CN113591236B - Aviation blade cross section molded line profile parameter evaluation method and system - Google Patents

Abstract

The invention discloses an aviation blade cross section molded line profile parameter evaluation method and system, and belongs to the field of aviation blade detection. Comprising the following steps: performing ICP matching on the theoretical data point set P and the leaf profile contour point set Q to obtain an initial matching result; q is divided into a front edge, a rear edge, a leaf basin and a leaf back sub-point set, and tolerance zone ranges are respectively set for the sub-point sets; giving coefficients to each measuring point in each sub-point set according to the position relation between the measuring point and the corresponding tolerance zone range; taking the x offset, the y offset and the torsion angle of Q as the quantity to be optimized, and taking the coefficient of each measuring point as the weight to construct an objective function with position constraint; the initial matching result is an initial value, and the objective function is solved to obtain an optimal matching result; and (3) carrying out rigid transformation by utilizing the optimal matching result, and comparing the aviation blade subjected to rigid transformation with a design model to judge whether the aviation blade is qualified or not. According to the invention, different coefficients are given to measuring points in different areas, a target function with position constraint is constructed, and torsion angles and offset are introduced, so that the actual detection requirements are more met.

Description

A method and system for evaluating the profile parameters of the cross-section of an aviation blade

Technical Field

The present invention belongs to the field of aviation blade detection, and more specifically, relates to a method and system for evaluating profile parameters of a cross-section profile of an aviation blade.

Background Art

Aircraft engines are key components of aircraft, and aircraft engine blades are the main components that determine engine performance, and have a great impact on the power, reliability and economy of aircraft flight. Therefore, the requirements for aircraft blade surface quality inspection are very strict.

Patent CN111400667A discloses a method and system for detecting the profile of an aviation blade based on a variable tolerance band constraint. The main idea is to estimate the curvature of all points in the blade profile measurement point set of the aviation blade to obtain the curvature distribution, and to perform downsampling in the blade profile measurement point set based on the curvature distribution to obtain the sampled blade profile measurement points; to perform random fitting circles using the sampled blade profile measurement points to segment the blade profile and obtain the leading edge point set, the trailing edge point set, the blade basin point set and the blade back point set; and to search for the points in the profile measurement point set of the designed blade. The nearest point of each point in the blade profile measurement point set is used to form the nearest point set. The blade profile measurement point set and the nearest point set are used for unconstrained matching to obtain the initial pose. The tolerance range and linear constraints are set for the leading edge point set, the trailing edge point set, the blade basin point set and the blade back point set respectively. The target matching function is constructed, and the initial pose is used as the initial value to solve the target matching function to obtain the optimal pose. The optimal pose is used to perform a rigid body transformation on the aviation blade, and the aviation blade after the rigid body transformation is compared with the designed blade to determine whether the aviation blade profile is qualified. However, there are the following defects: the Lagrange multiplier method is used to solve the objective function, which has many parameters and complex calculations. At the same time, the coupling relationship between the position error, torsion error and profile error is not considered, which will affect the final matching evaluation results.

Patent CN106407502A discloses a method for evaluating the profile parameters of a blade cross-section profile based on optimal matching. The main idea is: read in the theoretical data and measured data of the blade cross-section profile, and fit the profiles respectively; perform data preprocessing: according to the distribution characteristics of the profile data points, divide the data points into four parts: blade basin, blade back, leading edge and trailing edge, and store them separately; use the mid-arc line as a matching feature, perform a preliminary match between the measured data and the theoretical data, and remove the gross error data; use the optimal matching algorithm to accurately match the measured data with the theoretical data; based on the matching results, calculate the torsion angle Ψ and position w of the blade through the rotation and translation matrix required for the rigid body transformation, and use the limit profile error values of the blade basin, blade back, leading edge and trailing edge to obtain the profile profile; generate an error tolerance band with the standard shape as the skeleton according to the given tolerance parameters; and perform parameter evaluation and analysis on the profile of the blade cross-section profile according to the error tolerance band. However, there are the following defects: the torsion constraint and position constraint are not added to the actual matching process, but only the torsion error and position error are detected after each matching is completed to see whether they meet the detection requirements, and the blade matching process is separated from the constraint process of the offset and torsion angle.

In actual engineering inspections, the following evaluation method is often used for blade profile parameters, as proposed in the literature "Blade Model Registration Method Constrained by Contour Tolerance [J]. Computer Integrated Manufacturing Systems": ICP matching is performed between the blade measurement points and the blade design model, and then the blade is judged to be qualified by judging whether the contour error value of each point in the measurement point set is within the corresponding tolerance range. However, this method does not take into account that the tolerances of the leading edge, trailing edge, blade basin, and blade back of aviation blades are often different, that is, the accuracy requirements of each area are different, resulting in the error values of points in areas with small tolerances often exceeding the tolerance, thereby causing distortion of the matching results. At the same time, the torsion angle and offset of the blade matching are not evaluated, so that the blades with unqualified torsion angles and offsets may be misjudged as qualified when tested using existing methods.

Summary of the invention

In view of the defects of the prior art and the need for improvement, the present invention provides a method and system for evaluating the profile parameters of the cross-section profile of an aviation blade, with the aim of providing an effective and feasible parameter evaluation method, thereby reducing the number of deviations after alignment, thereby reducing the false scrap rate and preventing misjudgment of blade inspection results.

To achieve the above object, according to a first aspect of the present invention, a method for evaluating profile parameters of a cross-section profile of an aircraft blade is provided, the method comprising:

S1. Discretize the cross-sectional design model of the aviation blade to be evaluated into a theoretical data point set P, measure the measurement point data of the corresponding cross-sectional area of the aviation blade to be evaluated, and form a blade profile point set Q;

S2. Perform ICP matching using P and Q to obtain an initial matching result;

S3. segment the leaf profile, and divide Q into a leading edge sub-point set, a trailing edge sub-point set, a leaf basin sub-point set, and a leaf back sub-point set;

S4. For each measuring point in each sub-point set, find the corresponding nearest point in P, calculate the point-to-point distance between the measuring point and the corresponding nearest point, set the tolerance range for each sub-point set according to the profile evaluation standard, and assign a distance coefficient to each measuring point according to the positional relationship between the point-to-point distance and the corresponding tolerance range;

S5. Using the x-offset, y-offset and torsion angle of Q as decision variables and the distance coefficient of each measuring point as weight, an objective function with position constraints is constructed;

S6. Using the initial matching result as the initial value of the decision variable, solving the objective function, and obtaining the optimal matching result;

S7. Perform a rigid body transformation on the cross section of the aviation blade using the optimal matching result, compare the cross section of the aviation blade after the rigid body transformation with the design model, and determine whether the cross section profile of the aviation blade is qualified.

Preferably, step S3 comprises:

S31. Perform cubic spline interpolation on the measured points in Q to obtain a ^C2 -continuous curve L(x), discretize the curve according to the equal arc length principle, and obtain a discretized point set Q′;

S32. Calculate the curvature of each measuring point in Q' Among them, _Li ′ and _Li ″ represent the first-order derivative and the second-order derivative of the i-th measuring point on the curve L(x), respectively. According to the characteristic that the curvature of the leading and trailing edge regions is much greater than the curvature of the blade basin and blade back regions, the blade basin sub-point set, blade back sub-point set, leading edge sub-point set, and trailing edge sub-point set are extracted from Q.

Preferably, step S4 comprises:

S41. For each measurement point in the Q neutron point set, find the nearest point in P;

S42. Calculate the point-to-point distance between the measured point in Q and the corresponding nearest point in P;

S43. If the point is outside the theoretical blade profile, the directed distance s between the measuring point _qi in Q and the corresponding nearest point _pi in P is ＝|| _qi - _pi ||; if the point is inside the theoretical blade profile, the directed distance s is ＝-|| _qi - _pi ||;

S44. According to the profile evaluation standard, a tolerance zone range is set for each sub-point set. If the i-th measuring point falls within the corresponding tolerance zone range, the distance coefficient w _i of the measuring point is 1; if the i-th measuring point falls outside the corresponding tolerance zone range, the distance coefficient w _i of the point is greater than 1, which is jointly determined by the directed distance and the tolerance boundary.

Preferably, the calculation formula of the measuring point distance coefficient is as follows:

Among them, _wi represents the coefficient of the i-th measuring point _qi in Q, _si represents the directed distance of the i-th measuring point _qi in Q, U represents the upper limit of the tolerance band of the subset to which _qi belongs, and L represents the lower limit of the tolerance band of the subset to which _qi belongs.

Beneficial effect: The present invention assigns a distance coefficient to each point in Q in the objective function. Since the size of the distance coefficient depends on the point-to-point distance between the tolerance band range corresponding to the point and the corresponding nearest point, the constraint of the blade profile variable tolerance is taken into account, so that the final matching evaluation result of the blade profile can better meet the design requirements.

Preferably, the objective function with position constraint is as follows:

Among them, x′, y′, θ represent the x offset, y offset and torsion angle of Q respectively, _wi represents the coefficient of the i-th measuring point _qi in Q, M represents the number of measuring points in Q, _qix , _qiy represent the x coordinate and y coordinate of measuring point _qi respectively, _pix , _piy represent the x coordinate and y coordinate of measuring point _pi in P which is closest to measuring point _qi respectively, _x2 , _x1 represent the upper and lower tolerances of the offset respectively, _y2 , _y1 represent the upper and lower tolerances of the y offset respectively, _θ2 , _θ1 represent the upper and lower tolerances of the torsion angle respectively, and the above upper and lower tolerances are set according to the surface offset tolerance assessment standard.

Beneficial effect: The present invention adds constraints of x offset, y offset and torsion angle in the offset during the blade matching process. Since the offset and torsion angle will affect the matching state of the blade, the matching evaluation result obtained by this method is more in line with the tolerance design requirements.

Preferably, step S6 comprises:

S61. Take the initial matching result _x0 as the initial parameter and construct a quadratic model for the objective function:

S62. Set the number of iterations K and the precision ε, and initialize the number of iterations k = 0;

S63. When the number of iterations reaches k=K or the norm of the projected gradient ||P(x _k -g _k ,l,u)-x _k || _∞ is less than the precision ε, the iteration ends; wherein l=(x ₁ ,y ₁ ,γ1 ₁ ),u=(x ₂ ,y ₂ ,γ ₂ );

S64. Use the CauthyPoint method to obtain the approximate solution x _c of the Cauchy point for the quadratic model;

S65. First, ignore the constraints, start from x _c and perform the subspace minimization operation of the quadratic model of the objective function, and then satisfy the constraints by tracing the points back to the feasible region, and obtain the search direction d _k ;

S66. Perform a line search loop in the new search direction d _k until |g _k+1 ^T d _k |≤0.9|g _k ^T d _k | is satisfied, and x is updated in each loop to x _k+1 =x _k +λ _k d _k , where the initial value of λ is 1 and decreases linearly;

S67. Update the Hessian matrix and jump to step S63.

Beneficial effects: The present invention adopts a nonlinear optimization method to solve the constrained objective function, thereby obtaining a final matching result that takes into account the profile variation tolerance, position tolerance, and torsion tolerance.

Preferably, if all errors of the blade profile cross section meet the tolerance requirements, the evaluation result is qualified.

To achieve the above-mentioned object, according to a second aspect of the present invention, there is provided an aviation blade cross-section profile parameter evaluation system, comprising: a computer-readable storage medium and a processor;

The computer-readable storage medium is used to store executable instructions;

The processor is used to read the executable instructions stored in the computer-readable storage medium to execute the method for evaluating the profile parameters of the cross-section profile of an aviation blade according to the first aspect.

In general, the above technical solutions conceived by the present invention can achieve the following beneficial effects:

(1) In view of the variable tolerance of blades, the present invention sets different tolerance ranges for the leading edge, trailing edge, blade basin and blade back regions, and assigns different distance coefficients to each point according to whether the point is within the tolerance range, thereby constructing a new target matching function, so that the blade measurement model and the blade design model can be effectively aligned.

(2) After constructing the objective function based on the constraint area, the present invention adds constraints on the torsion angle and offset, which is more in line with the actual blade profile detection requirements. And using the result of ICP matching as the initial value, the constrained optimization method is used to solve the problem, ensuring that the blade profile error assessment is accurate and reliable.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG1 is a flow chart of a method for evaluating profile parameters of a cross-sectional profile of an aviation blade provided by the present invention.

FIG. 2 is a schematic diagram of a tolerance range provided by an embodiment of the present invention.

FIG. 3 is a schematic diagram of constraints of a torsion angle, an x offset, and a y offset provided by an embodiment of the present invention.

DETAILED DESCRIPTION

In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention. In addition, the technical features involved in the various embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

As shown in FIG1 , the present invention provides a method for evaluating the profile parameters of the cross-section of an aviation blade, comprising the following steps:

(1) Using the three-dimensional coordinate measuring machine method in contact measurement to obtain the measurement point data of each section of the blade, forming a blade profile point set Q;

(2) Using the ICP algorithm to find the point-pair relationship between the theoretical data point sets P and Q, and solving the ICP to obtain the initial result x ₀ of blade matching;

(3) According to the upper and lower deviation values of the blade basin, blade back, leading edge and trailing edge areas, set the tolerance range in each area;

(4) For each measuring point in the blade profile point set Q, different distance coefficients w are set according to the point-to-point distance between the measuring point and the nearest point and whether it is within the tolerance band.

(5) Add constraints on blade position errors such as x-offset, y-offset, and torsion angle, and construct a new objective function based on the distance coefficient of the measuring point and the initial objective function:

(6) Taking x ₀ =(x' ₀ ,y' ₀ ,θ ₀ ) as the initial parameter, the new objective function is solved to obtain the final match. The cross section of the aviation blade is compared with the design model through the final matching result and rigid body transformation to determine whether the cross section profile of the aviation blade is qualified.

Furthermore, in step (2), the ICP is solved using the singular value decomposition (SVD) method.

Furthermore, the specific implementation method of defining the tolerance range is as follows:

According to the design tolerance, the upper limit U and lower limit _L of the tolerance band are set for the leading edge, trailing edge, blade basin and blade back area. = According to the tolerance band range of each area, the theoretical profile of each area is offset along its normal by the distance of the upper deviation value and the lower deviation value (positive outward), and the strip area between the two offset curves is the tolerance band range. If the measurement point error in each area is between U and L, the measurement point is said to be within the tolerance band range, otherwise it is outside the tolerance band range.

Further, step (4) comprises:

(41) As shown in Figure 2, let the theoretical data point of the blade be p, the normal of the design model pointing inward at each data point p be n, the nearest point searched for each measuring point in the theoretical point set be q, and construct a vector t from point q to point p. If , it means that the measuring point is inside the theoretical blade, and L will be used to determine whether this point is within the tolerance range. Otherwise, it means that the measuring point is outside the theoretical blade, and U will be used to determine whether this point is within the tolerance range.

(42) Let d be the point-to-point distance between each measuring point in each region and the nearest point found in the designed blade profile. If the point is outside the theoretical blade profile, the directed distance s=d>0; if the point is inside the theoretical blade profile, the directed distance s=-d≤0.

(43) Construct the measurement point coefficient w:

Wherein, _si represents the point-to-point distance between the measured point _qi and its corresponding nearest point. U represents the upper limit of the tolerance band of the subset to which _qi belongs, and L represents the lower limit of the tolerance band of the subset to which _qi belongs. Furthermore, the construction method of the new objective function in step (5) includes:

(51) Add the measurement point coefficient to each measurement point and obtain the objective function:

Among them, M is the total number of measurement points, R and T are the rotation matrix around the Z axis and the translation matrix in the XOY plane in the rigid body transformation, respectively.

(52) Set the coordinates of q _i to (q _x , q _y ), the coordinates of _pi to (p _x , p _y ), and the matching parameters x = (x′, y′, θ), where x′ is the x offset, y′ is the y offset, and θ is the torsion angle, as shown in Figure 3. Expanding the objective function yields the objective function:

(53) Let the lower and upper deviations of x′ be x ₁ , x ₂ ; let the lower and upper deviations of y′ be y ₁ , y ₂ ; let the lower and upper deviations of θ be θ ₁ , θ ₂ . Adding this constraint to the above objective function, we get the final objective function:

Furthermore, the specific implementation method of solving the problem in step (6) is:

(61) The ICP initial solution obtained in step 1 is used as the initial parameter to construct a quadratic model for the objective function:

(62) Set the number of iterations K and the precision ε, and set k = 0;

(63) When the number of iterations reaches k = K or the norm of the projected gradient ||P(x _k -g _k ,l,u)-x _k || _∞ is less than the precision ε, the iteration ends; where l = (x ₁ ,y ₁ ,γ ₁ ),u = (x ₂ ,y ₂ ,γ ₂ );

(64) The Cauchy point method is used to obtain the approximate solution x _c of the Cauchy point for the quadratic model;

(65) First, ignore the constraints and start from x _c to perform the subspace minimization operation of the quadratic model of the objective function. Then, satisfy the constraints by backtracking the points into the feasible region and obtain the search direction d _k ;

(66) A line search loop is performed with the new search direction d _k until |g _k+1 ^T d _k |≤0.9|g _k ^T d _k | is satisfied. In each loop, x is updated to x _k+1 =x _k + λ _k d _k , where λ is initially 1 and decreases linearly.

(67) Update the Hessian matrix and jump to step (63).

The tolerance requirements of various regions of the blade in this embodiment are shown in Table 1.

Table 1

Leading edge area tolerance Trailing edge area tolerance Leaf basin area tolerance Tolerance of back area X offset Y offset Torsion Angle 0±0.15mm 0±0.15mm 0±0.075mm 0±0.075mm 0±0.15mm 0±0.15mm 0±20°

The same blade section is matched by ICP matching and matching according to the present invention, respectively. The evaluation results of existing ICP matching are shown in Table 2, and the evaluation results of matching according to the present invention are shown in Table 3.

Table 2

Leading edge maximum error Maximum trailing edge error Maximum error of leaf basin Maximum error of leaf back X offset Y offset Torsion Angle -0.138mm 0.101mm 0.080mm -0.076mm -0.160mm 0.101mm -12.765°

Table 3

Leading edge maximum error Maximum trailing edge error Maximum error of leaf basin Maximum error of leaf back X offset Y offset Torsion Angle -0.142mm 0.081mm 0.072mm 0.066mm -0.149mm 0.106mm -13.000°

From the above evaluation results, it can be seen that the maximum error of the blade basin, the maximum error of the blade back and the X offset obtained by comparing the blade profile obtained by the ICP matching method with the designed blade profile all exceed the tolerance requirements, so the evaluation result is that the blade is unqualified. However, the method described in the present invention is used to match and evaluate the blade profile section and the designed blade profile, and the errors of the blade profile section all meet the tolerance requirements, and the evaluation result is qualified.

It will be easily understood by those skilled in the art that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims (6)

1. The method for evaluating the profile parameters of the cross section profile of the aviation blade is characterized by comprising the following steps:

s1, dispersing a design model of the cross section of an aviation blade to be evaluated into a theoretical data point set P, and measuring measurement point data of the corresponding cross section of the aviation blade to be evaluated to form a blade profile contour point set Q;

S2, performing ICP matching by using P and Q to obtain an initial matching result;

s3, segmenting the profile of the blade profile, and dividing Q into a front edge sub-point set, a rear edge sub-point set, a She Penzi point set and a blade back sub-point set;

S4, for each measuring point in each sub-point set, searching a corresponding nearest point in P, calculating the point distance between the measuring point and the corresponding nearest point, setting a tolerance zone range for each sub-point set according to a profile evaluation standard, and giving a distance coefficient of each measuring point according to the position relation between the point distance and the corresponding tolerance zone range;

s5, taking the x offset, the y offset and the torsion angle of Q as decision variables, and taking the distance coefficient of each measuring point as a weight to construct an objective function with position constraint;

wherein the objective function with position constraint is as follows:

wherein x ', y', θ respectively represent the x offset, y offset and torsion angle of Q, w _i represents the coefficient of the ith measuring point Q _i in Q, M represents the number of measuring points of Q, Respectively representing the x coordinate and the y coordinate of the measuring point q _i,Respectively representing the x coordinate and the y coordinate of a measuring point P _i closest to a measuring point q _i in P, respectively representing the upper and lower tolerance of the offset by x ₂,x₁, respectively representing the upper and lower tolerance of the y offset by y ₂,y₁, respectively representing the upper and lower tolerance of the torsion angle by θ ₂,θ₁, wherein the upper and lower tolerances are set according to the profile offset tolerance evaluation criterion;

S61, constructing a secondary model for the objective function by taking an initial matching result x ₀ as an initial parameter:

S62, setting iteration times K and precision epsilon, and initializing the iteration times K=0;

s63, ending iteration when the iteration times k=K or the norm ||P of the projection gradient is reached (x _k-g_k,l,u)-x_k||_∞ is smaller than the precision epsilon, wherein l= (x ₁,y₁,γ₁),u＝(x₂,y₂,γ₂);

S64, obtaining an approximate solution x _c of Ke Xidian by using a CauthyPoint method on the secondary model;

S65, firstly ignoring constraint, starting from x _c, performing subspace minimization operation of the objective function quadratic model, and then backtracking points into a feasible region to meet constraint conditions and obtain a search direction d _k;

S66, performing line search circulation through the new search direction d _k until the requirement is met Updating x _k+1＝x_k+λ_kd_k for x in each cycle, wherein λ is initially 1 and decreases linearly;

s67, updating the Hessian matrix, and jumping to the step S63;

s7, performing rigid transformation on the cross section of the aviation blade by utilizing an optimal matching result, comparing the cross section of the aviation blade after rigid transformation with a design model, and judging whether the profile of the cross section molded line of the aviation blade is qualified or not.

2. The method of claim 1, wherein step S3 comprises:

S31, performing cubic spline curve interpolation on the measuring points in the Q to obtain a C ² continuous curve L (x), and dispersing the curve according to an equal arc length principle to obtain a discrete rear point set Q';

s32, calculating the curvature of each measuring point in the Q Wherein L _i′,L_i' respectively represents a first derivative and a second derivative of the ith measuring point at the curve L (x), and the leaf basin point set, the leaf back sub point set, the front edge sub point set and the rear edge sub point set are extracted from Q according to the characteristic that the curvature of the front edge area and the rear edge area is far greater than the curvature of the leaf back area of the leaf basin.

3. The method of claim 1, wherein step S4 comprises:

s41, searching the nearest point in P for each measuring point in the Q neutron point set;

S42, calculating the point distance between the measuring point in the Q and the corresponding nearest point in the P;

S43, if the point is positioned outside the theoretical leaf pattern, a directional distance s= ||q _i-p_i | between the measuring point Q _i in Q and a corresponding nearest point P _i in P, and if the point is positioned inside the theoretical leaf pattern, a directional distance s= - ||q _i-p_i |.

S44, setting a tolerance zone range for each sub-point set according to the profile evaluation standard, and if the ith measuring point falls within the corresponding tolerance zone range, setting a distance coefficient w _i of the measuring point to be 1; if the ith measuring point falls outside the corresponding tolerance zone, the distance coefficient w _i of the point is larger than 1, which is determined by the directional distance and the tolerance boundary.

4. The method of claim 1, wherein the station distance coefficient is calculated as follows:

Wherein w _i represents the coefficient of the ith measuring point Q _i in Q, s _i represents the directed distance of the ith measuring point Q _i in Q, U represents the upper tolerance zone limit of the subset to which Q _i belongs, and L represents the lower tolerance zone limit of the subset to which Q _i belongs.

5. A method according to any one of claims 1 to 4, wherein the evaluation is acceptable if the individual errors of the profile cross-section meet the tolerance requirements.

6. An aircraft blade cross-section profile parameter evaluation system, comprising: a computer readable storage medium and a processor;

The computer-readable storage medium is for storing executable instructions;

The processor is configured to read executable instructions stored in the computer readable storage medium and execute the aerovane cross-section profile parameter evaluation method of any one of claims 1 to 5.