A method, system, equipment, and medium for measuring car body bracing in harsh environments.

Abstract

This invention relates to the field of intelligent inspection technology, specifically to a method, system, equipment, and medium for measuring the stiffeners of a train carriage in harsh environments. By deploying top-down through-beam lidar on the front and rear tops of the carriage, point cloud data inside the carriage is collected. Even if the point cloud of the stiffener itself cannot be imaged due to environmental influences, the trailing point cloud appearing on the extension line of the stiffener can still be used, combined with a multi-radar spatial geometric inversion algorithm, to accurately calculate the true three-dimensional spatial position and size of the stiffener. Based on the spatial distribution of the trailing point cloud of the multi-lidar, a multi-line intersection inversion algorithm is used to achieve three-dimensional positioning and size measurement of the stiffener, without relying on the point cloud of the stiffener itself. It has strong environmental adaptability and automation capabilities, significantly improving the robustness, efficiency, and intelligence level of train carriage structure inspection.

Description

Technical Field

[0001] This invention relates to the field of intelligent monitoring technology, and more specifically, to a method, system, equipment, and medium for measuring the tensioning of train carriages in harsh environments. Background Technology

[0002] With the development of smart mines and intelligent logistics, automated loading systems are placing higher demands on the 3D perception of the internal structure of the loading compartment. LiDAR, as a mainstream 3D perception device, can efficiently acquire point cloud data of the internal structure of the loading compartment. However, in practical applications, environmental factors such as rain, fog, and dust significantly reduce the imaging capability of LiDAR for structures such as small tie rods inside the loading compartment, often resulting in point cloud loss and trailing phenomena. This makes it difficult to accurately obtain the true location and size of the tie rods, seriously affecting the path planning and obstacle avoidance capabilities of the automated loading robot arm, posing safety hazards and efficiency bottlenecks.

[0003] In scenarios such as truck body structure inspection, automated loading and unloading, and intelligent security inspection, accurately obtaining the three-dimensional position and dimensions of structural components such as tie rods inside the truck body is crucial. Traditional methods rely on lidar to directly image the point cloud of the tie rod. However, in harsh environments such as rain, fog, and dust, the point cloud of the tie rod itself often cannot be imaged, and only a trailing point cloud appears on the extension line of the tie rod, causing traditional detection methods to fail. Therefore, there is an urgent need for a technical solution that can invert the true spatial position and dimensions of the tie rod using the trailing point cloud when the tie rod itself is not visible.

[0004] Existing measurement methods mainly use lidar to directly scan the interior of the truck compartment, and achieve three-dimensional positioning and dimensional measurement of the tie rod by extracting and analyzing the point cloud of the tie rod itself. However, this method has the following drawbacks:

[0005] 1. Existing methods are highly dependent on the point cloud of the tie rod itself, mostly relying on lidar to directly image the point cloud of the tie rod. Accurate detection and measurement are only possible when there are sufficient point cloud echoes on the surface of the tie rod. When encountering harsh environments such as rain, fog, or dust, the point cloud of the tie rod is easily lost, leading to detection failure.

[0006] 2. Weak anti-interference ability: In complex environments, there are a lot of noise and interference points in point clouds. Existing algorithms have difficulty effectively distinguishing between the stretched point cloud and background noise, resulting in high false detection and false detection rates.

[0007] 3. Multi-radar data fusion is difficult. Existing technologies suffer from accuracy loss in calibration, coordinate unification, and data fusion, which affects the accuracy of the final measurement results.

[0008] 4. Traditional methods often require the radar to be installed in a fixed position and at a precise angle; otherwise, the accuracy of point cloud reconstruction and beam positioning will be affected, resulting in poor adaptability to the field.

[0009] 5. In the existing process, point cloud preprocessing, selection of reinforcement areas, and handling of abnormal situations often require manual intervention, which increases labor costs. Summary of the Invention

[0010] This invention addresses the problem that existing measurement methods cannot accurately detect the three-dimensional position and dimensions of car body tie rods in harsh environments. It proposes a car body tie rod measurement method, system, equipment, and medium based on harsh environments. Based on the spatial distribution of multiple lidar trailing point clouds, a multi-line intersection inversion algorithm is used to achieve three-dimensional positioning and dimension measurement of tie rods. It does not rely on the tie rod body point cloud, has strong environmental adaptability and automation capabilities, and significantly improves the robustness, efficiency, and intelligence level of car body structure inspection.

[0011] The specific implementation details of this invention are as follows:

[0012] A method for measuring the stiffeners of a train carriage in harsh environments includes the following steps:

[0013] Step S1: Preprocess the acquired raw interior point cloud data of the carriage to obtain preprocessed interior point cloud data of the carriage;

[0014] Step S2: Convert the preprocessed point cloud data of the interior of the carriage to the world coordinate system;

[0015] Step S3: Analyze the point cloud data inside the carriage in the world coordinate system, calculate the direction of each point in the point cloud relative to the radar, obtain the direction angle, and identify the trailing point cloud;

[0016] Step S4: Using the radar position corresponding to the trailing point cloud as the starting point and the direction relative to the radar as the direction vector, construct a spatial straight line;

[0017] Step S5: Use the multi-line intersection inversion algorithm and the least squares method to find the point with the smallest sum of squared distances to the constructed spatial straight line as the three-dimensional position of the tendon;

[0018] Step S6: Based on the three-dimensional position of the tie rod and the set search radius, obtain the dimensions of the tie rod and output them visually.

[0019] To better realize the present invention, step S1 further includes the following steps:

[0020] Step S11: Collect the original point cloud data of the interior space of the carriage based on the lidar A and lidar B set on both sides of the carriage.

[0021] Step S112: Based on the set voxel size, downsample the original point cloud data and merge points located in the same voxel into one point;

[0022] Step S113: Based on the obtained position of the carriage side panel, the original point cloud data after spatial cropping and downsampling is obtained to obtain the cropped point cloud data.

[0023] To better realize the present invention, step S2 further includes the following steps:

[0024] Step S21: Obtain the transformation matrix from the calibrated lidar to the world coordinate system;

[0025] Step S22: Based on the transformation matrix, transform the clipped point cloud data to the world coordinate system to obtain point cloud data in the world coordinate system.

[0026] To better realize the present invention, step S3 further includes the following steps:

[0027] Step S31: Analyze the point cloud data in the world coordinate system and calculate the direction vector of the point cloud data relative to the radar;

[0028] Step S32: Based on the direction vector and vertical direction vector of the point cloud data relative to the radar position, calculate the angle between the direction vector and the vertical direction to obtain the direction angle;

[0029] Step S33: Statistical angle interval density. Based on the set threshold or peak detection, identify the point cloud set corresponding to the potential tension direction to obtain the trailing point cloud.

[0030] To better realize the present invention, step S4 further includes the following steps:

[0031] Step S41: Take the radar position corresponding to the trailing point cloud as the starting point and the direction determined by the trailing point cloud as the direction vector to construct a spatial straight line;

[0032] Step S42: Form a set of straight lines from different radars and different trailing power sources.

[0033] To better realize the present invention, the specific operation of step S5 is as follows: calling the multi-line intersection inversion algorithm and the least squares method to find the point with the smallest sum of squared distances to the constructed set of lines as the three-dimensional position of the tendon.

[0034] To better realize the present invention, step S6 further includes the following steps:

[0035] Step S61: Based on the obtained three-dimensional position of the tendon, select points within a set range in the original point cloud with the three-dimensional position of the tendon as the center and according to the set search radius to obtain the point cloud set within the tendon area;

[0036] Step S62: Calculate the maximum and minimum range of the point cloud set in the X direction within the tensioning area to obtain the tensioning length;

[0037] Step S63: Calculate the maximum and minimum range of the point cloud set in the Y direction within the tensioning area to obtain the tensioning width;

[0038] Step S64: Calculate the maximum and minimum range of the point cloud set in the Z direction within the tensioning area to obtain the tensioning height;

[0039] Step S65: Visualize the original point cloud, brace location, and dimensions.

[0040] Based on the above-mentioned method for measuring the stiffener of a carriage under harsh environments, in order to better realize the present invention, a system for measuring the stiffener of a carriage under harsh environments is further proposed, which is used to perform the above-mentioned method for measuring the stiffener of a carriage under harsh environments; including a preprocessing unit, a conversion unit, an identification unit, a construction unit, an inversion calculation unit, and a visualization output unit;

[0041] The preprocessing unit is used to preprocess the acquired original point cloud data inside the carriage to obtain preprocessed point cloud data inside the carriage.

[0042] The conversion unit is used to convert the preprocessed point cloud data of the interior of the carriage to the world coordinate system;

[0043] The identification unit is used to analyze the point cloud data inside the carriage in the world coordinate system, calculate the direction of each point in the point cloud relative to the radar, obtain the direction angle, and identify the trailing point cloud.

[0044] The construction unit is used to take the radar position corresponding to the trailing point cloud as the starting point and the direction relative to the radar as the direction vector to construct a spatial straight line.

[0045] The inversion calculation unit is used to call the multi-line intersection inversion algorithm and the least squares method to find the point with the smallest sum of squared distances to the constructed spatial straight line as the three-dimensional position of the tendon.

[0046] The visualization output unit is used to obtain the dimensions of the tensioner based on its three-dimensional position and a set search radius, and then visualize and output these dimensions.

[0047] Based on the above-mentioned method for measuring the stiffeners of a carriage under harsh environments, in order to better realize the present invention, an electronic device is further proposed, which includes a memory and a processor; the memory stores a computer program; when the computer program is executed on the processor, the above-mentioned method for measuring the stiffeners of a carriage under harsh environments is realized.

[0048] Based on the above-mentioned method for measuring the car body bracing under harsh environments, in order to better realize the present invention, a computer-readable storage medium is further proposed, wherein computer instructions are stored on the computer-readable storage medium; when the computer instructions are executed on the above-mentioned electronic device, the above-mentioned method for measuring the car body bracing under harsh environments is realized.

[0049] The present invention has the following beneficial effects:

[0050] (1) The present invention enables accurate detection of the three-dimensional position and size of the tie rod even when the point cloud of the tie rod body is missing, which significantly improves the adaptability and robustness of the system in complex environments such as rain, fog and dust.

[0051] (2) The present invention reduces the impact of point cloud noise on the detection results, effectively improving the accuracy and reliability of tendon detection.

[0052] (3) This invention improves the accuracy of multi-radar data fusion and ensures high accuracy in measuring the spatial position and size of the bracing.

[0053] (4) The present invention reduces the dependence on the accuracy of equipment installation and enhances the on-site adaptability and ease of use of the system.

[0054] (5) The present invention realizes the full automation and high efficiency of the tendon testing process, meets the needs of real-time and batch testing, and enhances the value of industrial applications. Attached Figure Description

[0055] Figure 1 This is a schematic diagram of radar installation provided by the present invention.

[0056] Figure 2 This is a schematic diagram of the data flow processing flow provided by the present invention.

[0057] Figure 3 This is a schematic diagram of the processing steps provided by the present invention.

[0058] Figure 4 The diagram illustrates the principle of inversion of the tie rod body provided by this invention. Detailed Implementation

[0059] To more clearly illustrate the technical solutions of the embodiments of the present invention, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. It should be understood that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments, and therefore should not be regarded as a limitation on the scope of protection. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0060] In the description of this invention, it should be noted that, unless otherwise explicitly specified and limited, the terms "set up," "connected," and "linked" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0061] Example 1:

[0062] This embodiment proposes a method for measuring the stiffeners of a train carriage in harsh environments, specifically including the following steps:

[0063] Step S1: Preprocess the acquired raw interior point cloud data of the carriage to obtain preprocessed interior point cloud data of the carriage;

[0064] Step S1 specifically includes the following steps:

[0065] Step S11: Collect the original point cloud data of the interior space of the carriage based on the lidar A and lidar B set on both sides of the carriage.

[0066] Step S112: Based on the set voxel size, downsample the original point cloud data and merge points located in the same voxel into one point;

[0067] Step S113: Based on the obtained position of the carriage side panel, the original point cloud data after spatial cropping and downsampling is obtained to obtain the cropped point cloud data.

[0068] Step S2: Convert the preprocessed point cloud data of the interior of the carriage to the world coordinate system;

[0069] Step S2 specifically includes the following steps:

[0070] Step S21: Obtain the transformation matrix from the calibrated lidar to the world coordinate system;

[0071] Step S22: Based on the transformation matrix, transform the clipped point cloud data to the world coordinate system to obtain point cloud data in the world coordinate system.

[0072] Step S3: Analyze the point cloud data inside the carriage in the world coordinate system, calculate the direction of each point in the point cloud relative to the radar, obtain the direction angle, and identify the trailing point cloud;

[0073] Step S3 specifically includes the following steps:

[0074] Step S31: Analyze the point cloud data in the world coordinate system and calculate the direction vector of the point cloud data relative to the radar;

[0075] Step S32: Based on the direction vector and vertical direction vector of the point cloud data relative to the radar position, calculate the angle between the direction vector and the vertical direction to obtain the direction angle;

[0076] Step S33: Statistical angle interval density. Based on the set threshold or peak detection, identify the point cloud set corresponding to the potential tension direction to obtain the trailing point cloud.

[0077] Step S4: Using the radar position corresponding to the trailing point cloud as the starting point and the direction relative to the radar as the direction vector, construct a spatial straight line;

[0078] Step S4 specifically includes the following steps:

[0079] Step S41: Take the radar position corresponding to the trailing point cloud as the starting point and the direction determined by the trailing point cloud as the direction vector to construct a spatial straight line;

[0080] Step S42: Form a set of straight lines from different radars and different trailing power sources.

[0081] Step S5: Use the multi-line intersection inversion algorithm and the least squares method to find the point with the smallest sum of squared distances to the constructed spatial straight line as the three-dimensional position of the tendon;

[0082] The specific operation of step S5 is as follows: call the multi-line intersection inversion algorithm and the least squares method to find the point with the smallest sum of squared distances to the constructed set of lines as the three-dimensional position of the tendon.

[0083] Step S6: Based on the three-dimensional position of the tie rod and the set search radius, obtain the dimensions of the tie rod and output them visually.

[0084] Step S6 specifically includes the following steps:

[0085] Step S61: Based on the obtained three-dimensional position of the tendon, select points within a set range in the original point cloud with the three-dimensional position of the tendon as the center and according to the set search radius to obtain the point cloud set within the tendon area;

[0086] Step S62: Calculate the maximum and minimum range of the point cloud set in the X direction within the tensioning area to obtain the tensioning length;

[0087] Step S63: Calculate the maximum and minimum range of the point cloud set in the Y direction within the tensioning area to obtain the tensioning width;

[0088] Step S64: Calculate the maximum and minimum range of the point cloud set in the Z direction within the tensioning area to obtain the tensioning height;

[0089] Step S65: Visualize the original point cloud, brace location, and dimensions.

[0090] Working principle: This embodiment uses top-down laser radars deployed on the front and rear tops of the carriage to collect point cloud data inside the carriage. Even if the point cloud of the tie rod itself cannot be imaged due to environmental influences, the trailing point cloud appearing on the extension line of the tie rod can still be used. Combined with a multi-radar spatial geometric inversion algorithm, the true three-dimensional spatial position and size of the tie rod can be accurately calculated. Based on the spatial distribution of the trailing point cloud of the multi-laser radar, a multi-line intersection inversion algorithm is used to realize the three-dimensional positioning and size measurement of the tie rod. It does not rely on the point cloud of the tie rod itself, has strong environmental adaptability and automation capabilities, and significantly improves the robustness, efficiency and intelligence level of carriage structure inspection.

[0091] Example 2:

[0092] This embodiment is based on the above embodiment 1, such as... Figure 1 , Figure 2 , Figure 3 , Figure 4 As shown, a specific embodiment will be described in detail.

[0093] System structure: such as Figure 1 As shown, two downward-facing laser radars are installed on the front and rear tops of the carriage to collect point clouds inside the carriage.

[0094] Data flow: such as Figure 2 As shown, after data acquisition, the data goes through the following steps in sequence: voxel downsampling, spatial clipping, coordinate unification, trailing point cloud analysis, tendon inversion, size estimation, and result output.

[0095] The main data flow process is as follows:

[0096] 1. Parameter and calibration reading: Read the data directory and calibration matrix to obtain the radar's spatial transformation parameters.

[0097] 2. Point cloud data reading and preprocessing: acquire point cloud data from two radars and perform voxel downsampling; transform the point cloud to a unified world coordinate system according to the calibration matrix; identify the side panels of the vehicle compartment within the lane and obtain the edge parameters of the vehicle compartment; perform spatial cropping on the point cloud to retain only the effective area inside the vehicle compartment.

[0098] 3. Initial range detection of the tendon: For the two radar point clouds, the angular distribution of the trailing point clouds is statistically analyzed to preliminarily infer the possible x-position range of the tendon.

[0099] 4. Precise Positioning and Dimensional Measurement of Tie Rods: For each preliminary interval, the point cloud is further cropped, and the three-dimensional position of the tie rod is accurately calculated using a multi-line intersection inversion algorithm. The point cloud of the tie rod region is clustered to estimate the tie rod dimensions.

[0100] 5. Results Output and Visualization: Output the three-dimensional coordinates and dimensions of all tension members, and visualize the point cloud and tension member locations.

[0101] Principle explanation: The processing steps are as follows Figure 3 As shown in the schematic diagram, the principle is as follows: Figure 4 As shown.

[0102] Step S1: Data acquisition and preprocessing.

[0103] Step S11: Simultaneously acquire 3D point cloud data of the interior space of the truck using LiDAR A and LiDAR B located on both sides of the truck bed; denot the point cloud acquired by LiDAR A as P. A The point cloud acquired by radar B is denoted as P. B ;

[0104] Step S12: Process the original point cloud P A and P B Perform voxel downsampling; set an appropriate voxel size v (e.g., 0.03 meters) to merge points located within the same voxel into one point;

[0105] Step S13: Perform carriage side panel recognition. Based on the recognized positions of the front, rear, left, and right side panels of the carriage (fitting the carriage boundary for carriages without side panels), perform spatial cropping on the downsampled point cloud to remove background points and irrelevant noise outside the carriage, obtaining the cropped point cloud P. A ′ and P B ′.

[0106] Voxel downsampling:

[0107]

[0108] Step S2: Calibration and coordinate unification.

[0109] Step S21: Read the transformation matrix T from the LiDAR A to the world coordinate system, which was obtained through pre-calibration. A The transformation matrix T from lidar B to the world coordinate system B T A and T B It is a 4x4 homogeneous transformation matrix that contains rotation and translation information;

[0110] Step S22: Combine the pre-processed point clouds P from the two radars A ′ and P B Each point cloud P in the world coordinate system is obtained by transforming it to a unified world coordinate system using its corresponding calibration matrix. A "and P B ".

[0111] For a point in the radar coordinate system Its corresponding point in the world coordinate system The calculation is as follows:

[0112]

[0113] Right now

[0114]

[0115] Where T is the corresponding calibration matrix (T A and T B ), which includes the rotation matrix R and the translation vector t.

[0116] Step S3: Trailing point cloud recognition.

[0117] Step S31: Analyze the point cloud P transformed to the world coordinate system A "and P B For each radar, its position in the world coordinate system is: O A and O B For reference, calculate the orientation of each point in the point cloud relative to the radar.

[0118] Calculation points Relative to radar position Direction vector:

[0119]

[0120] Through calculation The angle between the direction and the vertical direction (Z-axis) is the direction angle α.

[0121]

[0122] Among them, V i V is the direction vector of the point cloud relative to the radar position. z It is a vertical vector.

[0123] Step S32: Divide the statistically analyzed angle range into several intervals and count the number of point clouds in each interval (voting). Since the trailing point clouds tend to concentrate along the extension direction of the tie rod, the point cloud density in these directional intervals will be relatively high. By setting a threshold or peak detection, identify the point cloud sets corresponding to potential tie rod directions; these are the trailing point clouds.

[0124] Step S4: Constructing spatial straight lines.

[0125] Step S41: For each identified trailing point cloud point P i (from P) A "and P B ”), with its corresponding radar position O(O A and O B Starting from point P, i Determined direction Given a direction vector, construct a spatial line (ray). Spatial line L i Parametric equations:

[0126]

[0127] Among them, O i It is a straight line L i The starting point (i.e., the corresponding radar position), d i It is a unit direction vector, and t' is the "range parameter" on the straight line, representing the radar O. i Distance to that point.

[0128] Step S42: Construct a set of lines {L1, L...} from different radars and different trailing point clouds. 2,… , L M}

[0129] Step S5: Multi-line intersection inversion.

[0130] Step S51: Using the multi-line intersection inversion algorithm, based on the least squares method, find a point Q such that it is connected to the set of lines {L1, L...} 2,…, L M The point Q is the optimal three-dimensional position estimate for the tendon because it minimizes the sum of squared distances between all straight lines in the given area.

[0131] Each straight line is represented as:

[0132]

[0133] Where q is the coordinate of a point in three-dimensional space, and o i It is a straight line L i The starting point (i.e., the corresponding radar position), d i It is the unit direction vector that determines a straight line between two points.

[0134] And construct an overdetermined system of linear equations

[0135]

[0136] Where G is a matrix consisting of the direction vectors and correlation coefficients of all lines, m is the unknown vector to be solved (in this case, the coordinates of the three-dimensional coordinate point Q of the optimal intersection point), and d is a vector consisting of constant terms related to the starting point and direction vectors of all lines.

[0137] The two ways to calculate Q are as follows:

[0138] ① Objective function: Find point Q and minimize:

[0139]

[0140] Where M represents the number of lines.

[0141] ② Solve in matrix form:

[0142]

[0143] in .

[0144] Step S6: Tie rod dimension estimation and result output

[0145] Step S61: Based on the inverted 3D position Q of the tendon, in the original point cloud Centered on Q, select points within a certain range.

[0146] The point cloud set within the tensioning area is as follows:

[0147]

[0148] Where r is the search radius;

[0149] Step S62: Cluster these points or directly count their maximum and minimum value ranges in the XYZ directions to estimate the length, width, height and other dimensional parameters of the tie rod.

[0150] Tie length (along the X direction):

[0151]

[0152] Similar to calculating width (Y direction):

[0153]

[0154] And height (Z direction):

[0155]

[0156] Step S63: Output the three-dimensional coordinates and dimensional parameters of all detected tension members.

[0157] Step S64: Visualize the original point cloud and the position and size of the reinforcing bar obtained by inversion using tools such as Open3D.

[0158] Working Principle: This embodiment is compatible with various vehicle models, requiring no special adaptation, and its robustness is greatly improved: even if the point cloud of the tie rod itself is not visible, it can still accurately invert the 3D position and size of the tie rod; it has strong environmental adaptability: suitable for complex environments such as rain, fog, and dust, and has strong anti-interference ability; it has a high degree of automation: the entire process is automated, requiring no manual intervention, and is suitable for batch inspection; it has lower cost: it can use medium and low line count radars, reducing reliance on high-precision equipment and manual labor, and the system integration is simple; it has higher efficiency: it has a fast processing speed, supports real-time detection and batch inspection of multiple vehicle compartments; it has strong anomaly adaptability: even if the point cloud of the tie rod itself is missing, it can still obtain effective results through trailing point cloud inversion, and the inspection process is not interrupted.

[0159] The other parts of this embodiment are the same as those in Embodiment 1 above, so they will not be described again.

[0160] Example 3:

[0161] Based on any one of Embodiments 1-2 above, this embodiment proposes a carriage tie rod measurement system for harsh environments, used to perform the above-mentioned carriage tie rod measurement method for harsh environments; it includes a preprocessing unit, a conversion unit, an identification unit, a construction unit, an inversion calculation unit, and a visualization output unit;

[0162] The preprocessing unit is used to preprocess the acquired original point cloud data inside the carriage to obtain preprocessed point cloud data inside the carriage.

[0163] The conversion unit is used to convert the preprocessed point cloud data of the interior of the carriage to the world coordinate system;

[0164] The identification unit is used to analyze the point cloud data inside the carriage in the world coordinate system, calculate the direction of each point in the point cloud relative to the radar, obtain the direction angle, and identify the trailing point cloud.

[0165] The construction unit is used to take the radar position corresponding to the trailing point cloud as the starting point and the direction relative to the radar as the direction vector to construct a spatial straight line.

[0166] The inversion calculation unit is used to call the multi-line intersection inversion algorithm and the least squares method to find the point with the smallest sum of squared distances to the constructed spatial straight line as the three-dimensional position of the tendon.

[0167] The visualization output unit is used to obtain the dimensions of the tensioner based on its three-dimensional position and a set search radius, and then visualize and output these dimensions.

[0168] This embodiment also proposes an electronic device, which includes a memory and a processor; the memory stores a computer program; when the computer program is executed on the processor, the above-described method for measuring the car body bracing in harsh environments is implemented.

[0169] This embodiment also proposes a computer-readable storage medium storing computer instructions; when the computer instructions are executed on the aforementioned electronic device, the aforementioned method for measuring the car body bracing in harsh environments is implemented.

[0170] The other parts of this embodiment are the same as any one of the above embodiments 1-2, so they will not be described again.

[0171] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Any simple modifications or equivalent changes made to the above embodiments based on the technical essence of the present invention shall fall within the protection scope of the present invention.

Claims

1. A method for measuring the stiffening of carriages in harsh environments, characterized in that, Specifically, the following steps are included: Step S1: Preprocess the acquired raw interior point cloud data of the carriage to obtain preprocessed interior point cloud data of the carriage; Step S2: Convert the preprocessed point cloud data of the interior of the carriage to the world coordinate system; Step S3: Analyze the point cloud data inside the carriage in the world coordinate system, calculate the direction of each point in the point cloud relative to the radar, obtain the direction angle, and identify the trailing point cloud; Step S4: Using the radar position corresponding to the trailing point cloud as the starting point and the direction relative to the radar as the direction vector, construct a spatial straight line; Step S5: Use the multi-line intersection inversion algorithm and the least squares method to find the point with the smallest sum of squared distances to the constructed spatial straight line as the three-dimensional position of the tendon; Step S6: Based on the three-dimensional position of the tie rod and the set search radius, obtain the dimensions of the tie rod and output them visually.

2. The method for measuring the car body bracing in harsh environments according to claim 1, characterized in that, Step S1 specifically includes the following steps: Step S11: Collect the original point cloud data of the interior space of the carriage based on the lidar A and lidar B set on both sides of the carriage. Step S112: Based on the set voxel size, downsample the original point cloud data and merge points located in the same voxel into one point; Step S113: Based on the obtained position of the carriage side panel, the original point cloud data after spatial cropping and downsampling is obtained to obtain the cropped point cloud data.

3. The method for measuring the car body bracing in harsh environments according to claim 2, characterized in that, Step S2 specifically includes the following steps: Step S21: Obtain the transformation matrix from the calibrated lidar to the world coordinate system; Step S22: Based on the transformation matrix, transform the clipped point cloud data to the world coordinate system to obtain point cloud data in the world coordinate system.

4. The method for measuring the car body bracing in harsh environments according to claim 3, characterized in that, Step S3 specifically includes the following steps: Step S31: Analyze the point cloud data in the world coordinate system and calculate the direction vector of the point cloud data relative to the radar; Step S32: Based on the horizontal and vertical direction vectors of the point cloud data relative to the radar position, calculate the angle between the direction vector and the vertical direction to obtain the direction angle; Step S33: Statistical angle interval density. Based on the set threshold or peak detection, identify the point cloud set corresponding to the potential tension direction to obtain the trailing point cloud.

5. The method for measuring the car body bracing in harsh environments according to claim 4, characterized in that, Step S4 specifically includes the following steps: Step S41: Take the radar position corresponding to the trailing point cloud as the starting point and the direction determined by the trailing point cloud as the direction vector to construct a spatial straight line; Step S42: Form a set of straight lines from different radars and different trailing power sources.

6. The method for measuring the car body bracing in harsh environments according to claim 5, characterized in that, The specific operation of step S5 is as follows: call the multi-line intersection inversion algorithm and the least squares method to find the point with the smallest sum of squared distances to the constructed set of lines as the three-dimensional position of the tendon.

7. The method for measuring the car body bracing in harsh environments according to claim 6, characterized in that, Step S6 specifically includes the following steps: Step S61: Based on the obtained three-dimensional position of the tendon, select points within a set range in the original point cloud with the three-dimensional position of the tendon as the center and according to the set search radius to obtain the point cloud set within the tendon area; Step S62: Calculate the maximum and minimum range of the point cloud set in the X direction within the tensioning area to obtain the tensioning length; Step S63: Calculate the maximum and minimum range of the point cloud set in the Y direction within the tensioning area to obtain the tensioning width; Step S64: Calculate the maximum and minimum range of the point cloud set in the Z direction within the tensioning area to obtain the tensioning height; Step S65: Visualize the original point cloud, brace location, and dimensions.

8. A carriage tie rod measurement system based on harsh environments, used to perform the carriage tie rod measurement method based on harsh environments as described in claim 1; characterized in that, It includes a preprocessing unit, a transformation unit, a recognition unit, a construction unit, an inversion calculation unit, and a visualization output unit; The preprocessing unit is used to preprocess the acquired original point cloud data inside the carriage to obtain preprocessed point cloud data inside the carriage. The conversion unit is used to convert the preprocessed point cloud data of the interior of the carriage to the world coordinate system; The identification unit is used to analyze the point cloud data inside the carriage in the world coordinate system, calculate the direction of each point in the point cloud relative to the radar, obtain the direction angle, and identify the trailing point cloud. The construction unit is used to take the radar position corresponding to the trailing point cloud as the starting point and the direction relative to the radar as the direction vector to construct a spatial straight line. The inversion calculation unit is used to call the multi-line intersection inversion algorithm and the least squares method to find the point with the smallest sum of squared distances to the constructed spatial straight line as the three-dimensional position of the tendon. The visualization output unit is used to obtain the dimensions of the tensioner based on its three-dimensional position and a set search radius, and then visualize and output these dimensions.

9. An electronic device, characterized in that, It includes a memory and a processor; the memory stores a computer program; when the computer program is executed on the processor, it implements the method for measuring the car body bracing in harsh environments as described in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions; when the computer instructions are executed on the electronic device as described in claim 9, the method for measuring the car body bracing in harsh environments as described in any one of claims 1-7 is implemented.

Images