ) 1 / 2 The composite angular displacement deviation reflects the overall degree of probe offset; For example: Δ i x ' =0.3°, Δ θᵧ'= 0.4°, k=0.1N・m / °, then the combined deviation=0.5°, M=0.1×0.5=0.05N・m.

[0056] The control chip sends control commands to the electromagnetic damper, which outputs a corresponding compensation torque M according to the command. This torque acts on the probe support and offsets the angular displacement caused by vibration through torque balance. For example, if the probe shifts to the right by 0.5°, a torque of 0.05 N·m is output to the left to pull the probe back to the initial angle, so that the angular displacement deviation of the probe is stably controlled within ±0.1°.

[0057] S34: Servo motor drive adjustment (to achieve precise probe positioning): Based on real-time angle compensation, the orientation of the sediment detection component is adjusted by driving a servo motor, so that the angle between the probe axis and the tangent direction at the bottom of the hole reaches the optimal detection angle. i p The specific steps are as follows: Initial angle calibration: Before adjustment, adjust the probe axes of all sediment detection components to the initial reference angle, which is usually vertically downward, parallel to the Z-axis, corresponding to an angle of 0°. Record the initial angle value using the angle sensor. i 0 , as the adjustment benchmark.

[0058] Adjustment calculation: For each detection point, based on the optimal detection angle... i p With initial angle i 0 Calculate the adjustment amount Δ of the servo motor. i reg : Δ i reg = i p -θ 0 -Δ i comp ; Where, Δ i comp The compensation angle output by the real-time angle compensation system is calculated from the angular displacement deviation, such as Δ. i x ' =0.3° corresponds to Δ i comp =0.3°, used to compensate for the angle deviation caused by the current vibration, ensuring that the adjustment accurately reflects the "difference between the target angle and the actual angle".

[0059] for example: i p =85°, i 0 =0°, Δ i comp =0.3°, then Δ i reg =85°-0°-0.3°=84.7°.

[0060] Servo motor drive and angle closed-loop control: The control chip sends adjustment commands to the servo motor, which drives the probe bracket to rotate via a reduction gear set, according to the adjustment amount Δ. i reg Adjust the probe's orientation; During the adjustment process, the angle sensor collects the current angle of the probe in real time. icurr Feedback is sent to the control chip to form a closed-loop control: If | i curr -θ p If the angle is greater than 0.1°, the target angle has not been reached. The control chip adjusts the servo motor speed. If the difference is large, the motor speeds up; if the difference is small, the motor speeds down. The adjustment continues. If | i curr -θ p If the angle is ≤0.1°, the target angle is reached, the control chip sends a stop command, the servo motor locks its current position, and the probe axis is kept stably at the optimal detection angle. i p ; The above adjustment process was performed on each sediment detection component corresponding to all detection points. After completion, the angular displacement deviation of all probes was checked again by gyroscope to ensure that the deviation was ≤ ±0.1° and that the angle between the probe axis and the tangent direction at the bottom of the hole was within the range of 80°-100°.

[0061] Example 4 Based on Example 1, Example 2, or Example 3, see [link to example]. Figure 2 As shown, the specific process of using wavelet transform algorithm to perform multi-scale decomposition of the reflected signal and extract the features of the upper loose layer and lower dense layer of sediment in step S4 is as follows: S41: Wavelet transform multi-scale decomposition: Decomposition parameter selection: The db6 wavelet basis is selected to perform 5-level multi-scale decomposition of the reflection signal. The db6 wavelet basis has good temporal locality and frequency domain resolution, and can effectively distinguish reflection pulses at different depths. The time interval between the reflection pulses in the upper loose zone and the reflection pulses in the lower dense zone may be only 1μs, corresponding to a sediment thickness difference of 0.17mm. The 5-level decomposition can decompose the signal into one approximate component (low frequency, corresponding to the overall trend) and five detail components (high frequency, corresponding to local reflection characteristics).

[0062] Decomposition results analysis: Detail components at different scales correspond to reflection information at different depths. The first and second layers of detail components (high frequency) correspond to near-surface reflection, such as the surface of the loose zone in the upper layer of sediment, with a depth <100mm; the third and fourth layers of detail components (mid frequency) correspond to mid-level reflection, such as the interface between the loose and dense zones, with a depth of 100-300mm; the fifth layer of detail components (low frequency) correspond to deep reflection, such as the surface of the dense zone, with a depth >300mm; and the approximate components correspond to the DC offset of the signal.

[0063] S42: Detection and Characterization of Layered Reflection Wave Peak Values: Peak detection: Peak detection is performed on each layer of detail components using a sliding window method with a window width of 5 sampling points. The voltage values ​​of detail components are traversed. When the voltage value of a certain sampling point is greater than that of all other sampling points in the window and is greater than a set threshold, such as 0.05V, which is 5 times the noise voltage, it is determined to be the peak point of the reflected wave. Its peak voltage Vp and the corresponding time t (time relative to the signal transmission time) are recorded.

[0064] Layering characteristics determination: The upper loose layer and the lower dense layer of sediment are distinguished based on the peak voltage Vp. If Vp < 0.3V: it is determined to be a reflected wave from the loose upper layer of sediment. The gaps between sediment particles in the loose zone are large, and the detection signal is scattered multiple times between the particles, resulting in severe energy attenuation and a low peak value of the reflected wave. At the same time, the time t1 (the reflection time of the upper surface) corresponding to this peak value is recorded.

[0065] If Vp > 0.5V: it is determined to be a reflected wave from the dense zone of the lower sediment layer. The sediment particles in the dense zone are compact, and the detection signal mainly undergoes specular reflection, resulting in small energy attenuation and a high peak value of the reflected wave. At the same time, the time t2 (the time of reflection from the lower surface) corresponding to this peak value is recorded.

[0066] If 0.3V≤Vp≤0.5V: it is determined to be a reflection wave in the transition zone, where the loose and dense zones are mixed and not included in the layered calculation. It is necessary to readjust the number of wavelet decomposition layers (e.g., increase to 6 layers) or the signal transmission parameters (e.g., increase the peak ultrasonic voltage to 8V), and re-acquire the signal for detection.

[0067] Step S4 involves detecting the signal propagation speed. v With the time of layered reflection t 1 、t 2 The detection distance from the probe to the surface of the upper loose zone and the lower dense zone of the sediment is calculated. The specific process is as follows: S43: Based on the fact that the detection signal is emitted from the probe, reflected from the sediment surface, and returned to the receiver, the propagation path is probe → sediment surface → probe, which is the round trip distance. Therefore, the formula for calculating the one-way distance (probe to sediment surface) is: d =( v × t ) / 2, unit: mm t The time difference between the reflection time and the emission time is expressed in seconds (s).

[0068] S44: Calculate the distance from the probe to the surface of the loose zone above the sediment. d 1: Will t=t 1 (Upper layer reflection time) v Substituting into the formula, we get d 1=( v ×t 1) / 2.

[0069] For example: ultrasonic detection. v =1481m / s=1.481×10 6 mm / s, t 1=100 μs =1×10⁻ 4 s ; but d 1 = (1.481 × 10 6 ×1×10⁻ 4 ) / 2 = 148.1 / 2 = 74.05mm ≈ 74.1mm; S45: Distance from the probe to the surface of the dense zone below the sediment layer d 2 :Will t=t 2 (Lower layer reflection time) v Substituting into the formula, we get d 2=( v × t 2) / 2.

[0070] For example: Same as above, t 2=300 μs =3×10⁻ 4 s , but d 2 = (1.481 × 10 6 ×3×10⁻ 4 ) / 2=444.3 / 2=222.15mm≈222.2mm.

[0071] In step S5, based on the dynamic three-dimensional cross-sectional profile model, the specific process of extracting the reference coordinates of the bottom of the borehole wall corresponding to each detection point and calculating the reference distance is as follows: S51: Locking the Calculation Baseline: In step S4, each detection point of the sediment detection component corresponds to a specific depth Z. p This falls within the depth range of the borehole bottom region determined in step S2. Z bottom , Z bottom +10mm], such as Z bottom When = 500mm, Z p ∈[500,510]mm; From the dynamic 3D cross-sectional profile model, select the Z-axis coordinate and Z... p A completely consistent cross-sectional profile layer, i.e., the bottom depth layer of the borehole wall, is a digital representation of the borehole wall morphology at the depth of the detection point, for example...Z p When Z=505mm, extract the contour layer from the model to ensure that the subsequent coordinate extraction matches the depth of the detection point.

[0072] S52: Perform probe point coordinate mapping: Know the plane coordinates of the probe point in the O-XY plane (perpendicular to the axis of the carrier unit) ( X p ,Y p The layout is determined by step S3. For example, when the four sets of probes are evenly distributed around the circumference, the coordinates of probe 1 are (0,50) mm. by( X p ,Y p Starting from the axis of the carrier unit, draw a ray in the radial direction away from the axis of the borehole (i.e., the direction of the borehole wall). The first intersection of the ray with the contour curve of the depth layer at the bottom of the borehole wall is defined as the reference point at the bottom of the borehole wall. P 0 , P 0 It is the theoretical boundary between the borehole wall and the sediment at the bottom of the borehole, and also the source of the reference coordinates; S53: Extract reference coordinates: The contour of each depth layer of the dynamic 3D model is generated by fitting the 3D feature points corrected in step S2, and each contour point stores accurate coordinates. X,Y,Z Spatial coordinates.

[0073] Directly reading the coordinate data of P0 in the model yields the reference coordinates of the bottom of the hole wall. X 0 ,Y 0 ,Z 0 ),in: Z 0 =Z p Consistent with the depth of the detection point, such as Z p =505mm Z 0 =505mm; X 0 、Y 0 Let P0 be the actual coordinates in the O-XY plane, as shown in the model. P 0 For (10, 55) mm, then X 0 =10mm, Y 0 =55mm; S54: Calculation of reference distance at the bottom of the hole wall: Determination and dynamic correction of probe spatial coordinates: Initial coordinates obtained: The probe coordinates are based on the axis of the carrier unit (origin of the O-XYZ coordinate system) and have been calibrated before leaving the factory. X t 、Y t The value is determined by the probe's mounting position at the bottom of the carrier unit and is a fixed value. For example, if probe 1 is installed at (0,50) mm, then... X t =0mm, Y t =50mm; Z t The vertical distance from the bottom of the carrier unit to the probe is a fixed value, such as... Z t =20mm, which is the initial Z-axis coordinate of the probe; Dynamic coordinate correction (caused by angle adjustment): In step S3, the probe is adjusted due to tilt (e.g., θ). p A displacement of 10° will occur along the Z-axis, which needs to be corrected using angle sensor data. Displacement deviation ΔZ t = L×sinθ p L is the length of the probe bracket. For example, if L = 50mm, the corrected Z-axis coordinate is... Z t '=Z t +Δ Z t .

[0074] for example: i p When =10°, Δ Z t =50×sin10°≈8.68mm, Z t ' =20+8.68≈28.68mm, ensuring that the coordinates reflect the actual position of the probe; The reference distance D0 is the straight-line distance from the probe to P0, based on the probe's corrected coordinates. X t ,Y t ,Z t ' ) and reference coordinates ( X 0,Y 0 ,Z 0 Substitute into the formula: D 0=[( X 0- X t )^2+( Y 0- Y t )^2+( Z 0- Z t ')^2)] 1 / 2 ; For example: probe coordinates (0, 50, 28.68) mm, reference coordinates (10, 55, 505) mm, calculated as follows: D 0 = [(10-0)^2 + (55-50)^2 + (505-28.68)^2] 1 / 2 ≈476.43mm.

[0075] Based on the layered detection distance and the baseline distance for this stage, the total thickness of the sediment and the thickness of the dense layer are calculated using the difference: Total thickness Ht otal = D 0− d 1 = 476.43 - 74.1 = 402.33 mm; Thickness of the dense sediment layer H dense = d 2− d 1 = 222.2 - 74.1 = 148.1 mm.

Claims

1. A method for adaptive directional measurement of the thickness of the bottom sediment of an irregularly shaped hole, characterized in that, The method comprises the following steps: S1: acquiring a plurality of groups of hole wall image data of an irregular hole to be measured from different directions by an image acquisition device, and simultaneously acquiring vertical distance data between the image acquisition device and the hole wall during each group of data acquisition; S2: extracting hole wall feature point coordinates, correcting the feature point coordinates to three-dimensional coordinates with the carrier unit axis as the origin, obtaining three-dimensional feature points, generating a dynamic three-dimensional cross-sectional profile model of the irregular hole by fitting the three-dimensional feature points, dividing the hole wall deformation level, and determining the hole wall inclination angle at different directions in the hole; S3: preferentially arranging detection points in the severely deformed area, calculating the optimal detection angle of each detection point, arranging a plurality of groups of sediment detection assemblies at the bottom of the carrier unit, monitoring the angular displacement of the probe by a gyroscope in real time, correcting the angular deviation caused by the hole wall vibration, driving a servo motor to adjust the direction of the detection assembly, and keeping the included angle between the probe axis and the corresponding direction hole bottom tangent direction within a specified range; S4: according to the hole wall deformation level, controlling the sediment detection assembly to start multi-frequency detection to emit ultrasonic signals of corresponding frequencies, receiving the reflected signals on the sediment surface, performing multi-scale decomposition on the reflected signals by a wavelet transform algorithm, extracting the features of the upper loose area and the lower dense area of the sediment, correcting the signal propagation speed, and respectively calculating the detection distances from the probe to the upper and lower surfaces of the sediment; S5: based on the dynamic three-dimensional cross-sectional profile model, extracting the hole wall bottom reference coordinates corresponding to each detection point and calculating the reference distance; subtracting the reference distance from the upper surface detection distance of the same detection point to obtain the total thickness of the sediment, and subtracting the upper and lower surface detection distances of the sediment to obtain the dense layer thickness.

2. The method for adaptive directional measurement of hole bottom sediment thickness for irregular hole shape according to claim 1, characterized in that, The specific process of extracting the hole wall feature point coordinates in step S2 is as follows: S211: Gaussian filtering algorithm is used on the hole wall image data to smooth the high-frequency interference signals in the image caused by uneven light in the hole and camera sensor noise, and the key contour information including the hole wall edge, accumulated water and bubble boundary is retained; S212: the gray value of each pixel is calculated by a weighted average formula, and the denoised color image is converted into a gray image; S213: according to the gray difference between the accumulated water, bubbles and hole wall, an Otsu adaptive threshold algorithm is used: traverse the gray value 0-255, calculate the inter-class variance of the hole wall area and the non-hole wall area under different threshold values, and select the gray value with the maximum inter-class variance as the segmentation threshold; the gray image is binarized according to the determined threshold value; S214: an edge detection algorithm with strong anti-noise ability is used to detect the edges of the binarized image, realize the extraction of the hole wall contour, and finally form a continuous and complete hole wall edge contour line, and exclude the isolated accumulated water and bubble edges; S215: Establish a pixel coordinate system with the top left corner of the image as the origin, and extract the two-dimensional coordinates of all pixels on the edge contour of the hole wall. u , v This forms the initial feature point set; The feature point density is calculated, processing is performed according to the feature point density to ensure that the feature point density of the whole hole wall meets the requirement, and the feature points are sorted in the circumferential direction to form an ordered feature point sequence u 1, v 1),( u 2, v 2),...,( u n , v n )}.

3. The method for adaptive directional measurement of hole bottom sediment thickness for irregular hole shape according to claim 2, characterized in that, The specific process of correcting the feature point coordinates to three-dimensional coordinates with the carrier unit axis as the origin in step S2 is as follows: S221: taking the axis of the measurement device carrier unit as the Z axis and the center of the cross-sectional circle of the carrier unit as the origin O, an X axis and a Y axis are established in the cross section to form a right-hand three-dimensional coordinate system O-XYZ; S222: Acquire focal length f , pixel size s , principal point coordinates u 0, v 0); S223: Coordinate transformation calculation: Transform each feature point ( uᵢ,vᵢ Convert to three-dimensional coordinates Xᵢ,Yᵢ,Zᵢ ); S224: convert the imaging plane coordinates into three-dimensional space coordinates.

4. The method for adaptive directional measurement of hole bottom sediment thickness for irregular hole shape according to claim 2, characterized in that, The specific process of generating a dynamic three-dimensional cross-sectional profile model of the irregular hole by fitting the three-dimensional feature points in step S2, dividing the hole wall deformation level, and determining the hole wall inclination angle in different directions is as follows: S231: Grouping three-dimensional feature points according to Z-axis coordinates: set the interval between layers Δ Z , divide the Z-axis into intervals including[ Z 1, Z 1+Δ Z ]、[ Z 1+Δ Z , Z 1+2Δ Z ], and Zᵢ group feature points belonging to the same interval into a group to form a depth layer feature point group, each group corresponding to a certain depth of the hole cross section; S232: The B-spline curve fitting algorithm is adopted, and the fitting result is optimized in combination with the hole wall features, so that a smooth and accurate single-depth layer profile is finally obtained; S233: The optimized profiles of all depth layers are arranged in sequence along the Z axis, and the profile details between adjacent depth layers are supplemented through linear interpolation to form a dynamic three-dimensional cross-sectional profile model; S234: The hole wall deformation level is divided based on the deviation degree of the hole wall profile in the model from the standard circular profile; S235: The angle between the hole wall tangent and the horizontal plane is calculated, wherein the angle between the hole wall tangent and the horizontal plane is the hole wall inclination angle; S236: The hole bottom position and range are determined based on the profile shape change in the Z-axis direction in the model.

5. The method for adaptive directional measurement of hole bottom sediment thickness for irregular hole shape according to claim 4, characterized in that, The specific process of step S3 is as follows: S31: According to the hole wall deformation level output by the dynamic three-dimensional cross-sectional profile model, the detection area is divided and the number of detection points in each area is determined to ensure that the severely deformed area has a higher detection density; S32: Based on the hole wall inclination angle, the optimal detection angle of each detection point is calculated in combination with the hole wall shape at the detection point position; S33: Real-time angle compensation is realized based on the dynamic correction of the gyroscope and the electromagnetic damper to correct the probe angle displacement deviation caused by hole vibration; S34: On the basis of real-time angle compensation, the orientation of the sludge detection assembly is adjusted by a servo motor to make the angle between the probe axis and the hole bottom tangent direction reach the optimal detection angle.

6. The method for adaptive directional measurement of hole bottom sediment thickness for irregular hole shape according to claim 1, characterized in that, The specific process of extracting the features of the upper loose and lower dense areas of the sludge in step S4 is as follows: S41: Select the decomposition parameters for multi-scale decomposition by wavelet transform, analyze the decomposition results, and the first to second layer detail components correspond to near-surface reflections, the third to fourth layer detail components correspond to middle-layer reflections, and the fifth layer detail component corresponds to deep-layer reflections; S42: Layered reflection peak value detection and feature judgment: peak value detection is performed on each layer of detail components, and the peak voltage is used to distinguish the upper loose area and the lower dense area of the sludge.

7. The method for adaptive directional measurement of hole bottom sediment thickness for irregular hole shape according to claim 6, characterized in that, The specific process of calculating the detection distance from the probe to the surface of the upper loose area and the surface of the lower dense area of the sludge in step S4 is as follows: S43: The one-way distance is calculated based on the path of the detection signal from the probe to the surface of the sludge reflected back to the receiver; S44: Calculate the distance of the probe to the surface of the upper loose zone of the sediment d 1; S45: Calculate the distance from the probe to the surface of the dense zone under the sediment d 2 .

8. The method for adaptive directional measurement of hole bottom sediment thickness for irregular hole shape according to claim 7, characterized in that, The specific process of extracting the hole wall bottom reference coordinates corresponding to each detection point and calculating the reference distance based on the dynamic three-dimensional cross-sectional profile model in step S5 is as follows: S51: From the dynamic three-dimensional cross-section profile model, the hole bottom depth corresponding to the Z-axis coordinate of the detection point is screened out Z p The cross-section profile layer that is completely consistent with the hole wall bottom depth layer is the profile layer with Z=505 mm in the extraction model, which ensures that the subsequent coordinate extraction matches the depth of the detection point. S52: mapping of the coordinates of the probing points: knowing the planar coordinates of the probing points in the O-XY plane ( X p ,Y p ) ; drawing a ray from the point ( X p ,Y p ) in a radial direction away from the axis of the carrier unit, the first intersection of the ray with the contour curve of the bottom depth of the hole wall defines the reference point of the bottom of the hole wall P 0 ; S53: Extract reference coordinates: The contour of each depth layer of the dynamic 3D model is generated by fitting the corrected 3D feature points, and each contour point stores accurate coordinates. X,Y,Z Spatial coordinates; read P 0 The coordinate data in the model are the reference coordinates of the bottom of the hole wall; S54: Hole wall bottom reference distance calculation: first, obtain the initial Z-axis coordinate of the probe, correct it in combination with the angle sensor data to obtain the corrected Z-axis coordinate, and then obtain the reference distance based on the corrected probe coordinate and the hole wall bottom reference coordinate.

Images