A consistent global constraint field of view planning method for blade profile line laser scanning

Abstract

The present invention belongs to the technical field of blade profile detection. The present invention discloses a consistent global constraint field of view planning method for blade profile line laser scanning, including step 1: calculating the change in the angle between the normal vectors of the profile data of the blade to be measured based on the blade CAD model and sensor parameters; step 2: obtaining the leading edge and trailing edge areas based on the change in the angle between the normal vectors of the profile data, and planning the field of view of the leading edge and the trailing edge; step 3: planning the field of view of the blade basin and the back of the blade by an optimization strategy of setting a dynamic threshold based on the change in the angle between the normal vectors of the profile points; step 4: optimizing the sensor posture in combination with the viewing angle; step 5: optimizing the sensor posture in combination with the viewing distance; step 6: confirming the field of view posture parameters and integrating the field of view. The present invention can effectively reduce the measurement uncertainty between different fields of view at the blade basin and the back of the blade, set a special planning mechanism for the leading and trailing edge areas, ensure the data collection accuracy of the area, and thus improve the overall detection accuracy and reliability.

Description

Technical Field

[0001] The invention belongs to the technical field of blade profile detection, and in particular relates to a consistent global constraint field of view planning method for blade profile line laser scanning. Background Art

[0002] In aircraft engines and gas turbines, blades are key power transmission components, and their surface accuracy directly affects the aerodynamic performance and operating efficiency of the engine. Therefore, blade surface detection is crucial to ensure the safety and efficient operation of the engine. In recent years, non-contact optical detection technology has gradually become the mainstream method for blade surface detection due to its advantages of high efficiency and flexibility without contacting the surface. According to the different structures of the light source and the data acquisition method, non-contact optical detection methods can be divided into point light source method, line light source method and surface light source method. Among them, the line light source method has the efficiency advantages of the surface light source and the accuracy advantages of the point light source, and has received widespread attention.

[0003] The Chinese invention patent with patent number 202110739936.6 discloses a blade multi-field point cloud registration method based on overlapping features and local distance constraints. This method can perform two-dimensional cross-section detection of blades without relying on the motion accuracy of the detection system. Due to the complex contours, high distortion, and large changes in the curvature of the leading and trailing edges of the blade, it is very easy to cause a significant difference between the imaging distance (viewing distance) and the incident angle (viewing angle) of the light source of any cross-sectional imaging segment of the blade during the scanning process, affecting the accuracy of the scanning data. The Chinese invention patent with patent number 202411073689.0 discloses a blade cross-section line laser scanning field of view planning method based on multivariate constraints, which improves the data acquisition accuracy; however, subsequent experiments found that this method has certain limitations and can only complete the optimal field of view planning of local contours. It is a local optimal strategy. In addition, due to the dense contour point cloud and complex geometric shape of the blade basin back area, it is easy to cause great differences in the scanning state between different fields of view, and the reliability and uncertainty of the measurement cannot be effectively controlled. At the same time, the characteristics of the leading and trailing edge areas of the blade, such as large changes in curvature and small radius, make it difficult for the above method to effectively guarantee the reliability and accuracy of the scanning data in this area. Summary of the invention

[0004] In view of the above problems, the purpose of the present invention is to propose a consistent global constraint field of view planning method for blade profile line laser scanning; this method can effectively reduce the measurement uncertainty between different fields of view at the blade basin and blade back, and improve the stability of the overall detection. At the same time, a special planning mechanism is set for the leading and trailing edge areas to ensure the data collection accuracy of this area, thereby improving the overall detection accuracy and reliability.

[0005] To achieve the above object, the present invention adopts the following technical solution:

[0006] A consistent global constraint field of view planning method for blade profile line laser scanning comprises the following steps:

[0007] Step 1: Calculate the change in the angle between the normal vector and the normal vector of the blade profile data set to be measured based on the blade CAD model and sensor parameters ;

[0008] Step 2: Obtain the leading edge and trailing edge areas based on the change in the angle between the normal vectors of the contour data, and plan the fields of view of the leading edge and trailing edge;

[0009] The change in the angle between the normal vectors of the contour points calculated in step 1 , perform clustering operation; based on the clustering results, extract the leading and trailing edge points , , the starting and ending points of the angle change between the normal vectors of the leading edge point , , and the starting and ending points of the angle change between the normal vectors of the trailing edge points , ;

[0010] Based on leading edge and its normal vector Calculate the beginning and end points of the leading edge field of view , ;extract and As the starting and ending points of the leading edge field of view 1, and As the starting and ending points of the leading edge field of view 2, that is, , ,in, , Respectively represent the contour point sets under the leading edge field of view 1 and field of view 2;

[0011] Based on the trailing edge and its normal vector Calculate the beginning and end points of the leading edge field of view , ; The trailing edge field of view is divided into: , ,in, , Respectively represent the contour point sets under the leading edge field of view 1 and field of view 2;

[0012] Step 3: Obtain the contour data set of the leaf basin and leaf back area, and plan the field of view of the leaf basin and leaf back area based on the optimization strategy of setting the dynamic threshold based on the change in the angle between the normal vectors of the contour points;

[0013] Step 3-1: Obtain the contour data set of the section to be measured based on step 1 and the starting and ending point coordinates of the leading and trailing edge fields of view calculated in step 2 , , , , integrate and remove the contour points of the leading edge and trailing edge areas, and obtain the contour data set of the leaf basin and leaf back area and its quantity P;

[0014] Step 3-2: Recalculate the normal vector of each contour point in the leaf base and leaf back area contour dataset The change in the angle between the normal vector and , and the change in the angle between the root normal vectors Preliminary planning of the field of view of the leaf basin and leaf back area; Preliminary determination of the minimum field of view number n required to detect the leaf basin and leaf back area, and the contour points in each field of view are ; The number of contour points in each field of view;

[0015] Step 3-3: Calculate the dynamic threshold value based on the maximum value of the angle change between normal vectors , minimum point Calculate the dynamic threshold of the angle change between the normal vectors of contour points , through the dynamic threshold Control the average distribution of the angle change between the normal vectors of the contour points between the planned fields of view, and guide the subsequent update of the field of view number;

[0016] Step 3-4: Update the field of view allocation; based on the dynamic threshold dz, update the number of fields of view and the number of contour points in the field of view. After the update, the contour points in each field of view are ,in Indicates the number of contour points of the nth field of view;

[0017] Step 3-5: Measurement range constraint: The distance between the start and end points of the contour data in each field of view should be less than the upper base width A of the sensor's trapezoidal measurement range;

[0018] Step 3-6: Iterative update; if the constraints in step 3-5 are not met, the field of view number is incremented, and the incremented field of view number is substituted back into step 3-2, the dynamic threshold dz is recalculated, and the field of view allocation is updated until the constraints in step 3-5 are met;

[0019] Finally, the number of fields of view for iterative update based on the global optimal strategy in the blade basin and back region of the section to be measured is determined to be n, and the contour point set in the field of view is: ,in Indicates the number of contour points of the nth field of view;

[0020] Step 4: Based on the planned field of view, optimize the sensor's pose in combination with the viewing angle;

[0021] Take the front edge field of view 1 as an example for optimization, and the optimization of other fields of view is the same;

[0022] Based on the starting and ending points of the leading edge field of view 1 planned in step 2 and , calculate the midpoint coordinates of two points , and calculate the average normal vector of all contour data between the start and end points , is the number of contour points under the leading edge field of view 1;

[0023] Based on midpoint coordinates Average normal vector to the leading edge field of view 1 Adjust the midpoint of the trapezoidal measurement range of the line laser sensor to the midpoint of the beginning and end points of the leading edge field of view 1 Coincident, at the same time, the vector of the sensor trapezoidal measurement range along the depth of field direction must coincide with the average normal vector of the front edge field of view 1 parallel to each other;

[0024] By performing a point-based Around the mean normal vector The rotation adjustment is performed to update the vector of the incident light, and the angle sum between all contour points and the corresponding incident light vector is calculated and minimized as the optimization target;

[0025] ;

[0026] In the formula, Indicates that the measurement range is based on points Around the mean normal vector The rotation angle range is , represents the number of contour points in the leading edge field of view 1, represents the normal vector of the i-th contour point, The incident light vector corresponding to the i-th contour point, and the incident light vector Changes with angle;

[0027] Filter out the included angle and the rotation angle corresponding to the minimum value from all calculation results , complete the optimal rotation angle corresponding to the viewing angle optimization; when the optimal rotation angle After being determined, based on the vector of the sensor laser surface in the depth direction Calculate The angle between the blade coordinate system o-xyz and the y-axis ; The rotation matrix of field of view 1 is: ;

[0028] Step 5: Based on the rotated field of view, optimize the sensor's position and posture in combination with the viewing distance;

[0029] Based on the rotated trapezoidal measurement range in step 4, calculate the start and end points of the leading edge field of view 1 , All contour points between The distance from the xd axis in the scan data coordinate system od-xdyd , so that it is minimized to meet the line of sight constraint condition, that is, satisfying:

[0030] ;

[0031] In the formula, express The projection distance of the cth point in , It represents the slope of the straight line corresponding to the yd axis in od-xdyd in the blade coordinate system o-xyz, G is the intercept of the straight line, and the equation of the straight line is: , , P1 and P2 are the vertices in the trapezoidal measurement range; Indicate point The distance between the point and the axis yd;

[0032] By updating the coordinates of the vertices of the trapezoidal measurement range at this time, the translation parameter T1 between the trapezoidal measurement range and the blade coordinate system o-xyz can be calculated;

[0033] Step 6: Confirmation of field of view posture parameters and field of view integration;

[0034] Calculate the planned field of view numbers of the leading and trailing edges of the blade section to be measured and the blade basin and back area respectively; based on step 2, the total number of fields of view of the leading and trailing edge areas is , where the field of view of the leading edge of the blade is f1 and f2, and the field of view of the trailing edge is b1 and b2; based on step 3, the field of view of the blade basin and back area is further subdivided ; The field of view number of the leaf basin area is , then the field of view number of the leaf back area is ; Perform steps 4 and 5 for each contour point in the planning field of view to calculate the first The rotation matrix of the field of view With the translation matrix , and then determine the relative pose parameters of the sensor relative to the initial field of view ;

[0035] The determined field of view posture parameters are sorted and processed to achieve the field of view integration of the leading and trailing edges of the blade, the blade basin, and the blade back area.

[0036] Furthermore, in step 1, the change in the angle between the normal vectors of the blade profile data to be measured is calculated based on the blade CAD model and the sensor parameters. The specific steps are:

[0037] The theoretical contour of the section S1 to be measured is extracted based on the blade CAD model. The spacing between contour points is determined based on the resolution Lx of the sensor in the X-axis direction. The theoretical contour of the section to be measured is discretized to obtain the contour data set of the section to be measured. , is the three-dimensional coordinate data set of the contour points after the CAD model of the section to be measured is discretized, M is the total number of contour points after the section to be measured is discretized, is the three-dimensional coordinate of the mth point on the contour;

[0038] Calculate the normal vector of each contour point in the contour data set of the section to be measured ,in is the normal vector of the mth point on the contour;

[0039] Calculate the normal vector angle between any two adjacent contour points to obtain the change in the angle between the normal vectors of the contour points ; is the change in the angle between the normal vectors of the mth contour point on the contour, is the normal vector of the m+1th point on the contour.

[0040] Furthermore, the normal vector of the mth contour point The specific calculation is as follows:

[0041] Calculate the backward tangent vector between the mth contour point and the m+1th contour point of the measured section , and the forward tangent vector between the m-1th contour point and the mth contour point ; Average and normalize the two tangent vectors to determine the average tangent vector of the mth contour point , reverse vertical rotation of the average tangent vector of the mth contour point Get the normal vector of the mth contour point .

[0042] Furthermore, the starting and ending points of the leading edge field of view in step 2 are calculated in the same direction as the starting and ending points of the leading edge field of view; the starting and ending points of the leading edge field of view are calculated in the following steps:

[0043] Select leading edge point Normal vector along the leading edge The reverse movement distance h, based on this moved point, draw a line with a slope perpendicular to The straight line is obtained, and the intersection of the straight line and the cross-sectional profile data is calculated to obtain the starting and ending points of the leading edge field of view. , .

[0044] Furthermore, the moving distance h is based on the minimum reverse extension distance as a reference value only, and during the scanning process of blades with different geometric features, it can be flexibly adjusted according to the shape, size and detection requirements of the blades; the distance between the starting and ending points of the contour data in the field of view planned based on the moving distance h must meet the upper width constraint of the trapezoidal measurement range.

[0045] Furthermore, the sorting rule in step 6 is ;

[0046] Further broken down into:

[0047] ;

[0048] in, Represents the relative pose parameters for the initial field of view, is the rotation matrix, is the translation matrix, represents the field of view of the leaf basin area, represents the field of view in the back area of ​​the leaf, represents the field of view in the leading edge region, Represents the field of view in the trailing edge region.

[0049] Compared with the prior art, the beneficial effects of the present invention are as follows: in view of the different precision requirements for different areas (such as the blade basin, blade back and leading and trailing edges) of complex blades during optical inspection, a consistent global constrained field of view planning method for blade profile line laser scanning is proposed in combination with the line of sight, viewing angle and measurement range of the sensor during sensor imaging; through this method, the accuracy, reliability and overall scanning efficiency of the scanning data during blade inspection can be effectively improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 Schematic diagram of the field of view and parameters of the line laser scanning sensor.

[0051] Figure 2 This is a schematic diagram of the process of step 1 of the present invention.

[0052] Figure 3 Schematic diagram of the change in the angle between the normal vectors of the contour points in step 2 of the present invention.

[0053] Figure 4 This is a schematic diagram of the leading edge field of view profile data in step 2 of the present invention.

[0054] Figure 5 This is a schematic diagram of the field of view of the leaf basin and leaf back area and the number of contour points evenly divided in step 3 of the present invention.

[0055] Figure 6 This is a schematic diagram of obtaining the initialization dz based on the change in the angle between normal vectors in step 3 of the present invention.

[0056] Figure 7 This is a schematic diagram of updating dz based on the change in the angle between normal vectors in step 3 of the present invention.

[0057] Figure 8 This is a schematic diagram of solving the midpoint and average normal vector in step 4 of the present invention.

[0058] Fig. 9 This is a schematic diagram of solving the angle in step 4 of the present invention.

[0059] Fig.10 It is a schematic diagram of the rotation range in step 4 of the present invention.

[0060] Fig.11 This is a schematic diagram of the optimal rotation angle in step 4 of the present invention.

[0061] Fig.12 This is a schematic diagram of step 5 of the present invention.

[0062] Fig.13 This is a schematic diagram of the process of step 6 of the present invention.

[0063] Markings in the figure: A and B are the upper and lower base widths of the sensor trapezoidal measurement area, respectively; Mr is the width along the depth of field direction; CD is the distance from the sensor light outlet to the near-end field of view; L d is the benchmark record; d -x d y d is the data coordinate system; is the change in the angle between normal vectors; S1 is the section to be measured; Lx is the resolution of the sensor in the X-axis direction; is the backward tangent direction between the mth contour point and the m+1th contour point of the measured section; is the forward tangent vector between the m-1th contour point and the mth contour point of the section to be measured; is the average tangent vector of the mth contour point; is the normal vector of the mth contour point; , , is the change in the angle between the normal vectors of the first, mth and last point M on the contour; is the mth point on the contour; , are the coordinates of the leading and trailing edge points; , are the starting and ending points of the angle change between the leading edge normal vectors; , are the starting and ending points of the change in the angle between the normal vectors of the trailing edge; h is the moving distance; , are the starting and ending points of the leading edge field of view; , Respectively represent the contour point sets under the leading edge field of view 1 and field of view 2; , are the starting and ending points of the trailing edge field of view; is the change in the angle between the normal vectors of each contour point; , is the change in the angle between the normal vectors The maximum and minimum points in ; , are the coordinates of the start and end points of the contour data within the field of view; is the distance between the start and end points of the contour data in the field of view; is the dynamic threshold; , are the coordinates of the start and end points of the contour data in the updated field of view; Indicates the number of contour points of the nth field of view; The starting and ending points of the leading edge field of view 1 and The midpoint coordinates of is the average normal vector; is the rotation angle; is the angle between the normal vector of the contour point and the incident light vector; is the incident light vector corresponding to the i-th contour point; is the projection distance; P1 and P2 are the vertices in the trapezoidal measurement range. DETAILED DESCRIPTION

[0064] Figure 1 is the field of view of the line laser scanning sensor and the parameter definition of the sensor in this embodiment; the measurement range of the sensor covers the area from the near-end field of view to the far-end field of view, which is defined as a trapezoidal measurement area; the upper base width of the trapezoidal area is recorded as A, the lower base width is B, and the width along the depth of field is Mr, and the distance from the sensor light outlet to the near-end field of view is CD. According to the imaging characteristics of the sensor, the measurement accuracy is the minimum angle between the normal vector of the contour point in the field of view and the incident light of the sensor and the reference distance L from the sensor. d Therefore, when planning the field of view, the contour points in the field of view are usually positioned as close to the reference distance as possible while being collinear with the incident light of the sensor, thereby improving the accuracy and reliability of the data. Figure 1 o in d -x d y d The coordinate system is the data coordinate system, and all raw data scanned by the sensor are represented based on this coordinate system.

[0065] This embodiment provides a consistent global constraint field of view planning method for blade profile line laser scanning, comprising the following steps:

[0066] Step 1: Calculate the change in the angle between the normal vector and the normal vector of the blade profile data set to be measured based on the blade CAD model and sensor parameters ;

[0067] like Figure 2 As shown, this embodiment extracts the theoretical contour of the section to be measured S1 based on the blade CAD model, and then determines the spacing between contour points in combination with the resolution Lx of the sensor in the X-axis direction, discretizes the theoretical contour of the section to be measured to obtain the coordinate value of the contour data.

[0068] Profile dataset of the section to be measured , is the three-dimensional coordinate data set of the contour points after the CAD model of the section to be measured is discretized, M is the total number of contour points after the section to be measured is discretized, is the three-dimensional coordinate of the mth point on the contour, are the coordinate values ​​of the point respectively.

[0069] Calculate the backward tangent vector between the mth contour point and the m+1th contour point of the measured section , and the forward tangent vector between the m-1th contour point and the mth contour point ; Average and normalize the two tangent vectors to determine the average tangent vector of the mth contour point , reverse vertical rotation of the average tangent vector of the mth contour point Get the normal vector of the mth contour point . The normal vector of each contour point is calculated by the above method ,in is the normal vector of the mth point on the contour.

[0070] In order to effectively reflect the geometric characteristics such as the change of blade curvature, the normal vector of any contour point is combined to calculate the angle between any two adjacent contour points, so as to obtain the change of the angle between the normal vectors of any contour points. ; is the change in the angle between the normal vectors of the mth contour point on the contour, The value range is ; is the normal vector of the m+1th point on the contour.

[0071] Step 2: Obtain the leading edge and trailing edge areas based on the change in the angle between the normal vectors of the contour data, and plan the fields of view of the leading edge and trailing edge;

[0072] Since the curvature of the leading and trailing edge areas of the blade varies greatly, the radius is small, and the detection accuracy requirement is high, the essence of field of view planning is to pre-select appropriate segments from the cross-section CAD contour for subsequent field of view planning.

[0073] The change in the angle between the normal vectors of the contour points calculated in step 1 , perform K-means clustering operation; based on the clustering results, extract the index and coordinate values ​​of the starting and ending points of the front and rear edge points and the angle change between the front and rear edge normal vectors; the coordinates of the front and rear edge points are expressed as , , the starting and ending points of the angle change between the leading edge normal vectors are expressed as , , the starting and ending points of the angle change between the trailing edge normal vectors are expressed as , ;like Figure 3 shown.

[0074] Select leading edge point Normal vector along the leading edge The reverse movement distance h, based on this moved point, draw a line with a slope perpendicular to The straight line is obtained, and the intersection of the straight line and the cross-sectional profile data is calculated to obtain the starting and ending points of the leading edge field of view. , ;like Figure 3 shown.

[0075] Therefore, the initial extraction and As the starting and ending points of the leading edge field of view 1, and As the starting and ending points of the leading edge field of view 2, Figure 4 As shown, , ,in, , Respectively represent the contour point sets under the leading edge field of view 1 and field of view 2, , Respectively represent the number of contour points in the leading edge field of view 1 and field of view 2.

[0076] The planning method of the trailing edge area is the same as that of the leading edge area, and the division of the trailing edge field of view adopts the same strategy as that of the leading edge, namely , ,in, , Respectively represent the contour point sets under the leading edge field of view 1 and field of view 2, , Respectively represent the number of contour points in the leading edge field of view 1 and field of view 2.

[0077] The total field of view number of the leading and trailing edge areas is ns. In the above steps of this embodiment, ns=4. It can be adjusted for different blades according to actual conditions. The field of view number can be flexibly adjusted according to the geometric characteristics, size and specific detection requirements of the blades to meet the requirements of efficient scanning and precise detection of different types of blades.

[0078] It should be noted that the moving distance h described in this embodiment is not fixed, but an adjustable parameter. In actual application, the value of h should be dynamically adjusted according to the geometric characteristics of the blade being measured and the scanning requirements. Specifically, in the field of view planning process, the minimum reverse extension distance is only used as a reference value for the design. In the scanning process of blades with different geometric characteristics, this parameter can be flexibly adjusted according to the shape, size and detection requirements of the blades, thereby meeting the efficient scanning and precise detection requirements of different types of blades. In addition, if the parameter h is adjusted, , Therefore, in the planning process, the value of parameter h needs to be strictly controlled and cannot be increased at will. It must be ensured that the distance between the start and end points of the contour data in the field of view planned based on parameter h always meets the upper length of the trapezoidal measurement range (i.e. Figure 1 The parameters A in the above formula are constrained to ensure the detection accuracy and reliability.

[0079] Step 3: Obtain the contour data set of the leaf basin and leaf back area, and plan the field of view of the leaf basin and leaf back area based on the optimization strategy of setting the dynamic threshold based on the change in the angle between the normal vectors of the contour points;

[0080] Based on step 1, obtain the contour data set of the section to be measured and the starting and ending point coordinates of the leading and trailing edge fields of view calculated in step 2 , , , , integrate and remove the contour points of the leading edge and trailing edge areas, and obtain the leaf basin and leaf back area contour data set and its quantity P, as shown in Figure 5 shown.

[0081] Then, recalculate the normal vector of each contour point in the leaf basin and leaf back area contour dataset The change in the angle between the normal vector and , and the change in the angle between the root normal vectors Preliminary planning of the field of view of the leaf basin and leaf back area. At the same time, obtain the angle change between the normal vectors The maximum point in With minimum point .

[0082] In order to ensure that the number of leaf basin and leaf back regions is evenly distributed in different fields of view, reduce the fluctuation of detection error, and improve detection accuracy; this embodiment preliminarily determines the minimum number of fields of view n required to detect the leaf basin and leaf back region based on the number of contour points in the leaf basin and leaf back region and the measurement principle of the sensor, such as Figure 5 As shown in the figure, the contour points of the leaf basin and leaf back area are evenly distributed to each field of view, ensuring that the number of contour points in each field of view is , the coordinates of each contour point in the field of view are recorded as Due to the limited measurement range of the sensor, one scan cannot simultaneously obtain all the contour data in the leaf basin and leaf back area, so the initial value of the minimum field of view number n is set to 2.

[0083] In addition, the distance between the start and end points of the contour data in the field of view It must also be smaller than the upper length A of the trapezoidal measurement range of the line laser sensor, that is, it satisfies If this constraint is not met, the number of fields of view n=n+1 is increased and the remaining contour points are evenly divided again until the measurement range constraint is met, thereby initially determining the number of planned fields of view and the number of contour points in each field of view.

[0084] However, the method of simply dividing the number of remaining contour points equally ignores the influence of geometric features such as the change in leaf curvature, which may lead to uneven curvature distribution between different fields of view, thereby reducing the detection accuracy. With minimum point , combined with the field of view number n, calculate the dynamic threshold dz based on the change in the angle between the normal vectors of the contour points, and dynamically adjust the number of initially planned fields of view and the number of contour points in the field of view.

[0085] Specifically, dynamically adjusting the number of initially planned fields of view and the number of contour points within the fields of view includes the following steps:

[0086] Step 3-1: Calculate the dynamic threshold; Figure 6 As shown in the figure, the number of contour points in the leaf basin and leaf back area is evenly distributed in each field of view, based on the maximum value of the angle change between the normal vectors. , minimum point Calculate the dynamic threshold of the angle change between the normal vectors of contour points , through the dynamic threshold Control the average distribution of the angle change between the normal vectors of the contour points between the planned fields of view, and guide the subsequent update of the field of view number.

[0087] Step 3-2: Update the field of view allocation; Figure 6 As shown, based on the dynamic threshold dz, the number of fields of view and the number of contour points in the field of view are updated. After the update, the coordinates of the contour points in each field of view are recorded as ,in Indicates the number of n-th field of view contour points.

[0088] Step 3-3: Measurement range constraint: The distance between the start and end points of the contour data in each field of view should be less than the upper base A of the sensor's trapezoidal measurement range, that is, .

[0089] Step 3-4: Iterative update; if the constraint of step 3-3 is not met, the field of view number n1=n1+1 is incremented, and the incremented field of view number n1 is substituted back into step 3-1 (i.e., n=n1), as shown in 7, the dynamic threshold dz is recalculated, and then steps 3-1, 3-2, 3-3, and 3-4 are iterated until the constraint of step 3-3 is met.

[0090] Through the above steps, if Figure 7 As shown, the number of fields of view for iterative update of the blade basin and back region of the section to be measured based on the global optimal strategy is finally determined to be n1, and the contour point set in the field of view is expressed as: ,in Represents the number of contour points in the nth field of view. It should be noted that although the number of contour points between different fields of view may not be completely evenly divided at this time, by dynamically adjusting the threshold dz of the angle change between normal vectors, the consistency of the scanning state and measurement uncertainty between different fields of view is ensured, further improving the detection accuracy.

[0091] Step 4: Based on the planned field of view, optimize the sensor's pose in combination with the viewing angle;

[0092] This embodiment will optimize the sensor imaging posture based on the viewing angle and viewing distance; since the contour data of each area has been segmented, the viewing angle and viewing distance optimization method described in this embodiment will be applicable to the contour data under all fields of view.

[0093] Based on the starting and ending points of the leading edge field of view 1 planned in step 2 and , calculate the midpoint coordinates of two points , and calculate the average normal vector of all contour data between the start and end points , is the number of contour points under the leading edge field of view 1; based on the midpoint coordinates Average normal vector to the leading edge field of view 1 Adjust the midpoint of the trapezoidal measurement range of the line laser sensor to the midpoint of the beginning and end points of the leading edge field of view 1 Overlap, such as Figure 8 As shown, the vector of the sensor's trapezoidal measurement range along the depth of field direction must be aligned with the average normal vector of the front edge field of view 1 To ensure the detection accuracy of the scanned data, Fig. 9As shown, when optimizing the field of view, it is necessary to ensure that the angle between the normal vector of each contour point and the incident light vector corresponding to the point is as small as possible.

[0094] However, simply relying on the average normal vector of all contour data between the start and end points of the field of view to determine the trapezoidal measurement range often cannot effectively ensure that the angle between each contour point and the incident light corresponding to the point is the smallest. Around the mean normal vector Rotation adjustment, such as Fig.10 As shown in the figure, the vector of the incident light is updated to optimize the angle between each contour point and the corresponding incident light. Specifically, by updating the vector of the incident light, the sum of the angles between all contour points and the corresponding incident light vector is calculated and used as the optimization target for minimization. The core of this process is to find the sensor-based point that minimizes the sum of the angles. Around the mean normal vector Rotation angle ; that is, satisfy:

[0095] ;

[0096] In the formula, Indicates that the measurement range is based on points Around the mean normal vector The rotation angle range is , represents the number of contour points in the leading edge field of view 1, represents the normal vector of the i-th contour point, The incident light vector corresponding to the i-th contour point, and the incident light vector The angle changes.

[0097] The calculation process is as follows: Calculate the coordinates of the i-th contour point based on the sensor coordinate system , and calculate the coordinates of the sensor light outlet based on the sensor , the incident light vector corresponding to the i-th contour point pass and Two-point coordinate calculation; specifically, .

[0098] Represents the angle between the normal vector of the contour point and the incident light vector, and its range is .

[0099] Therefore, the core purpose of viewing angle optimization is to determine the optimal rotation angle. ; Since the normal vector of the contour point is known, the rotation angle The change of will change the incident light vector, thus affecting the angle and between the normal vector of the contour point and the corresponding incident light vector. Within the rotation range, traverse with a step size of 1° All possible values ​​of , and each traversal calculates the angle between the normal vector of each contour point within the trapezoidal measurement range and its corresponding nearest incident light, such as Fig.10 After the traversal is completed, the included angle and the rotation angle corresponding to the minimum value are selected from all the calculation results. , which is the optimal rotation angle corresponding to completing the viewing angle optimization.

[0100] When the optimal rotation angle After being determined, the viewing angle optimization of the leading edge field of view 1 is completed, such as Fig.11 As shown. Optimal rotation angle The determination of also means that the trapezoidal measurement range of the line laser sensor is determined, and the coordinates of the four vertices of the sensor's measurement range will be updated accordingly. Based on these updated coordinate values, the vector of the sensor laser surface in the depth direction is further calculated. At this point, the view angle optimization is complete, and the rotation process of the trapezoidal measurement range has ended. The next step only requires positive or negative translation in the updated depth of field vector direction without rotation adjustment. The field of view of the trailing edge, blade basin, and blade back area adopts the same view angle optimization strategy as above.

[0101] Finally, based on calculate The angle between the blade coordinate system o-xyz and the y-axis ; At this time, the rotation matrix of field of view 1 is: .

[0102] Step 5: Based on the rotated field of view, optimize the sensor's position and posture in combination with the viewing distance;

[0103] Based on the rotated trapezoidal measurement range in step 4, calculate the start and end points of the leading edge field of view 1 , All contour points between The distance between the xd axis and the scan data coordinate system od-xdyd, i.e. the projection distance ,express The distance between point c and the xd axis is Fig.12 As shown in the figure. To ensure the accuracy of the scanning data, the contour points should be placed as close to the sensor reference distance as possible during the field of view planning. However, due to the complex geometric characteristics of the blade cross section, it is difficult to effectively ensure that each contour point is placed near the sensor reference distance. Therefore, the average value of the projection distance of all contour points in the field of view is used as an indicator to minimize it to solve the scanning posture, that is, the line of sight constraint.

[0104] Since the contour point set in the field of view All coordinates of are based on the blade coordinate system o-xyz. Therefore, by calculating The average projection distance of all contour points in the image is minimized to meet the viewing distance constraint condition, that is, it satisfies:

[0105] ;

[0106] In the formula, express The projection distance of the cth point in , It represents the slope of the straight line corresponding to the yd axis in the blade coordinate system o-xyz in od-xdyd, and G is the intercept of the straight line, that is, the equation of the straight line can be expressed as: . Indicate point The distance between a point and the axis yd. In the equation of a straight line, It can be solved by the vertices P1 and P2 in the trapezoidal measurement range, that is, .

[0107] Therefore, the core purpose of the sight distance constraint is to solve the optimal intercept G. All coordinate values ​​of are known, and the optimal intercept G can be directly solved by the least squares method to minimize The sum of the projection distances of all contour points to the straight line. That is, when the optimal viewing distance optimization condition is clearly met, the spatial position relationship between the trapezoidal measurement range and the contour data has been clarified, thereby determining the field of view posture corresponding to the "viewing angle-viewing distance" optimization, such as Fig.12 Finally, by updating the coordinates of the vertices of the trapezoidal measurement range at this time, the translation parameter T1 between the trapezoidal measurement range and the blade coordinate system o-xyz can be calculated.

[0108] Steps 4 and 5 respectively calculate the trapezoidal measurement range rotation parameter R1 corresponding to the optimal viewing angle in the leading edge field of view 1 and the translation parameter T1 corresponding to the optimal viewing distance from the perspectives of viewing angle and viewing distance, thus completing the "viewing angle-viewing distance" optimization.

[0109] The field of view of the trailing edge, blade basin, and blade back area adopts the same "viewing angle-viewing distance" optimization strategy as mentioned above.

[0110] Step 6: Confirmation of field of view posture parameters and field of view integration;

[0111] Determination of the field of view position parameters: T and R represent the translation axis and rotation axis of the scanning system, respectively, which are used to adjust the relative position between the sensor and the blade, that is, the field of view Based on T and R, the sensor is driven to move from the initial position to the planned field of view position to complete the acquisition of blade cross-sectional profile data under different fields of view. Fig.13 As shown, represents the pose parameters of the scanning sensor under the field of view n, represents the rotation matrix of the blade based on the ob-yb axis of the inertial coordinate system in the field of view n, It represents the translation vector of the sensor between the nth field of view and the initial field of view. For any other field of view, the pose parameters , the definition rules are consistent.

[0112] Integration of the field of view of the leading and trailing edges, blade basin, and blade back areas: Calculate the planned field of view numbers of the leading and trailing edges, blade basin, and blade back areas of the blade section to be measured. Based on step 2, the total number of fields of view of the leading and trailing edges is , where the field of view of the leading edge of the blade is f1 and f2, and the field of view of the trailing edge is b1 and b2; based on step 3, the field of view of the blade basin and back area is further subdivided ; indicates the field of view number of the leaf basin area is , then the field of view number of the leaf back area is On this basis, steps 4 and 5 are performed for each contour point in the planning field of view, such as Figure 6 As shown, calculate the The rotation matrix of the field of view With the translation matrix , and then determine the relative pose parameters of the sensor relative to the initial field of view .

[0113] In order to ensure the continuity of scanning during the detection process of the line laser sensor, it is necessary to sort the determined field of view posture parameters to achieve the field of view integration of the leading and trailing edges of the blade and the blade basin and back area. The rule is: .

[0114] Further broken down into:

[0115] ;

[0116] in, Represents the relative pose parameters for the initial field of view, is the rotation matrix, is the translation matrix, represents the field of view of the leaf basin area, represents the field of view in the back area of ​​the leaf, represents the field of view in the leading edge region, Represents the field of view in the trailing edge region.

[0117] It should be noted that the scanning order of the blade basin, leading edge, blade back, and trailing edge regions is not unique. The actual scanning order is determined by the region where the initial point is located and the scanning direction.

[0118] The above description is only a preferred implementation manner of the present invention, but the protection scope of the present invention is not limited thereto, and any modification and replacement based on the technical solution and inventive concept provided by the present invention should be included in the protection scope of the present invention.

Claims

1. A consistent global constraint field of view planning method for blade profile line laser scanning, characterized in that: The steps include: Step 1: Calculate the change in the angle between the normal vector and the normal vector of the blade profile data set to be measured based on the blade CAD model and sensor parameters ; Step 2: Obtain the leading edge and trailing edge areas based on the change in the angle between the normal vectors of the contour data, and plan the fields of view of the leading edge and trailing edge; The change in the angle between the normal vectors of the contour points calculated in step 1 , perform clustering operation; Based on the clustering results, extract the leading and trailing edge points , , the starting and ending points of the angle change between the normal vectors of the leading edge point , , and the starting and ending points of the angle change between the normal vectors of the trailing edge points , ; Based on leading edge and its normal vector Calculate the beginning and end points of the leading edge field of view , ;extract and As the starting and ending points of the leading edge field of view 1, and As the starting and ending points of the leading edge field of view 2, that is, , ,in, , Respectively represent the contour point sets under the leading edge field of view 1 and field of view 2; Based on the trailing edge and its normal vector Calculate the beginning and end points of the leading edge field of view , ; The trailing edge field of view is divided into: , ,in, , Respectively represent the contour point sets under the leading edge field of view 1 and field of view 2; Step 3: Obtain the contour data set of the leaf basin and leaf back area, and plan the field of view of the leaf basin and leaf back area based on the optimization strategy of setting the dynamic threshold based on the change in the angle between the normal vectors of the contour points; Step 3-1: Obtain the contour data set of the section to be measured based on step 1 and the starting and ending point coordinates of the leading and trailing edge fields of view calculated in step 2 , , , , integrate and remove the contour points of the leading edge and trailing edge areas, and obtain the contour data set of the leaf basin and leaf back area and its quantity P; Step 3-2: Recalculate the normal vector of each contour point in the leaf base and leaf back area contour dataset The change in the angle between the normal vector and the , and the change in the angle between the root normal vectors Preliminary planning of the field of view of the leaf basin and leaf back area; Preliminary determination of the minimum field of view number n required to detect the leaf basin and leaf back area, and the contour points in each field of view are ; The number of contour points in each field of view; Step 3-3: Calculate the dynamic threshold value based on the maximum value of the angle change between normal vectors , minimum point Calculate the dynamic threshold of the angle change between the normal vectors of contour points , through the dynamic threshold Control the average distribution of the angle change between the normal vectors of the contour points between the planned fields of view, and guide the subsequent update of the field of view number; Step 3-4: Update the field of view allocation; based on the dynamic threshold dz, update the number of fields of view and the number of contour points in the field of view. After the update, the contour points in each field of view are ,in Indicates the number of contour points of the nth field of view; Step 3-5: Measurement range constraint: The distance between the start and end points of the contour data in each field of view should be less than the upper base width A of the sensor's trapezoidal measurement range; Step 3-6: Iterative update; if the constraints in step 3-5 are not met, the field of view number is incremented, and the incremented field of view number is substituted back into step 3-2, the dynamic threshold dz is recalculated, and the field of view allocation is updated until the constraints in step 3-5 are met; Finally, the number of fields of view for iterative update based on the global optimal strategy in the blade basin and back region of the section to be measured is determined to be n, and the contour point set in the field of view is: ,in Indicates the number of contour points of the nth field of view; Step 4: Based on the planned field of view, optimize the sensor's pose in combination with the viewing angle; Take the front edge field of view 1 as an example for optimization, and the optimization of other fields of view is the same; Based on the starting and ending points of the leading edge field of view 1 planned in step 2 and , calculate the midpoint coordinates of two points , and calculate the average normal vector of all contour data between the start and end points , is the number of contour points under the leading edge field of view 1; Based on midpoint coordinates Average normal vector to the leading edge field of view 1 Adjust the midpoint of the trapezoidal measurement range of the line laser sensor to the midpoint of the beginning and end points of the leading edge field of view 1 Coincident, at the same time, the vector of the sensor trapezoidal measurement range along the depth of field direction must coincide with the average normal vector of the front edge field of view 1 parallel to each other; By performing a point-based Around the mean normal vector The rotation adjustment is performed to update the vector of the incident light, and the angle sum between all contour points and the corresponding incident light vector is calculated and minimized as the optimization target; ； In the formula, Indicates that the measurement range is based on points Around the mean normal vector The rotation angle range is , represents the number of contour points in the leading edge field of view 1, represents the normal vector of the i-th contour point, The incident light vector corresponding to the i-th contour point, and the incident light vector Changes with angle; Filter out the included angle and the rotation angle corresponding to the minimum value from all calculation results , complete the optimal rotation angle corresponding to the viewing angle optimization; when the optimal rotation angle After being determined, based on the vector of the sensor laser surface in the depth direction Calculate The angle between the blade coordinate system o-xyz and the y-axis ; The rotation matrix of field of view 1 is: ； Step 5: Based on the rotated field of view, optimize the sensor's position and posture in combination with the viewing distance; Based on the rotated trapezoidal measurement range in step 4, calculate the start and end points of the leading edge field of view 1 , All contour points between The distance from the xd axis in the scan data coordinate system od-xdyd , so that it is minimized to meet the line of sight constraint condition, that is, satisfying: ； In the formula, express The projection distance of the cth point in , It represents the slope of the straight line corresponding to the yd axis in od-xdyd in the blade coordinate system o-xyz, G is the intercept of the straight line, and the equation of the straight line is: , , P1 and P2 are the vertices in the trapezoidal measurement range; Indicate point The distance between the point and the axis yd; By updating the coordinates of the vertices of the trapezoidal measurement range at this time, the translation parameter T1 between the trapezoidal measurement range and the blade coordinate system o-xyz can be calculated; Step 6: Confirmation of field of view posture parameters and field of view integration; Calculate the planned field of view numbers of the leading and trailing edges of the blade section to be measured and the blade basin and back area respectively; based on step 2, the total number of fields of view of the leading and trailing edge areas is , where the field of view of the leading edge of the blade is f1 and f2, and the field of view of the trailing edge is b1 and b2; based on step 3, the field of view of the blade basin and back area is further subdivided ; The field of view number of the leaf basin area is , then the field of view number of the leaf back area is ; Perform steps 4 and 5 for each contour point in the planning field of view to calculate the first The rotation matrix of the field of view With the translation matrix , and then determine the relative pose parameters of the sensor relative to the initial field of view ; The determined field of view posture parameters are sorted and processed to achieve the field of view integration of the leading and trailing edges of the blade, the blade basin, and the blade back area.

2. The consistent global constraint field of view planning method for blade profile line laser scanning according to claim 1 is characterized in that: In step 1, the change in the angle between the normal vectors of the blade profile data to be measured is calculated based on the blade CAD model and sensor parameters. The specific steps are: The theoretical contour of the section S1 to be measured is extracted based on the blade CAD model. The spacing between contour points is determined based on the resolution Lx of the sensor in the X-axis direction. The theoretical contour of the section to be measured is discretized to obtain the contour data set of the section to be measured. , is the three-dimensional coordinate data set of the contour points after the CAD model of the section to be measured is discretized, M is the total number of contour points after the section to be measured is discretized, is the three-dimensional coordinate of the mth point on the contour; Calculate the normal vector of each contour point in the contour data set of the section to be measured ,in is the normal vector of the mth point on the contour; Calculate the normal vector angle between any two adjacent contour points to obtain the change in the angle between the normal vectors of the contour points ; is the change in the angle between the normal vectors of the mth contour point on the contour, is the normal vector of the m+1th point on the contour.

3. The consistent global constraint field of view planning method for blade profile line laser scanning according to claim 2 is characterized in that: The normal vector of the mth contour point The specific calculation is as follows: Calculate the backward tangent vector between the mth contour point and the m+1th contour point of the measured section , and the forward tangent vector between the m-1th contour point and the mth contour point ; Average and normalize the two tangent vectors to determine the average tangent vector of the mth contour point , reverse vertical rotation of the average tangent vector of the mth contour point Get the normal vector of the mth contour point .

4. The consistent global constraint field of view planning method for blade profile line laser scanning according to claim 1 is characterized in that: The starting and ending points of the leading edge field of view in step 2 are calculated in the same direction as the starting and ending points of the leading edge field of view; the starting and ending points of the leading edge field of view are calculated as follows: Select leading edge point Normal vector along the leading edge The reverse movement distance h, based on this moved point, draw a line with a slope perpendicular to The straight line is obtained, and the intersection of the straight line and the cross-sectional profile data is calculated to obtain the starting and ending points of the leading edge field of view. , .

5. The consistent global constraint field of view planning method for blade profile line laser scanning according to claim 4 is characterized in that: The moving distance h is based on the minimum reverse extension distance as a reference value only, and in the process of scanning blades with different geometric features, it can be flexibly adjusted according to the shape, size and detection requirements of the blades; the distance between the starting and ending points of the contour data in the field of view planned based on the moving distance h must meet the upper width constraint of the trapezoidal measurement range.

6. The consistent global constraint field of view planning method for blade profile line laser scanning according to claim 1 is characterized in that: The sorting rules in step 6 are , Further broken down into: ； in, Represents the relative pose parameters for the initial field of view, is the rotation matrix, is the translation matrix, represents the field of view of the leaf basin area, represents the field of view in the back area of ​​the leaf, represents the field of view in the leading edge region, Represents the field of view in the trailing edge region.

Images