An assembly key point-based complex product deviation transmission calculation method

Abstract

The application provides a complex product deviation transmission calculation method based on assembly key points, first obtains assembly relationship and key feature information of a complex product component, constructs an assembly information coordination tree and an assembly deviation transmission network, extracts assembly level information and deviation node relationship, calculates deviation transmission entropy, obtains an assembly deviation transmission function and determines the assembly deviation transmission network, substitutes data calculation and outputs a group of deviation values of key control points, so that a group of deviation values of key control points are obtained. Reasonable adjustment is carried out in combination with technical experience accumulation to verify that the output deviation values meet the assembly parameters. The application comprehensively considers factors influencing assembly effect, introduces an intelligent optimization algorithm in combination with the assembly deviation transmission network, and finally achieves the effect of improving assembly performance.

Description

Technical Field

[0001] This invention patent relates to the field of aircraft manufacturing and assembly technology, and further to a method for calculating the transmission of deviations in complex products based on assembly key points. Background Technology

[0002] With the continuous improvement of the precision and performance of complex products, the requirements for safety and economy are receiving increasing attention from manufacturing enterprises. Currently, traditional deviation analysis methods rely on the experience of process engineers to determine approximate deviation values, which are then verified through simulation. This approach struggles to accurately construct assembly deviation analysis models, resulting in deviation values ​​that fail to account for the large uncertainty, nonlinearity, and multi-level characteristics of assembly, and making it impossible to accurately calculate deviation values ​​at key points. Therefore, this invention proposes an assembly deviation propagation method based on a transfer entropy network graph. This method analyzes the sources of assembly deviations, combines assembly deviation detection data to construct a multi-level assembly coordination tree, and integrates transfer entropy and complex network graph theory to construct an assembly deviation propagation model, revealing the topological relationships of assembly deviations and measuring causality indicators. Summary of the Invention

[0003] The purpose of this application is to provide a method for calculating the transmission of deviations in complex products based on assembly key points, extracting key deviation sources in the product, and calculating the initial deviation value of assembly deviation transmission according to the transmission entropy algorithm and the mean shift algorithm, for use in aircraft assembly quality control.

[0004] To achieve the above objectives, this application provides a method for calculating the propagation of deviations in complex products based on assembly key points. The method includes:

[0005] Step 1: Obtain the assembly relationship and key feature data of complex product components, namely, obtain its assembly process flow, the main transmission path of assembly deviation, positioning reference and control target of key feature points;

[0006] Step 2: Construct an assembly information coordination tree to extract assembly hierarchy information and deviation node relationships;

[0007] An assembly information coordination tree is constructed, and a hierarchical transfer entropy network diagram for the accumulation and decomposition of assembly deviations is constructed by selecting typical components. As the reverse process of structural decomposition, the assembly process is progressive and based on the specific relationships and tolerance constraints of the assembly objects. Parts are assembled into components, then parts and components are assembled into parts, then parts are assembled into large parts, and finally large parts are assembled into the entire complex product.

[0008] Step 3: Calculate the assembly deviation propagation entropy and obtain the assembly deviation transfer function;

[0009] For two assembly unit sequences I and J with stationary Markov properties at assembly levels, where the J-th level assembly unit sequence J = {x1, x2, ..., x...}n ,x n+1 The assembly unit sequence of level I is I = {y1, y2, ... y}. n There is a certain coupling relationship between the two assembly unit sequences. The transmission relationship is used to characterize the amount of assembly deviation transferred from J to I:

[0010]

[0011] Among them, T J→I Let x represent the entropy relationship of the transfer information from the J-th level assembly unit sequence to the I-th level assembly unit sequence. n+1 ,x n For the J-th level assembly unit, y n Let p(·|·) be the assembly unit at level I; here p(·|·) represents the conditional entropy, i.e., p(x) n+1 |x n ) indicates that in the known J-level assembly unit x n Under the conditions, assembly unit x n+1 Uncertainty; p(x) n+1 |x n ,y n ) represents a known assembly unit y n and x n Under the conditions, assembly unit x n+1 Uncertainty;

[0012] Step 4: Construct an assembly deviation transmission network to reveal the assembly deviation transmission mechanism;

[0013] For assembly unit sequences with nonlinear characteristics, calculating the deviation propagation relationship requires effective probability density estimation of the detection data. Let x1, x2, ..., x... i , ..., x n For n assembly units in an assembly unit sequence X, the kernel density estimation is... for:

[0014]

[0015] In the formula, h represents the kernel function bandwidth; a larger value indicates a smoother kernel density function curve, less variation, and more accurate results; N is the length of the assembly unit sequence for complex products; K(·) represents the unit step kernel function.

[0016]

[0017] The probability density M of the sample is calculated using the mean shift method. h (x):

[0018]

[0019] Where w(x) i )≥0 is the weight of each assembly unit;

[0020] The integrated assembly deviation transmission relationship is expressed as follows:

[0021]

[0022] Step 5: Combining the key feature data extracted in Step 1 and the assembly hierarchy information in Step 2, calculate a set of deviation values ​​for key control points based on the assembly deviation transmission relationship expression.

[0023] This invention proposes a method for calculating the propagation of deviations in complex products based on assembly key points. It primarily addresses the problem of difficulty in effectively estimating the probability density of detection data when calculating the propagation relationship of assembly unit sequences with nonlinear characteristics. The method utilizes a mean-shift algorithm to solve this problem. The characteristics of the mean-shift algorithm are: (1) it is a parameter-free mean estimation algorithm that finds the peak of the distribution along the ascending direction of the probability gradient; (2) because it defines a kernel function, it is equivalent to adding a weight coefficient to each sample. This weight coefficient makes the weights of different samples different, which causes the "contribution" of the offset value to the offset vector to vary with the distance between the sample and the offset point. Attached Figure Description

[0024] Figure 1 This is a flowchart illustrating the calculation of complex product deviation transmission at key assembly points in this invention.

[0025] Figure 2 For the overall assembly process of complex products.

[0026] Figure 3 A hierarchical entropy transfer network diagram for assembly deviations. Implementation

[0027] The purpose of this invention is to provide a method for calculating the transmission of deviations in complex products based on key assembly points. This method extracts the key deviation sources in the product and calculates the initial deviation value of the assembly deviation transmission based on the transmission entropy algorithm and the mean shift algorithm, which can be used for aircraft assembly quality control.

[0028] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0029] Figure 1 This is a flowchart illustrating the calculation of complex product deviations at key assembly points in this invention, as shown below. Figure 1 As shown, this invention provides a method for calculating the transmission of aircraft nose deviation based on assembly key points. The transmission quantification analysis method includes:

[0030] Step S1: Obtain the assembly relationship and key feature data of the aircraft nose;

[0031] Step S2: Construct an assembly information coordination tree and extract assembly hierarchy information and deviation node relationships;

[0032] Step S3: Calculate the assembly deviation propagation entropy and obtain the assembly deviation transfer function;

[0033] Step S4: Construct an assembly deviation propagation network to reveal the assembly deviation propagation mechanism.

[0034] Step S5: Substitute the data to calculate and output a set of deviation values ​​for key control points;

[0035] The following is a detailed discussion of each step:

[0036] Step S1: Obtain the assembly relationships and key feature data of the aircraft nose section. Specifically, this refers to obtaining the assembly process flow, main transmission routes of assembly deviations, positioning references, and control targets for key feature points based on design data. The aircraft nose section follows the assembly process of "parts → components → sub-components → large components → nose section". Figure 2 As shown, the assembly of aircraft components, parts, and docking of large components are achieved through positioning, clamping, connecting, releasing, and springback mechanisms.

[0037] Step S2: Construct an assembly information coordination tree, extract assembly hierarchy information and deviation node relationships, specifically including:

[0038] Construct an assembly information coordination tree, and select typical components to construct a hierarchical transfer entropy network diagram for the accumulation and decomposition of assembly deviations, as the inverse process of structural decomposition. Figure 2 The overall assembly process for the aircraft nose section is a progressive one, based on the specific relationships and tolerance constraints of the assembled objects. Parts are assembled into components, then parts and components are assembled into sub-components, then sub-components are assembled into larger sub-components, and finally the large sub-components are assembled into the entire aircraft nose section. Figure 2 As shown, its mathematical expression is:

[0039]

[0040] Where: P i Components include: cockpit frame, cockpit floor, frame assembly, escape hatch, boarding door panel, top panel, forward service bay floor beam, forward landing gear bay, lower forward side panel, lower left and right side panels, lower rear left, center and right side panels, electrical and electronic bay door, and forward service bay floor panel; C i The components include: cockpit components, floor and frame components; S i The components include: cockpit engagement assembly, service compartment assembly, and lower forward assembly; L iFor large components, including: upper part of the head and lower part of the head; H represents the head; ∑ represents the geometric relationship and tolerance constraints between them;

[0041] Step S3: Calculate the assembly deviation propagation entropy and obtain the assembly deviation transfer function;

[0042] For two assembly unit sequences I and J with stationary Markov properties at assembly levels, where the J-th level assembly unit sequence J = {x1, x2, ..., x...} n ,x n+1 The assembly unit sequence of level I is I = {y1, y2, ... y}. n There is a certain coupling relationship between the two assembly unit sequences. The transmission relationship is used to characterize the amount of assembly deviation transferred from J to I:

[0043]

[0044] Among them, T J→I Let x represent the entropy relationship of the transfer information from the J-th level assembly unit sequence to the I-th level assembly unit sequence. n+1 x n For the J-th level assembly unit, y n For the I-th level assembly unit; p(·) is the probability density;

[0045] Figure 3 The following is a diagram of the hierarchical entropy transfer network for assembly deviations in an embodiment, such as... Figure 3 As shown, the directed edge representation of the information transmission relationship between assembly unit deviations is given, forming a hierarchical transfer entropy network diagram that can express the accumulation and decomposition of assembly deviations, T. J→I It can describe the cumulative bias, T I→J It can describe the deviation decomposition.

[0046] For two assembly unit sequences J and I with linear Gaussian distribution characteristics, the transfer entropy relationship can be transformed into the following expression using the covariance matrices of I and J:

[0047]

[0048] Where C represents the covariance matrix;

[0049] Step S4: Construct an assembly deviation propagation network to reveal the assembly deviation propagation mechanism.

[0050] Step S41: For assembly unit sequences with nonlinear characteristics, calculating the deviation propagation relationship requires effective probability density estimation of the detection data. For assembly unit sequence X... t For each cell, the kernel density estimation function is:

[0051]

[0052] In the formula, h represents the kernel function bandwidth; a larger value indicates a smoother kernel density function curve, less variation, and more accurate results; N is the length of the aircraft nose assembly unit sequence; K(·) represents the unit step kernel function.

[0053]

[0054] Step S42: Select the mean shift algorithm to calculate the probability density of the sample. The mean shift formula is as follows:

[0055]

[0056] Where w(x) i )≥0 is the weight of each assembly unit;

[0057] The assembly deviation propagation relationship expression is integrated using the mean shift algorithm as follows:

[0058]

[0059] Step S5: Substitute the data to calculate and output a set of deviation propagation entropy for key control points;

[0060] Combining the key feature data extracted in step S1 and the assembly hierarchy information in step S2, a set of deviation values ​​for key control points are calculated based on the assembly deviation transmission relationship expression.

Claims

1. A method for calculating the propagation of deviations in complex products based on assembly key points, the method comprising: Step 1: Obtain the assembly relationship and key feature data of complex product components, namely, obtain its assembly process flow, the main transmission path of assembly deviation, positioning reference and control target of key feature points; Step 2: Construct an assembly information coordination tree to extract assembly hierarchy information and deviation node relationships; Construct an assembly information coordination tree and select typical components to construct a hierarchical transfer entropy network diagram for the accumulation and decomposition of assembly deviations. As the reverse process of structural decomposition, the assembly process is progressive and based on the specific relationships and tolerance constraints of the assembly objects. Parts are assembled into components, then parts and components are assembled into parts, then parts are assembled into large parts, and finally large parts are assembled into the entire complex product. Step 3: Calculate the assembly deviation propagation entropy and obtain the assembly deviation transfer function; For two assembly unit sequences I and J with stationary Markov properties at assembly levels, where the J-th level assembly unit sequence J = {x1, x2, ..., x...} n ,x n+1 The assembly unit sequence of level I is I = {y1, y2, ... y}. n There is a certain coupling relationship between the two assembly unit sequences. The transmission relationship is used to characterize the amount of assembly deviation transferred from J to I: Among them, T J→I Let x represent the entropy relationship of the transfer information from the J-th level assembly unit sequence to the I-th level assembly unit sequence. n+1 ,x n For the J-th level assembly unit, y n Let p(·|·) be the assembly unit at level I; here p(·|·) represents the conditional entropy, i.e., p(x) n+1 |x n ) indicates that in the known J-level assembly unit x n Under the conditions, assembly unit x n+1 Uncertainty; p(x) n+1 |x n ,y n ) represents a known assembly unit y n and x n Under the conditions, assembly unit x n+1 Uncertainty; Step 4: Construct an assembly deviation transmission network to reveal the assembly deviation transmission mechanism; For assembly unit sequences with nonlinear characteristics, calculating the deviation propagation relationship requires effective probability density estimation of the detection data. Let x1, x2, ..., x... i , ..., x n For n assembly units in an assembly unit sequence X, the kernel density estimation is... for: In the formula, h represents the kernel function bandwidth; a larger value indicates a smoother kernel density function curve, less variation, and more accurate results; N is the length of the assembly unit sequence for complex products; K(·) represents the unit step kernel function. The probability density M of the sample is calculated using the mean shift method. h (x): Where w(x) i )≥0 is the weight of each assembly unit; The integrated assembly deviation transmission relationship is expressed as follows: Step 5: Combining the key feature data extracted in Step 1 and the assembly hierarchy information in Step 2, calculate a set of deviation values ​​for key control points based on the assembly deviation transmission relationship expression.

Images