Dimple quality online detection method and system based on 3D scanning and virtual assembly

Abstract

The invention provides a dimple quality online detection method and system based on 3D scanning and virtual assembly, and the method comprises the steps: S1, obtaining dimple hole point cloud data, segmenting the point cloud data, and obtaining conical surface point cloud data; s2, performing conical surface fitting on the conical surface point cloud data through a fitting algorithm to obtain a local point set and a cone parameter; and S3, according to the local point set and the cone parameters, pit diameter contour plane fitting is carried out, the pit diameter contour plane and the nail head are assembled in a simulated mode, and a dimple depth error value is obtained. On the basis of 3D scanning and virtual assembly, pit hole surface point cloud analysis, nail rod interference analysis and compensation functions are achieved, pit depth measurement is achieved, and compared with a monocular or binocular vision scheme, the measurement precision is higher.

Description

Technical Field

[0001] The present invention relates to the field of aviation assembly quality inspection, and specifically, to an on-line inspection method and system for countersink quality based on 3D scanning and virtual assembly. Background Art

[0002] The processing and inspection of holes on the aircraft surface are still a major bottleneck in the field of aircraft automated assembly, mainly due to reasons such as large engineering quantities, complex and diverse materials, and high precision requirements. For the inspection of countersink hole quality, traditional manual inspection needs to be completed with a variety of inspection tools such as hole diameter gauges, hole depth gauges, perpendicularity gauges, roughness meters, etc., with extremely large workload and low efficiency.

[0003] With the development of modern aviation technology, aircraft assembly technology is gradually entering a new stage of intelligence and digitization. Machine vision measurement is a non-contact measurement method, which has advantages such as high efficiency, high precision, low cost, and easy automation compared with manual contact measurement. It is now widely used and developing rapidly. In order to meet the high-precision, high-efficiency, and high-reliability requirements for geometric quantity inspection in the process of assembling a new generation of aircraft, an on-line quality inspection function for countersink holes is realized based on 3D scanning and virtual assembly analysis technology. Summary of the Invention

[0004] Aiming at the defects in the prior art, the purpose of the present invention is to provide an on-line inspection method and system for countersink quality based on 3D scanning and virtual assembly.

[0005] An on-line inspection method for countersink quality based on 3D scanning and virtual assembly according to the present invention includes:

[0006] Step S1: Obtain the point cloud data of the countersink hole, and segment the point cloud data to obtain the cone point cloud data;

[0007] Step S2: Perform cone fitting on the cone point cloud data through a fitting algorithm to obtain an inlier set and cone parameters;

[0008] Step S3: Perform fitting of the countersink diameter contour plane according to the inlier set and cone parameters, simulate the assembly of the countersink diameter contour plane and the nail head, and obtain the countersink depth error value.

[0009] Preferably, the step S1 includes: calculating the point cloud normal using the neighborhood point cloud to fit a small plane, and performing point cloud segmentation using the normal vector and the Z-axis direction vector of the measurement coordinate system.

[0010] Preferably, the following method is adopted for the point cloud segmentation:

[0011] Calculate the point cloud normal vector n L using the neighborhood point cloud to fit a small plane, and use the local normal vector n L and the Z-axis direction vector of the measurement coordinate system Perform point cloud segmentation and calculate the vector n L and the vector n z The minimum included angle θ is as follows:

[0012]

[0013] Segment the point cloud according to the included angle θ.

[0014] Preferably, the step S2 includes: using the random sample consensus algorithm to fit the point cloud data to obtain the inlier set of the conical surface and the conical parameters;

[0015] The conical parameters include: the conical vertex coordinate P C , the conical axis vector V C , the conical angle α C .

[0016] Preferably, the socket diameter contour plane and the nail head simulation assembly in the step S3 include:

[0017] Define the nail head cone angle as α F , the vertex P of the nail head cone angle of the fastener in the measurement coordinate system F and the axis direction V F are as follows:

[0018]

[0019] where x F , y F , z F , β F , γ F are the pose parameters of the fastener nail head;

[0020] Then the distance d from a point P i on the socket hole surface to the nail head is: i :

[0021]

[0022] When d i is greater than 0, it indicates that the point is outside the nail head and there is a gap between the point and the nail head; when d i is less than 0, the point is inside the nail head and there is interference between the point and the nail head; the ratio of the interfering point cloud to the total point cloud is defined as the interference amount μ between the nail head and the point cloud.

[0023] Preferably, the nail head pose parameters are determined as follows:

[0024] Use the LM method to solve the optimization equation shown in Equation (4) to achieve rough matching of the point cloud and the nail head. When solving, let p F = p C , vF = v C , the initial value of the parameter to be identified is obtained by inverse solution of Equation (2). After the solution is completed, the identified parameter is substituted into Equation (2) to obtain the and

[0025]

[0026] When and are used as the pose of the nail head, there is often a large interference amount between the point cloud and the nail head of the fastener; define and corresponding to d i The cumulative distribution function is f(z) = P(d ≤ z), then F(0) is the interference amount when the nail head parameters are and ; set the target interference amount to μ t , then the nail head pose parameters for simulating assembly are:

[0027]

[0028] When μ t = 0.1 - 0.3, it is in a fitting state.

[0029] Preferably, the countersink diameter profile plane is determined by the following method:

[0030] Within the countersink diameter range, the point cloud is intercepted at equal intervals by planes perpendicular to the X-axis and Y-axis of the measurement coordinate system respectively to obtain the point cloud within the corresponding X-section and Y-section. The random sample consensus algorithm is used to fit the point cloud curves within the X-section and Y-section respectively, and the bisection method is used to calculate the intersection coordinates of the two curves, that is, the point cloud of the intersection of the countersink and the upper surface; the random sample consensus algorithm is used to perform plane fitting on the point cloud to obtain the countersink diameter profile plane.

[0031] Preferably, the countersink depth error value is calculated by the following method:

[0032] Calculate and The intersection point q C of the axis of the fastener represented by and the countersink diameter profile plane, and the final countersink depth deviation is obtained by using the following formula (6):

[0033]

[0034] where h F is the distance from the vertex of the conical surface of the fastener nail head to the nail head end face, which can be obtained from the fastener database; h C is the distance from the vertex of the conical surface of the fastener nail head to q CThe distance represents the depth of the vertex of the countersink cone surface after virtual assembly.

[0035] An on-line detection system for countersink quality based on 3D scanning and virtual assembly provided by the present invention includes:

[0036] Module M1: Obtain the point cloud data of the countersink hole, segment the point cloud data, and obtain the point cloud data of the cone surface;

[0037] Module M2: Perform cone surface fitting on the point cloud data of the cone surface through a fitting algorithm to obtain an inlier set and cone parameters;

[0038] Module M3: Perform fitting of the countersink diameter contour plane according to the inlier set and cone parameters, simulate the assembly of the countersink diameter contour plane and the nail head, and obtain the countersink depth error value.

[0039] Compared with the prior art, the present invention has the following beneficial effects:

[0040] 1. Based on 3D scanning and virtual assembly, the present invention realizes point cloud analysis of the countersink hole surface, nail rod interference analysis and compensation function, and realizes countersink depth measurement. Compared with monocular or binocular vision schemes, the measurement accuracy is higher, and the countersink depth measurement accuracy can reach ±0.015 mm.

[0041] 2. The present invention uses the RANSAC algorithm for point cloud segmentation, which can reduce the influence of noise and outliers on the measurement accuracy, and has higher efficiency compared with the method of calculating the slope of each point cloud point in each contour line in the prior art for feature segmentation.

[0042] 3. Through the virtual assembly process, the present invention can more accurately evaluate the installation quality of the countersink hole. Exclude the problem of unqualified flatness of fastener installation caused by non-concentricity between the countersink hole and the hole. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] By reading the following detailed description of the non-limiting embodiments with reference to the accompanying drawings, other features, objects and advantages of the present invention will become more apparent:

[0044] Figure 1 It is a flowchart of an on-line detection method for countersink hole quality disclosed by the present invention;

[0045] Figure 2 It is a schematic diagram of point cloud segmentation and cone surface fitting in the present invention;

[0046] Figure 3 It is a schematic diagram of virtual assembly of the nail head of the fastener in the present invention;

[0047] Figure 4 It is a schematic diagram of calculating the countersink depth deviation in the present invention;

[0048] Figure 5This is a flow chart of countersink quality detection disclosed in the present invention. DETAILED DESCRIPTION

[0049] The present invention will be described in detail below with reference to specific embodiments. The following examples will help those skilled in the art to further understand the present invention, but are not intended to limit the present invention in any form. It should be noted that, for those skilled in the art, several changes and improvements can be made without departing from the scope of the present invention. These all fall within the scope of protection of the present invention.

[0050] This patent proposes an online detection method for the quality of countersink holes. The point cloud data of the countersink holes is obtained by a 3D profile scanner. Based on the random sampling consensus algorithm (RANSAC, hereinafter referred to as the RANSAC algorithm) and virtual assembly technology, point cloud segmentation and high-precision measurement of the hole depth are completed, realizing the online detection function and quality evaluation of the hole depth. The overall process of this detection method is as follows Figure 1 and Figure 2 shown.

[0051] The specific implementation process is as follows:

[0052] Using a 3D scanner to acquire point cloud data, the point cloud normal is calculated, and the conical surface point cloud is segmented. The segmented conical surface is fitted using the RANSAC algorithm, resulting in a local point set and cone parameters. Once the conical surface parameters are obtained, the dimple diameter contour is fitted and the nail head assembly is simulated to determine the countersink depth error.

[0053] First, the point cloud normal vector n is calculated by fitting the small plane using the neighboring point cloud L , and using the local normal data vector n L and the Z-axis direction vector of the measurement coordinate system Perform point cloud segmentation and calculate vector n L and vector n z The minimum angle θ is as follows:

[0054]

[0055] When θ approaches 0°, the points are located on the outer surface. When θ approaches (90° - countersink angle / 2), where the countersink angle is a known parameter, typically 100° or 120°, the points are located on the conical surface. When θ approaches 90°, the points are located at the bottom. During point cloud segmentation, an appropriate threshold can be set to eliminate the influence of noise points.

[0056] After the point cloud is segmented, Figure 3 Then, in order to reduce the interference of noise, the RANSAC algorithm is used to fit the initial cone, and finally the filtered local point set Φ and the cone vertex coordinate P are obtained.C 、Conical axis vector V C 、Conical angle α of the cone C , as Figure 3 (b) shown.

[0057] Next, the points in the inlier set Φ are matched with the stud head theoretical model.

[0058] As Figure 4 shown,[[]]END]]

[0059] Define the stud head cone angle as α F , the vertex P of the stud head cone angle in the measurement coordinate system F and the axis direction V F are as follows:

[0060]

[0061] where x F , y F , z F , β F , γ F are the pose parameters of the stud head of the fastener;

[0062] Then the distance from a point P i on the countersink surface to the stud head is d i :

[0063]

[0064] When d i is greater than 0, it indicates that the point is outside the stud head and there is a gap between the point and the stud head; when d i is less than 0, the point is inside the stud head and there is interference between the point and the stud head. The ratio of the interfering point cloud to the total point cloud is defined as the interference amount μ between the stud head and the point cloud.

[0065] To obtain the optimal stud head pose parameters, first use the LM method to solve the optimization equation shown in Equation (4) to achieve a rough match between the point cloud and the stud head. When solving, let p F = p C , v F = v C , and the initial values of the parameters to be identified are obtained by inverse-solving Equation (2). After the solution is completed, substitute the identified parameters into Equation (2) to obtain the rough-matched and

[0066]

[0067] Substitute and When in the position of the fastener head, there is often a large interference amount between the point cloud and the fastener head. Considering the actual fitting state of the fastener head and the countersink, it is necessary to compensate the position of the fastener head to achieve precise matching of the point cloud and the fastener head. As shown in the figure, define and corresponding to the cumulative distribution function of d i as f(z) = P(d ≤ z), then F(0) is the interference amount when the fastener head parameters are and . Set the target interference amount as μ t , then the final fastener head position parameters for simulated assembly are:

[0068]

[0069] Among them, considering the existence of measurement errors, it can be considered that when μ t = 0.1 - 0.3, it conforms to the actual fitting state.

[0070] After completing the simulated assembly of the fastener head, in order to further analyze the deviation of the countersink depth, it is necessary to know the plane where the countersink diameter contour is located. As Figure 5 shown, within the nominal countersink diameter range, the point cloud is intercepted at equal intervals by planes perpendicular to the X-axis and Y-axis of the measurement coordinate system respectively, and the point cloud within the corresponding X-section and Y-section is obtained. Among them, the intersection line of the X / Y-section and the upper surface of the workpiece can be simplified as a parabola, and the intersection line of the X / Y-section and the countersink cone surface is defined as a conic curve. After respectively fitting the upper surface parabola and the countersink cone surface conic curve using the RANSAC algorithm, the intersection coordinates of the two curves can be obtained by using the bisection method. Through the above method, the point cloud of the intersection of the countersink and the upper surface can be obtained. Then, the RANSAC algorithm is used to perform plane fitting on the point cloud to obtain the plane of the countersink diameter contour.

[0071] Finally, calculate and to represent the intersection point q C of the fastener axis and the plane of the countersink diameter contour, and obtain the final deviation of the countersink depth using the following formula.

[0072]

[0073] Among them, h F is the distance from the vertex of the fastener head conical surface to the end face of the fastener head, which can be obtained from the fastener database; h C is the distance from the vertex of the fastener head conical surface to q C , representing the depth of the vertex of the fastener head conical surface after virtual assembly.

[0074] Those skilled in the art know that, in addition to implementing the system and its various devices, modules, and units provided by the present invention in the form of pure computer-readable program code, the method steps can be logically programmed to enable the system and its various devices, modules, and units provided by the present invention to be implemented in the form of logic gates, switches, application-specific integrated circuits, programmable logic controllers, and embedded microcontrollers, etc. to achieve the same functions. Therefore, the system and its various devices, modules, and units provided by the present invention can be considered as a kind of hardware component, and the devices, modules, and units included therein for implementing various functions can also be regarded as the structures within the hardware component; the devices, modules, and units for implementing various functions can also be regarded as either software modules for implementing the method or structures within the hardware component.

[0075] The specific embodiments of the present invention have been described above. It should be understood that the present invention is not limited to the above specific embodiments, and those skilled in the art can make various changes or modifications within the scope of the claims, which do not affect the essence of the present invention. Without conflict, the embodiments of the present application and the features in the embodiments can be combined with each other arbitrarily.

Claims

1. An on-line detection method for countersink quality based on 3D scanning and virtual assembly, characterized in that Including: Step S1: Obtain the countersunk hole point cloud data, segment the point cloud data, and obtain the conical point cloud data; Step S2: Perform conical fitting on the conical point cloud data through a fitting algorithm to obtain the inlier set and conical parameters; Step S3: Perform countersink diameter contour plane fitting based on the inlier set and conical parameters, and simulate the assembly of the countersink diameter contour plane and the nail head to obtain the countersunk depth error value.

2. The online detection method for countersinking quality based on 3D scanning and virtual assembly according to claim 1, characterized in that The said Step S1 includes: calculating the point cloud normal using the neighborhood point cloud to fit a small plane, and segmenting the point cloud using the normal vector and the Z-axis direction vector.

3. The online detection method for counterboring quality based on 3D scanning and virtual assembly according to claim 2, characterized in that The following method is adopted for the point cloud segmentation: Calculating the normal vector n of the point cloud by fitting a small plane using neighborhood point clouds L , and using the local normal vector n L and the direction vector of the Z-axis of the measurement coordinate system to perform point cloud segmentation, and calculating the minimum included angle θ between the vector n L and the vector n z as follows: Segment the point cloud according to the included angle θ.

4. The online detection method for countersinking quality based on 3D scanning and virtual assembly according to claim 1, characterized in that, The said Step S2 includes: performing fitting on the point cloud data using the random sample consensus algorithm to obtain the inlier set and conical parameters of the conical surface; The conical parameters include: the coordinates P of the conical vertex C , the vector V of the conical axis C , and the conical angle α C .

5. The online detection method for counterboring quality based on 3D scanning and virtual assembly according to claim 1, wherein The simulation assembly of the countersink diameter contour plane and the nail head in the said Step S3 includes: Define the nail head cone angle as α F , the vertex P of the fastener nail head cone angle in the measurement coordinate system F and the axis direction V F are as follows: where x F , y F , z F , β F , γ F are the pose parameters of the fastener head; Then the distance from a point P on the surface of the socket hole i to the nail head is d i : When d i is greater than 0, it indicates that the point is outside the nail head and there is a gap between the point and the nail head; when d i is less than 0, the point is inside the nail head and there is interference between the point and the nail head; the ratio of the interfering point cloud to the total point cloud is defined as the interference amount μ between the nail head and the point cloud.

6. The online detection method for counterboring quality based on 3D scanning and virtual assembly according to claim 5, characterized in that The pose parameters of the nail head are determined in the following manner: The LM method is used to solve the optimization equation shown in Equation (4) to achieve the rough matching of the point cloud and the nail head. During the solution, let p F = p C , v F = v C . The initial values of the parameters to be identified are obtained by the inverse solution of Equation (2). After the solution is completed, the identified parameters are substituted into Equation (2) to obtain the and When and are used as the nail head poses, there is often a large interference amount between the point cloud and the fastener nail head; define and corresponding to d i The cumulative distribution function is f(z) = P(d ≤ z), then F(0) is the interference amount when the nail head parameters are and ; set the target interference amount to μ t , then the nail head pose parameters for simulating assembly are: When μ t is 0.1 to 0.3, it is in a conforming state.

7. The online detection method for countersink quality based on 3D scanning and virtual assembly according to claim 6, characterized in that The said countersink diameter contour plane is determined by the following method: Within the countersink diameter range, the point cloud is intercepted at equal intervals by planes perpendicular to the X-axis and Y-axis of the measurement coordinate system respectively to obtain the point cloud within the corresponding X-section and Y-section. The point cloud curves within the X-section and Y-section are respectively fitted using the random sample consensus algorithm, and the intersection coordinates of the two curves are calculated using the bisection method, which are the point cloud of the intersection of the countersink and the upper surface; the point cloud is fitted by a plane using the random sample consensus algorithm to obtain the countersink diameter contour plane.

8. The online detection method for countersinking quality based on 3D scanning and virtual assembly according to claim 6, characterized in that, characterized in that, The said countersunk depth error value is calculated by the following method: Calculation and the intersection point q of the fastener axis represented by and the counterbore diameter profile plane C , and obtain the final counterbore depth deviation by using the following formula (6): where h F is the distance from the vertex of the conical surface of the fastener head to the end face of the head, which can be obtained from the fastener database; h C is the distance from the vertex of the conical surface of the fastener head to q C and represents the depth of the vertex of the conical surface of the head after virtual assembly.

9. An on-line detection system for countersinking quality based on 3D scanning and virtual assembly, characterized in that, Including: Module M1: Obtain the countersunk hole point cloud data, segment the point cloud data, and obtain the conical point cloud data; Module M2: Perform conical fitting on the conical point cloud data through a fitting algorithm to obtain the inlier set and conical parameters; Module M3: Perform countersink diameter contour plane fitting based on the inlier set and conical parameters, and simulate the assembly of the countersink diameter contour plane and the nail head to obtain the countersunk depth error value.

Images