Automatic attitude adjustment docking system and method based on autonomous mobile platform

Abstract

The invention relates to the technical field of mechanical automatic docking, in particular to an automatic posture adjusting docking system and method based on an autonomous mobile platform. The data acquisition unit is used for acquiring distance data and angle data of a butt joint surface and feature points of the butt joint surface; the central control unit analyzes the transmitted data of the laser radar and the depth camera, compares the data with a target position to obtain the position and the angle of the part needing to be adjusted, generates an adjusting instruction and transmits the instruction to the posture adjusting unit; and the posture adjusting unit adjusts the posture of the component according to the received posture adjusting instruction. According to the automatic attitude adjustment docking system and method based on the autonomous mobile platform, median filtering is introduced in position deviation analysis for optimization, the robustness of position deviation calculation can be improved through median filtering, the position deviation can be kept in various environments, the filtering strength can be dynamically adjusted according to the actual situation, and the docking precision is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of mechanical automatic docking, and in particular to an automatic posture adjustment docking system and method based on an autonomous mobile platform. Background Art

[0002] In the aviation manufacturing and maintenance industry, the docking of aircraft parts is a critical and complex task. Aircraft parts such as wings, fuselage sections, engines, etc. are usually very large in size and can reach tens of meters in length. The weight of these parts is often between several tons and tens of tons. The handling and positioning require powerful equipment support. The docking of aircraft parts must achieve micron-level accuracy. Any slight error may lead to serious quality problems and even flight safety problems. In the prior art, the docking of aircraft parts usually relies on manual operation or partially automated equipment. The handling and docking of large aircraft parts requires a lot of physical labor. Workers are prone to fatigue after working for a long time, which affects work efficiency and quality. It is difficult for manual operation to achieve high precision requirements, especially in complex assembly environments. Slight errors may cause docking failure or require rework. Therefore, automatic docking technology has been developed, but the partial automation equipment of the existing automatic docking technology is usually fixed in a specific position and cannot flexibly cope with aircraft parts of different models and sizes. Although the partial automation equipment can improve the accuracy, the accuracy in complex environments is still limited, especially in dynamically changing environments. Therefore, an automatic attitude adjustment docking system and method based on an autonomous mobile platform are provided. Summary of the invention

[0003] The purpose of the present invention is to provide an automatic attitude adjustment and docking system and method based on an autonomous mobile platform, so as to solve the problems proposed in the above background technology that some automated equipment is usually fixed at a specific position and cannot flexibly cope with aircraft components of different models and sizes, and although some automated equipment can improve accuracy, the accuracy is still limited in complex environments.

[0004] To achieve the above object, on the one hand, the present invention aims to provide an automatic posture adjustment docking system based on an autonomous mobile platform, comprising:

[0005] A data acquisition unit, the data acquisition unit is used to acquire distance data, angle data and characteristic points of the docking surface, and transmit the distance data, angle data and characteristic points of the docking surface to the central control unit;

[0006] A central control unit, which analyzes the data transmitted from the laser radar and the depth camera, obtains the position and angle of the component to be adjusted compared with the target position, generates an adjustment instruction, and transmits the instruction to the posture adjustment unit;

[0007] A posture adjustment unit is used to adjust the posture of the component according to the received posture adjustment instruction.

[0008] As a further improvement of the technical solution, the data acquisition unit includes a laser radar, a depth camera and an image processing module, wherein the laser radar can scan the docking surface in real time, obtain distance data and angle data, and transmit the distance data and angle data to the central control unit, and the depth camera can capture a three-dimensional image of the docking surface and the surrounding environment in real time; the image processing module is used to identify feature points in the captured three-dimensional image and transmit the feature points to the central control unit;

[0009] As a further improvement of the technical solution, the image processing module identifies feature points in the captured three-dimensional image, and the specific steps are as follows:

[0010] S1. Use Sobel filter to calculate the gradient of the image in the horizontal and vertical directions;

[0011] S2. At each pixel point (x, y), construct an autocorrelation matrix M;

[0012] S3, calculating the response function of each pixel;

[0013] S4. At each pixel point (x, y), check the response function value within its 26 neighborhood. If R(x, y) is the maximum value within its 26 neighborhood and is greater than the preset threshold T, select the point whose response function R is greater than the preset threshold as the feature point.

[0014] As a further improvement of the technical solution, the S1 is specifically as follows:

[0015]

[0016]

[0017] Among them, I x is the horizontal gradient of the image; I y is the vertical gradient of the image; I z is the gradient of the image in the depth direction; I is the input image.

[0018] As a further improvement of this technical solution, S2 is specifically as follows:

[0019] M(x,y,z)=∑ (x′，y′，z′)∈W w(x′, y′, z′)·P;

[0020]

[0021] Where M(x,y,z) is the autocorrelation matrix; w(x′,y′,z′) is the Gaussian weight function; W is a small window centered on (x,y); P is a matrix; (x,y,z) are the coordinates of the center of the small window; (x′,y′,z′) is the offset of the pixel segment of the small window relative to the center point (x,y,z).

[0022] As a further improvement of the technical solution, S3 is specifically as follows:

[0023] R = det(M) - k·trace(M) 2 ;

[0024] det(M)=λ1λ2λ3;

[0025] trace(M)=λ1+λ2+λ3;

[0026] Among them, R is the response function; det(M) is the determinant of the autocorrelation matrix M; trace(M) is the trace of the autocorrelation matrix M; λ1 is the first eigenvalue of the autocorrelation matrix M; λ2 is the second eigenvalue of the autocorrelation matrix M; λ3 is the third eigenvalue of the autocorrelation matrix M; k is a constant.

[0027] As a further improvement of the present technical solution, the central control unit analyzes the data transmitted from the laser radar and the depth camera, and the analysis includes position deviation analysis, roll angle analysis, pitch angle analysis and yaw angle analysis.

[0028] As a further improvement of the technical solution, the position deviation analysis, roll angle analysis, pitch angle analysis and yaw angle analysis are specifically as follows:

[0029] Position deviation:

[0030] ΔX=X1-X0;

[0031] ΔY=Y1-Y0;

[0032] ΔZ=Z1-Z0;

[0033] Among them, (X1, Y1, Z1) is the position of the feature point in the current coordinates; (X0, Y0, Z0) is the target position where the component needs to move; ΔX is the deviation in the horizontal direction; ΔY is the deviation in the vertical direction; ΔZ is the deviation in the depth direction;

[0034] Roll Angle Adjustment:

[0035]

[0036] Among them, θ rollis the roll angle; (x1, y1) is the coordinate point of one end of the feature line segment captured by the camera in the image; (x2, y2) is the coordinate point of the other end of the feature line segment captured by the camera in the image;

[0037] Pitch angle adjustment:

[0038]

[0039] Among them, θ pitch is the pitch angle; (x1, y1, z1) is the coordinate point measured by the upper laser radar; (x2, y2, z2) is the coordinate point measured by the lower laser radar;

[0040] Yaw angle adjustment:

[0041]

[0042] Among them, θ yaw is the yaw angle; (x3, y3, z3) is the coordinate point measured by the left lidar; (x4, y4, z4) is the coordinate point measured by the right lidar.

[0043] As a further improvement of the technical solution, median filtering is introduced in the position deviation analysis for optimization, and the optimization is as follows:

[0044] ΔX′=X1-X0+α·median(ΔX n );

[0045] ΔY′=Y1-Y0+α·median(ΔY n );

[0046] ΔZ′=Z1-Z0+α·median(ΔZ n );

[0047] Among them, ΔX′ is the deviation in the horizontal direction after optimization; ΔY′ is the deviation in the vertical direction after optimization; ΔZ′ is the deviation in the depth direction after optimization; α is the adaptive parameter; ΔX n is a set of n measurement points in the horizontal direction; ΔY n is a set of n measurement points in the vertical direction; ΔZ n is a set of n measurement points in the depth direction; median(△X i ) is the median of the position deviations of n measuring points in the horizontal direction; median(ΔY i ) is the median of the position deviations of n measuring points in the vertical direction; median(ΔZ i ) is the median value of the position deviations of n measuring points in the depth direction.

[0048] On the other hand, the present invention provides an automatic posture adjustment and docking method based on an autonomous mobile platform, which is used in any of the above-mentioned automatic posture adjustment and docking systems based on an autonomous mobile platform, comprising the following steps:

[0049] S10.1, the data acquisition unit is used to acquire distance data, angle data and characteristic points of the docking surface, and transmit the distance data, angle data and characteristic points of the docking surface to the central control unit;

[0050] S10.2, the central control unit analyzes the data transmitted from the laser radar and the depth camera, and obtains the position and angle of the component that needs to be adjusted compared with the target position, generates an adjustment instruction, and transmits the instruction to the attitude adjustment unit;

[0051] S10.3. The posture adjustment unit adjusts the posture of the component according to the received posture adjustment instruction.

[0052] Compared with the prior art, the present invention has the following beneficial effects:

[0053] In an automatic posture adjustment and docking system and method based on an autonomous mobile platform, median filtering is introduced in position deviation analysis for optimization. Median filtering can improve the robustness of position deviation calculation so that it can be maintained in various environments. The intensity of the filter can also be dynamically adjusted according to actual conditions, thereby better eliminating noise and improving the accuracy of docking. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 It is the overall flow chart of the present invention;

[0055] Figure 2 The figure is a flow chart of the overall method in the present invention.

[0056] The meaning of each number in the figure is:

[0057] 1. Data acquisition unit; 2. Central control unit; 3. Posture adjustment unit. DETAILED DESCRIPTION

[0058] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0059] See also Figure 1 As shown, an automatic posture adjustment docking system based on an autonomous mobile platform is provided, comprising a data acquisition unit 1, a central control unit 2 and a posture adjustment unit 3;

[0060] It also includes an autonomous mobile platform, which can carry and transport large aircraft components as a mobile carrier. The platform has omnidirectional mobility and can flexibly turn in a small space. The attitude adjustment device is installed on the top of the autonomous mobile platform and is responsible for adjusting the attitude of the large components, including position, tilt angle, pitch angle and yaw angle. The attitude adjustment device adopts a multi-degree-of-freedom mechanical arm or parallel robot structure, which can achieve precise attitude adjustment; among them,

[0061] The data acquisition unit 1 is used to acquire the distance data, angle data and characteristic points of the docking surface, and transmit the distance data, angle data and characteristic points of the docking surface to the central control unit 2;

[0062] The data acquisition unit 1 includes a laser radar, a depth camera and an image processing module, wherein the laser radar can scan the docking surface in real time, obtain distance data and angle data, and transmit the distance data and angle data to the central control unit 2, and the depth camera can capture a three-dimensional image of the docking surface and the surrounding environment in real time; the image processing module is used to identify feature points in the captured three-dimensional image and transmit the feature points to the central control unit 2;

[0063] Multiple high-precision lidars are installed around the attitude adjustment device to form a 3D scanning layout, which is used to measure the distance and angle between the docking surface and the mobile platform. The lidar can obtain multi-point data at the same time to improve the accuracy and reliability of the measurement; a high-resolution depth camera is installed on the attitude adjustment device to capture three-dimensional images of the docking surface and the environment. The camera can identify the feature points of the docking surface and generate a depth map for precise positioning of the docking position.

[0064] It is crucial to ensure that the relative position between the two platforms is correct during the docking process. Using LiDAR, the system can understand the specific conditions of the docking surface in real time, and maintain good performance even in poor lighting conditions or changing environments. The three-dimensional structure of the scene captured by the depth camera is very useful for understanding the shape and layout of objects in complex environments, helping the system to more accurately identify docking targets, avoid collisions, and optimize path planning;

[0065] By analyzing the captured data, key feature points can be effectively extracted from the background, which helps to improve the accuracy of the final docking. In addition, image processing technology can also help the system identify and adapt to different docking objects, increasing the versatility and flexibility of the system.

[0066] The image processing module identifies feature points in the captured 3D image. The specific steps are as follows:

[0067] Harris corner detection is a classic feature point detection method, which is mainly used to detect corners in images. Its basic idea is to determine whether a point is a corner by calculating the grayscale change of each pixel in different directions.

[0068] S1. Use Sobel filter to calculate the gradient of the image in the horizontal and vertical directions;

[0069] S2. At each pixel point (x, y), construct an autocorrelation matrix M;

[0070] S3, calculating the response function of each pixel;

[0071] S4. In order to reduce the number of detected corner points and retain the most significant corner points, non-maximum suppression is performed.

[0072] At each pixel point (x, y), check the response function value in its 26 neighborhoods. If R(x, y) is the maximum value in its 26 neighborhoods and is greater than the preset threshold T, then select the points whose response function R is greater than the preset threshold as feature points. These points are the feature points in the image. The 26 neighborhood refers to the 26 adjacent points of the point (x, y, z);

[0073] S1 is as follows:

[0074]

[0075] Among them, I x is the horizontal gradient of the image, that is, the grayscale change rate in the horizontal direction; I y is the vertical gradient of the image, that is, the grayscale change rate in the vertical direction; I z is the gradient of the image in the depth direction, that is, the grayscale change rate in the depth direction; I is the input image; these gradient values ​​are used to detect edges and corners in the image, because corners are usually points with large grayscale changes in multiple directions.

[0076] S2 is as follows:

[0077] M(x,y,z)=∑ (x′，y′，z′)∈W w(x′, y′, z,)·P;

[0078]

[0079] Where M(x,y,z) is the autocorrelation matrix, which is used to describe the local structure of the image near each pixel; w(x′, y1, z′) is the Gaussian weight function, which is used to perform weighted averaging of the gradient product within the window, usually using a Gaussian kernel with a standard deviation of σ; W is a small window centered at (x,y), usually a 3×3×3 cubic window. The diagonal elements of the autocorrelation matrix M reflect the intensity of grayscale changes in all directions, and the non-diagonal elements reflect the correlation of grayscale changes in different directions; P is a matrix; (x,y,z) are the coordinates of the center of the small window; (x′,y′,z′) are the offsets of the pixel segments in the small window relative to the center point (x,y,z);

[0080] S3 is as follows:

[0081] R = det(M) - k·trace(M) 2 ;

[0082] det(M)=λ1λ2λ3;

[0083] trace(M)=λ1+λ2+λ3;

[0084] Among them, R is the response function; det(M) is the determinant of the autocorrelation matrix M, which represents the volume of the matrix. The larger the value of the determinant, the stronger the grayscale change of the point in multiple directions; trace(M) is the trace of the autocorrelation matrix M, which represents the sum of the diagonal elements of the matrix. The larger the value of the trace, the stronger the grayscale change of the point in certain directions; λ1 is the first eigenvalue of the autocorrelation matrix M; λ2 is the second eigenvalue of the autocorrelation matrix M; λ3 is the third eigenvalue of the autocorrelation matrix M; k is a constant;

[0085] Through the combination of determinant and trace, we can measure whether the point is a corner point. If the value of R is large, it means that the grayscale changes of the point in multiple directions are strong, which may be a corner point.

[0086] The central control unit 2 analyzes the data transmitted from the laser radar and the depth camera, obtains the position and angle of the component that needs to be adjusted compared with the target position, generates an adjustment instruction, and transmits the instruction to the posture adjustment unit 3;

[0087] The central control unit 2 analyzes the data transmitted from the laser radar and the depth camera, including position deviation analysis, roll angle analysis, pitch angle analysis, and yaw angle analysis;

[0088] The position deviation analysis, roll angle analysis, pitch angle analysis and yaw angle analysis are as follows:

[0089] Position deviation:

[0090] ΔX=X1-X0;

[0091] ΔY = Y1 - Y0;

[0092] ΔZ = Z1 - Z0;

[0093] Where, (X1 , Y1, Z1) is the position in the current coordinates of the feature point; (X0, Y0, Z0) is the target position to which the component needs to be moved; ΔX is the deviation in the horizontal direction; ΔY is the deviation in the vertical direction; ΔZ is the deviation in the depth direction; the central control unit 2 drives the attitude adjustment unit 3 to adjust the position of the component according to the calculated deviation.

[0094] The depth camera calculates the roll angle by capturing the horizontal line feature points of the docking surface. Roll angle adjustment:

[0095]

[0096] Where, θ roll is the roll angle; (x1, y1) is the coordinate point of one end of the feature line segment captured by the camera in the image; (x2, y2) is the coordinate point of the other end of the feature line segment captured by the camera in the image; by comparing the difference between the calculated roll angle and the target roll angle, the central control unit 2 drives the attitude adjustment unit 3 to adjust the roll angle.

[0097] The calculation of the pitch angle is based on the distance measurement data of the lidar system. Pitch angle adjustment:

[0098]

[0099] Where, θ pitch is the pitch angle; (x1, y1, z1) is the coordinate point measured by the upper lidar; (x2, y2, z2) is the coordinate point measured by the lower lidar; by comparing the difference between the calculated pitch angle and the target pitch angle, the central control unit 2 drives the attitude adjustment unit 3 to adjust the pitch angle.

[0100] The calculation of the yaw angle is based on the distance difference between the left and right sides of the docking surface measured by the lidar. Yaw angle adjustment:

[0101]

[0102] Where, θ yaw is the yaw angle; (x3, y3, z3) is the coordinate point measured by the left lidar; (x4, y4, z4) is the coordinate point measured by the right lidar; by comparing the difference between the calculated yaw angle and the target yaw angle, the central control unit 2 drives the attitude adjustment unit 3 to adjust the yaw angle.

[0103] Median filtering is introduced for optimization in the position deviation analysis. Specifically after optimization:

[0104] ΔX′=X1-X0+α·median(ΔX n );

[0105] ΔY′=Y1-Y0+α·median(ΔY n );

[0106] ΔZ′=Z1-Z0+α·median(ΔZ n );

[0107] Among them, ΔX′ is the deviation in the horizontal direction after optimization; ΔY′ is the deviation in the vertical direction after optimization; ΔZ′ is the deviation in the depth direction after optimization; α is the adaptive parameter; ΔX n is a set of n measurement points in the horizontal direction; ΔY n is a set of n measurement points in the vertical direction; ΔZ n is a set of n measurement points in the depth direction; median(△X i ) is the median of the position deviations of n measuring points in the horizontal direction; median(ΔY i ) is the median of the position deviations of n measuring points in the vertical direction; median(ΔZ i ) is the median value of the position deviation of n measuring points in the depth direction;

[0108] Median filtering is a nonlinear filtering method that can effectively reduce the impact of noise by selecting the median value in a set of data to replace the original value. Especially in the presence of outliers, median filtering is more robust than mean filtering; for example, if a measurement point is disturbed, resulting in an abnormally large deviation value, median filtering can ignore this outlier, thereby improving the accuracy of the calculation result;

[0109] Median filtering can improve the robustness of position deviation calculation, so that it can maintain high accuracy in various environments. Robustness refers to the ability of the system to maintain good performance in the face of uncertainty and interference. For example, in outdoor environments, the deviation values ​​of some measurement points may be unstable due to factors such as light changes and occlusion. Median filtering can help the system maintain stable performance in this case.

[0110] The adaptive parameter α can dynamically adjust the intensity of the filter according to the actual situation. When the noise is large, the value of α can be increased to enhance the filtering effect; when the noise is small, the value of α can be reduced to reduce the impact of the filter on the real signal. For example, if the system detects that the current environmental noise is large, it can automatically increase α , making the median filtering effect stronger and thus better eliminating noise; conversely, if the environmental noise is small, α can be reduced to maintain a higher response speed.

[0111] The posture adjustment unit 3 adjusts the posture of the component according to the posture adjustment instruction received. The posture adjustment unit 3 includes a high-precision servo motor, a linear actuator and a posture adjustment device, which is used to adjust the posture according to the instruction of the central control unit 2 to ensure the docking accuracy.

[0112] This embodiment also provides an automatic posture adjustment and docking method based on an autonomous mobile platform, which is used in the above-mentioned automatic posture adjustment and docking system based on an autonomous mobile platform, and includes the following steps:

[0113] S10.1, the data acquisition unit 1 is used to acquire distance data, angle data and characteristic points of the docking surface, and transmit the distance data, angle data and characteristic points of the docking surface to the central control unit 2;

[0114] S10.2, the central control unit 2 analyzes the data transmitted from the laser radar and the depth camera, and obtains the position and angle of the component that needs to be adjusted compared with the target position, generates an adjustment instruction, and transmits the instruction to the posture adjustment unit 3;

[0115] S10.3. The posture adjustment unit 3 adjusts the posture of the component according to the received posture adjustment instruction.

[0116] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited by the above embodiments, and the above embodiments and descriptions are only preferred examples of the present invention, and are not intended to limit the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, and these changes and improvements all fall within the scope of the present invention to be protected.

Claims

1. An automatic posture adjustment docking system based on an autonomous mobile platform, characterized in that: include: A data acquisition unit (1), the data acquisition unit (1) being used to acquire distance data, angle data and characteristic points of the docking surface, and transmit the distance data, angle data and characteristic points of the docking surface to a central control unit (2); A central control unit (2), wherein the central control unit (2) analyzes the data transmitted from the laser radar and the depth camera, obtains the position and angle of the component that need to be adjusted compared with the target position, generates an adjustment instruction, and transmits the instruction to the posture adjustment unit (3); A posture adjustment unit (3), wherein the posture adjustment unit (3) performs posture adjustment of the component according to a received posture adjustment instruction.

2. The automatic posture adjustment docking system based on the autonomous mobile platform according to claim 1 is characterized in that: The data acquisition unit (1) comprises a laser radar, a depth camera and an image processing module, wherein the laser radar can scan the docking surface in real time, obtain distance data and angle data, and transmit the distance data and angle data to the central control unit (2), and the depth camera can capture a three-dimensional image of the docking surface and the surrounding environment in real time; the image processing module is used to identify feature points in the captured three-dimensional image, and transmit the feature points to the central control unit (2).

3. The automatic posture adjustment docking system based on the autonomous mobile platform according to claim 2 is characterized in that: The image processing module identifies feature points in the captured three-dimensional image, and the specific steps are as follows: S1. Use Sobel filter to calculate the gradient of the image in the horizontal and vertical directions; S2. At each pixel point (x, y), construct an autocorrelation matrix M; S3, calculating the response function of each pixel; S4. At each pixel point (x, y), check the response function value in its 26 neighborhood. If R(x, y) is the maximum value in its 26 neighborhood and is greater than the preset threshold T, select the point whose response function R is greater than the preset threshold as the feature point.

4. The automatic posture adjustment docking system based on the autonomous mobile platform according to claim 3 is characterized in that: The S1 is specifically as follows: Among them, I x is the horizontal gradient of the image; I y is the vertical gradient of the image; I z is the gradient of the image in the depth direction; I is the input image.

5. The automatic posture adjustment docking system based on the autonomous mobile platform according to claim 4 is characterized in that: The S2 is specifically as follows: M(x,y,z)=∑ (x′，y′，z′)∈W w(x′,y′,z,)·P; Among them, M(x,y,z) is the autocorrelation matrix; w(x′, y′, z′) is the Gaussian weight function; W is a small window centered on (x,y); P is a matrix; (x,y,z) is the coordinates of the center of the small window; (x′, y′, z′) is the offset of the pixel segment of the small window relative to the center point (x,y,z).

6. The automatic posture adjustment docking system based on the autonomous mobile platform according to claim 5 is characterized in that: The S3 is as follows: R=det(M)-k·trace(M) 2 ; det(M)=λ1λ2λ3; trace(M)=λ1+λ2+λ3; Among them, R is the response function; det(M) is the determinant of the autocorrelation matrix M; trace(M) is the trace of the autocorrelation matrix M; λ1 is the first eigenvalue of the autocorrelation matrix M; λ2 is the second eigenvalue of the autocorrelation matrix M; λ3 is the third eigenvalue of the autocorrelation matrix M; k is a constant.

7. The automatic posture adjustment docking system based on the autonomous mobile platform according to claim 6 is characterized in that: The central control unit (2) analyzes the data transmitted from the laser radar and the depth camera, and the analysis includes position deviation analysis, roll angle analysis, pitch angle analysis and yaw angle analysis.

8. The automatic posture adjustment docking system based on the autonomous mobile platform according to claim 7 is characterized in that: The position deviation analysis, roll angle analysis, pitch angle analysis and yaw angle analysis are specifically as follows: Position deviation: ΔX=X1-X0; ΔY=Y1-Y0; ΔZ=Z1-Z0; Among them, (X1, Y1, Z1) is the position of the feature point in the current coordinates; (X0, Y0, Z0) is the target position where the component needs to move; ΔX is the deviation in the horizontal direction; ΔY is the deviation in the vertical direction; ΔZ is the deviation in the depth direction; Roll Angle Adjustment: Among them, θ roll is the roll angle; (x1, y1) is the coordinate point of one end of the feature line segment captured by the camera in the image; (x2, y2) is the coordinate point of the other end of the feature line segment captured by the camera in the image; Pitch angle adjustment: Among them, θ pitch is the pitch angle; (x1, y1, z1) is the coordinate point measured by the upper laser radar; (x2, y2, z2) is the coordinate point measured by the lower laser radar; Yaw angle adjustment: Among them, θ yaw is the yaw angle; (x3, y3, z3) is the coordinate point measured by the left lidar; (x4, y4, z4) is the coordinate point measured by the right lidar.

9. The automatic posture adjustment docking system based on the autonomous mobile platform according to claim 8, characterized in that: In the position deviation analysis, median filtering is introduced for optimization, and the optimization is as follows: ΔX′=X1-X0+α·median(ΔX n ); ΔY′=Y1-Y0+α·median(ΔY n ); ΔZ′ = Z1 - Z0 + α·median(ΔZ n ); Among them, ΔX′ is the deviation in the horizontal direction after optimization; ΔY′ is the deviation in the vertical direction after optimization; ΔZ′ is the deviation in the depth direction after optimization; α is the adaptive parameter; ΔX n is a set of n measurement points in the horizontal direction; ΔY n is a set of n measurement points in the vertical direction; ΔZ n is a set of n measurement points in the depth direction; median(△X i ) is the median of the position deviations of n measuring points in the horizontal direction; median(ΔY i ) is the median of the position deviations of n measuring points in the vertical direction; median(ΔZ i ) is the median value of the position deviations of n measuring points in the depth direction.

10. An automatic posture adjustment and docking method based on an autonomous mobile platform, used in an automatic posture adjustment and docking system based on an autonomous mobile platform as claimed in any one of claims 1 to 9, characterized in that: The steps include: S10.1, the data acquisition unit (1) is used to acquire distance data, angle data and characteristic points of the docking surface, and transmit the distance data, angle data and characteristic points of the docking surface to the central control unit (2); S10.2, the central control unit (2) analyzes the data transmitted from the laser radar and the depth camera, and obtains the position and angle of the component that needs to be adjusted compared with the target position, generates an adjustment instruction, and transmits the instruction to the posture adjustment unit (3); S10.

3. The posture adjustment unit (3) adjusts the posture of the component according to the received posture adjustment instruction.

Images