Spacecraft cabin part process route updating method considering design change

Abstract

The invention discloses a spacecraft cabin part process route updating method considering design change, and the method comprises the steps: comprehensively considering the part attribute similarity, feature structure similarity and feature change similarity, and building a comprehensive similarity calculation model; and when the similarity reaches a set threshold value, process route updating is carried out based on the typical parts. And secondly, dynamically adjusting the step priority graph and the 0-1 step matrix according to the feature difference (addition, deletion and transformation) of the target part and the typical part, and solving the optimal process route by adopting the existing optimization algorithm. And finally, through a fuzzy comprehensive evaluation method, evaluating the rationality of the changed process route from five dimensions of process cost, processing time, processing difficulty, manufacturing resource requirements and process feasibility. Therefore, the problems that the updating efficiency of an existing process route of similar parts of a spacecraft is low, and the multi-dimensional change requirement of aerospace complex products cannot be met are effectively solved.

Description

Technical Field

[0001] The present invention relates to a process route updating method, in particular to a process route updating method for spacecraft cabin parts taking design changes into consideration. Background Art

[0002] Design changes are an inevitable and crucial part of machining process planning. Design changes typically refer to modifications to the original design during the product design or manufacturing process due to factors such as design optimization, functional requirements adjustments, manufacturing constraints, or changes in the external environment. Design changes not only involve adjustments to product geometry, material properties, or functional requirements but can also impact machining processes, production processes, and resource allocation.

[0003] According to statistics, only 20% of new spacecraft development requires complete redesign; the remainder can either directly leverage existing products or undergo minor modifications. Especially for high-variety, low-volume aerospace products, component designs often require frequent iteration and optimization to meet stringent performance requirements and process standards. Furthermore, evolving customer demands and supply chain fluctuations often necessitate design revisions. These changes not only increase the complexity of process planning but also impact product cycle time, manufacturing costs, and process quality.

[0004] Currently, design changes have been applied in many fields of the manufacturing industry, but the application of this technology in the aerospace field is still insufficient, which makes it impossible to effectively improve the efficiency of aerospace product process design; secondly, traditional design methods mainly rely on manual operation and experience judgment, which makes it difficult to cope with the multi-dimensional change requirements of complex aerospace products. Summary of the Invention

[0005] In order to address the deficiencies of the above-mentioned technologies, the present invention provides a method for updating the process route of spacecraft cabin parts taking into account design changes.

[0006] In order to solve the above technical problems, the technical solution adopted by the present invention is: a method for updating the process route of spacecraft cabin parts considering design changes, comprising the following steps:

[0007] Step 1: Build a part information model and design a multi-dimensional comprehensive similarity calculation formula between two parts based on the part information model. The multi-dimensional calculation formula integrates part attribute similarity, part feature structure similarity, and feature change similarity;

[0008] Step 2: When the similarity between two parts reaches a set threshold, a feature-oriented part process route update plan is designed based on the repetitive process of typical parts;

[0009] The process of updating the part process route includes: dynamically adjusting the work step priority diagram and 0-1 work step matrix based on the feature differences between the target part and the typical part, which include feature addition, deletion, and transformation, and using the existing optimization algorithm to solve the optimal process route.

[0010] Step 3: Based on the fuzzy comprehensive evaluation method, an evaluation mechanism for the updated part process route is constructed;

[0011] Step 4: Evaluate the process route update plan designed in step 2 based on the evaluation mechanism. If the evaluation requirements are met, the process route update plan is determined to be feasible.

[0012] Preferably, in step 1, the constructed part information model includes three aspects and a total of 9 indicators, namely:

[0013] Part attribute information: part type, part material;

[0014] Part feature structure information: feature location, feature main type, feature subtype, and feature number;

[0015] Characteristic parameters: dimensional parameters, accuracy grade, surface roughness.

[0016] Preferably, in step 1, the comprehensive similarity between the target part and the typical part is calculated by integrating the part attribute similarity, the part feature structure similarity, and the feature change similarity, as shown in formula (15):

[0017] S=α Ι S Ι +α ΙΙ S ΙΙ +α ΙΙΙ S ΙΙΙ (15)

[0018] In the formula, S represents the comprehensive similarity of the target part to the typical part, α Ι , α ΙΙ , α ΙΙΙ Respectively represent the influence weights of part attributes, part feature structures and feature parameters on the comprehensive similarity, and satisfy α Ι +α ΙΙ +α ΙΙΙ =1.

[0019] Preferably, in step 1, the part attribute similarity is the part type similarity Similarity to part material The product of the two terms is shown in formula (3).

[0020]

[0021] Where S ΙIndicates the similarity of part attributes; similarity of part types As shown in formula (1), the part material similarity It is expressed as shown in formula (2);

[0022]

[0023] Where, and Represents the tensile strength limit of the target part and the typical part respectively; HBS obj and HBS typ Represents the Brinell hardness of the target part and the typical part respectively; α σ and α H Respectively represent the influence weights of strength and hardness on the similarity of part materials, and satisfy α σ +α H =1.

[0024] Preferably, in step 1, the calculation of the similarity of the part feature structure is performed in the following three steps:

[0025] 1) Define the feature judgment factor and judge whether two features have high similarity based on the feature main type, subtype and position requirements, as shown in formula (4);

[0026] 2) In the target part and the typical part, the number of parts with the same manufacturing features is counted based on the feature judgment factor, as shown in formula (5);

[0027] 3) Calculate the feature structure similarity according to formula (6);

[0028]

[0029]

[0030] Where, Represents the feature judgment factor, F x ,F y Respectively represent two manufacturing features that are compared with each other; represents the kth manufacturing feature under the jth subtype of the ith main type in the target part and the typical part respectively; N M Indicates the number of feature main types, N S (i) represents the number of characteristic subtypes under the i-th main type, N F (ij) represents the number of features under the jth subtype of the i-th main type, N FS Represents the total number of identical features between the target part and the typical part; N obj ,N tpy Represents the total number of features of target parts and typical parts respectively; S ΙΙ Indicates the similarity of part feature structure.

[0031] Preferably, in step 1, the calculation process of feature change similarity is as follows:

[0032] Feature changes are divided into feature addition, feature deletion, and feature transformation;

[0033] Among them, the similarity calculation of the newly added features is shown in formula (7):

[0034]

[0035] Where, Indicates the similarity of a single newly added feature of the target part relative to the typical part;

[0036] The similarity calculation of feature deletion is shown in formula (8):

[0037]

[0038] Where, Indicates the similarity of the target feature to a single deleted feature of a typical part;

[0039] The similarity calculation of feature transformation is shown in formula (9):

[0040]

[0041] Where, represents the similarity of a single transformed feature of the target part relative to the typical part; φ represents the transformation similarity operator, which is used to calculate the similarity of two features with the same main type and subtype; F obj ,F typ Represent the manufacturing features of target parts and typical parts respectively;

[0042] φ(F obj ,F typ ) It is necessary to calculate the similarity of the size parameters, accuracy level and surface roughness, and then calculate the feature transformation similarity according to their respective weights, as shown in formulas (10)-(13):

[0043]

[0044] φ(F obj ,F typ )=α D S SIZ +α AC S ACC +α R S ROU (13)

[0045] Where S SIZ 、S ACC 、SROU They represent the similarity of dimensional parameters, accuracy level and surface roughness respectively; Represent the qth size of the target part and the typical part under the pth transformation feature respectively; Represent the qth accuracy level of the target part and the typical part under the pth transformation feature, They represent the qth precision surface roughness of the target part and the typical part under the pth transformation feature respectively; M(p), N(p), and L(p) represent the number of sizes, the number of precision grades, and the number of surface roughness under the pth transformation feature respectively; α D , α AC , α R Respectively represent the influence weights of size parameters, accuracy level and surface roughness on the similarity of feature parameters, and satisfy α D +α AC +α R =1;

[0046] Since addition, deletion, and transformation have the same impact on feature change similarity, the feature change similarity is calculated as shown in formula (14):

[0047]

[0048] Where S ΙΙΙ Indicates the similarity of feature changes, N add 、N del 、N alt They respectively represent the number of new features, the number of deleted features, and the number of transformed features of the target part relative to the typical part.

[0049] Preferably, in step 2, the process of updating the part process route specifically includes the following steps:

[0050] 1) Determine the processing plan for each changed feature;

[0051] 2) Determine the relative position of the work step corresponding to the changed feature in the work step priority graph, and form an updated work step priority graph of the target part based on the addition, deletion, and transformation operations of the feature differences; a 0-1 work step matrix can be obtained based on the work step priority graph of the target part;

[0052] (3) Taking the 0-1 process step matrix as input, the reoptimization genetic algorithm is used to solve the optimal process route.

[0053] Preferably, in step 2, a similarity formula between the process parameters of the changed feature and the available process parameters of the processing scheme is established to determine the optimal processing scheme for the changed feature;

[0054] Among them, the similarity formula for changing characteristic process parameters is shown in formula (16):

[0055]

[0056] Where X represents the vector composed of the changed characteristic size parameters, accuracy level and surface roughness; Y represents the optimal parameter vector of the processing solution; i Indicates the value of the i-th parameter in X, y i represents the value of the i-th parameter in Y; n represents the number of times the processing plan is selected;

[0057] The similarity formula of the process parameters of the processing plan can be used, as shown in formula (17):

[0058]

[0059] Where β represents the utilization coefficient of the i-th processing scheme; α represents the influence weight of the i-th process parameter.

[0060] Preferably, in step 3, the comprehensive evaluation model is as shown in formula (28):

[0061] B=A·R (28)

[0062] In the formula, B represents the evaluation result vector, and A represents the weight set;

[0063] According to the maximum membership principle, the process route evaluation result is obtained, and then the feasibility of the part process route update is judged based on the evaluation result.

[0064] Preferably, in step 3, during the construction of the fuzzy comprehensive evaluation method:

[0065] Establish the comprehensive evaluation factor set U of the process route, as shown in formula (21):

[0066] U={u1,u2,u3,u4,u5} (21)

[0067] Where U represents the factor set of fuzzy comprehensive evaluation of process route, u1, u2, u3, u4, and u5 represent the five factors affecting the process route, namely process cost, processing time, processing difficulty, degree of requirement of process engineer, and feasibility of process route;

[0068] 2) Taking the qualitative value as the evaluation alternative set, the fuzzy evaluation set V of the process route is established, as shown in formula (22):

[0069] V=(v1,v2,v3,v4,v5) (22)

[0070] Where, represents the fuzzy evaluation set of the process route, v1, v2, v3, v4, and v5 represent excellent, good, medium, poor, and bad, respectively;

[0071] The weight coefficients of the factors determined by combining literature and expert evaluation are required to meet specific conditions for the weight coefficients affecting the process route, as shown in formula (23):

[0072]

[0073] Where λ i represents the weight coefficient of the i-th influencing factor, i = 1, 2, 3, 4, 5;

[0074] Combine the literature and expert opinions to determine each weight coefficient λ i , forming a weight set A, the specific values are shown in formula (24):

[0075] A=(0.1,0.1,0.3,0.2,0.3) (24)

[0076] The membership degree of the evaluation set V is determined by the judgment matrix; the single factor evaluation set is shown in formula (25):

[0077] r i =(r i1 ,r i2 ,r i3 ,r i4 ,r i5 ) (25)

[0078] Where r i represents the single factor evaluation set of the i-th factor in the factor set, r i1 ,r i2 ,r i3 ,r i4 ,r i5 They respectively represent the membership of the i-th factor to the evaluation set V.

[0079] Perform single factor evaluation on the factor set and generate a fuzzy judgment matrix, as shown in formula (26):

[0080]

[0081] In the formula, R represents the fuzzy judgment matrix, r ij It represents the membership value of the i-th factor in the factor set to the j-th element in the evaluation set;

[0082] After normalization, r ij Satisfaction relationship:

[0083]

[0084] Currently, spacecraft cabin parts are characterized by numerous features, diverse shapes, and complex processes, and similar parts often exhibit a high degree of similarity. For these parts, this paper proposes a process route update method for spacecraft cabin parts that takes design changes into account, based on similarity calculation. First, a comprehensive similarity calculation model is established by comprehensively considering part attribute similarity, feature structure similarity, and feature change similarity. When the similarity reaches a set threshold, the process route is updated based on a typical part. Second, based on the feature differences (additions, deletions, and transformations) between the target part and the typical part, the process step priority diagram and 0-1 process step matrix are dynamically adjusted, and the optimal process route is determined using existing optimization algorithms. Finally, a fuzzy comprehensive evaluation method is used to evaluate the rationality of the modified process route based on five dimensions: process cost, processing time, processing difficulty, manufacturing resource requirements, and process feasibility. This method effectively addresses the low efficiency of existing process route updates for similar spacecraft parts and their inability to address the multi-dimensional change requirements of complex aerospace products. BRIEF DESCRIPTION OF THE DRAWINGS

[0085] Figure 1 This is the parts information model diagram constructed by the present invention.

[0086] Figure 2 This figure illustrates the changes to the features of the present invention.

[0087] Figure 3 This is a comprehensive evaluation factor diagram of the process route quality after the design change of the present invention.

[0088] Figure 4 3D model diagram of typical parts and target part A in an embodiment of the present invention.

[0089] Figure 5 This is a design change process diagram of an embodiment of the present invention.

[0090] Figure 6 It is the 0-1 step matrix of the embodiment of the present invention.

[0091] Figure 7 This is the iteration diagram of target part A in the embodiment of the present invention. DETAILED DESCRIPTION

[0092] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0093] The present invention discloses a method for updating the process route of spacecraft cabin parts taking into account design changes. It effectively uses digital technology to apply the change concept to the process route change of spacecraft products, and proposes a multi-dimensional comprehensive similarity calculation method that integrates part attribute similarity, feature structure similarity and feature change similarity, providing a quantitative evaluation basis for the feature changes of similar parts of the spacecraft cabin; and a dynamic process route updating technology that takes into account design changes, establishes a dynamic adjustment mechanism of the work step priority diagram and the 0-1 work step matrix for the feature differences (addition / deletion / transformation) of similar spacecraft parts, combines the re-optimization genetic algorithm with the fuzzy comprehensive evaluation, and realizes the rapid update and feasibility verification of the process route.

[0094] The overall technical route for updating the process route of spacecraft cabin parts considering design changes proposed in the present invention includes the following steps:

[0095] Step 1: Build a part information model and design a multi-dimensional comprehensive similarity calculation formula between two parts based on the part information model. The multi-dimensional calculation formula integrates part attribute similarity, feature structure similarity, and feature change similarity;

[0096] Step 2: When the similarity between two parts reaches a set threshold, a feature-oriented part process route update plan is designed based on the repetitive process of typical parts;

[0097] The process of updating the part process route includes: dynamically adjusting the work step priority diagram and 0-1 work step matrix based on the feature differences between the target part and the typical part, which include feature addition, deletion, and transformation, and using the existing optimization algorithm to solve the optimal process route.

[0098] Step 3: Based on the fuzzy comprehensive evaluation method, an evaluation mechanism for the updated part process route is constructed;

[0099] Step 4: Evaluate the process route update plan designed in step 2 based on the evaluation mechanism. If the evaluation requirements are met, the process route update plan is determined to be feasible.

[0100] Evaluate the rationality of the changed process route from five dimensions: process cost, processing time, processing difficulty, manufacturing resource requirements, and process feasibility

[0101] 1. Design of multi-dimensional comprehensive similarity calculation formula:

[0102] In product structure design, different parts with high similarity usually do not need to be redesigned, and design changes can be made based on existing typical parts. The same is true for process planning of similar parts. In order to comprehensively and effectively evaluate the similarity between two parts, the present invention considers three major aspects that affect process planning, totaling 9 indicators, and establishes a part information model, which includes: part attribute information (part type, part material), part feature structure information (feature position, feature main type, feature subtype, feature number) and feature parameters (dimensional parameters, accuracy level, surface roughness), such as Figure 1 Next, we can use the part information model to calculate various sub-similarity and further obtain the comprehensive similarity between the two parts.

[0103] 1.1 Part attribute similarity

[0104] Part attribute information mainly includes part type and part material. Part type can be divided into two categories: rotational body type and non-rotational body type. Since design changes occur between parts with high similarity, process routes can only be reused and changed when the part types are the same. Therefore, the part type similarity can be defined as shown in formula (1). For part materials, their performance parameters include various types, but the material strength and hardness have the greatest impact on process route planning. In subsequent calculations, more commonly used and representative performance indicators are selected: tensile strength is selected for strength, and Brinell hardness is selected for hardness. The part material similarity can be expressed as shown in formula (2).

[0105]

[0106] Where, Represent the part type similarity and part material similarity respectively; and Represents the tensile strength limit of the target part and the typical part respectively; HBS obj and HBS typ Represents the Brinell hardness of the target part and the typical part respectively. σ and α H Respectively represent the influence weights of strength and hardness on the similarity of part materials, and satisfy α σ +α H =1.

[0107] The part attribute similarity is defined as the product of part type similarity and part material similarity, as shown in formula (3).

[0108]

[0109] Where S Ι Represents the similarity of part attributes.

[0110] 1.2 Part feature structure similarity

[0111] Feature structure similarity mainly refers to the similarity of the type, geometric dimensions, and relative position of two features. If two features have high similarity, first, the main type of the feature must be the same, second, the subtype of the feature must be the same, and finally, the position requirements must be met. In spacecraft cabin parts, the main type of feature can generally include holes, slots, bosses, windows, contours, etc., and the subtype of feature is based on the further subdivision of a main type. For example, holes can be further subdivided into through holes, tapered holes, countersunk holes, countersunk holes, and threaded holes. In terms of position requirements, the most important factors affecting feature processing and process route planning are the edge and centripetality of the feature. Therefore, features can be divided into four categories according to position requirements: centripetal and edge, centripetal non-edge, non-centripetal but edge, and non-centripetal non-edge.

[0112] Under the above conditions, the characteristic structure similarity of the present invention is carried out in the following three steps:

[0113] (1) Define the feature judgment factor and judge whether two features have high similarity based on the feature main type, subtype and position requirements, as shown in formula (4);

[0114] (2) In the target part and the typical part, the number of parts with the same manufacturing features is counted based on the feature judgment factor, as shown in formula (5);

[0115] (3) Calculate the feature structure similarity according to formula (6);

[0116]

[0117] Where, Represents the feature judgment factor, F x ,F y Respectively represent two manufacturing features that are compared with each other; represents the kth manufacturing feature under the jth subtype of the ith main type in the target part and the typical part respectively; N M Indicates the number of feature main types, N S (i) represents the number of characteristic subtypes under the i-th main type, N F (ij) represents the number of features under the jth subtype of the i-th main type, N FS Represents the total number of identical features between the target part and the typical part; N obj ,N tpy Represents the total number of features of target parts and typical parts respectively; S ΙΙ Indicates the feature structure similarity.

[0118] 1.3 Feature Change Similarity

[0119] For the same features in two parts, the feature structure similarity can be calculated. However, during the feature design change process, the target part also has a certain similarity with the changed features of the typical part, which can also be calculated using the similarity formula. For this reason, the present invention consciously introduces feature change similarity to accurately express the similarity of the changed features. Based on the analysis of the feature structure of similar parts, the present invention divides feature changes into feature addition, feature deletion, and feature transformation. Next, the feature change similarity will be calculated based on these three change types respectively.

[0120] For newly added features, similarity calculation can be performed based on the main type / subtype of the newly added features, as shown in formula (7):

[0121]

[0122] Where, Indicates the similarity of a single newly added feature of the target part relative to the typical part.

[0123] Similarly, the feature deletion calculation formula is as shown in formula (8):

[0124]

[0125] Where, Indicates the similarity of the target feature to a single subtracted feature of a typical part.

[0126] However, for feature transformation, which includes many variable situations such as feature main type transformation, subtype transformation, and size parameter transformation, similarity calculation is often more complex and diverse, as shown in formula (9):

[0127]

[0128] Where, represents the similarity of a single transformed feature of the target part relative to the typical part; φ represents the transformation similarity operator, which is used to calculate the similarity of two features with the same main type and subtype; F obj ,F typ Represent the manufacturing features of the target part and the typical part respectively. For two features with the same main type and sub-type, their geometric dimensions, accuracy levels and surface roughness are different. Therefore, φ(F obj ,F typ ) It is necessary to calculate the similarity of the size parameters, accuracy level and surface roughness, and then calculate the feature transformation similarity according to their respective weights, as shown in formulas (10)-(13):

[0129]

[0130] φ(F obj ,Ftyp )=α D S SIZ +α AC S ACC +α R S ROU (13)

[0131] Where S SIZ 、S ACC 、S ROU They represent the similarity of dimensional parameters, accuracy level and surface roughness respectively; Represent the qth size of the target part and the typical part under the pth transformation feature respectively; Represent the qth accuracy level of the target part and the typical part under the pth transformation feature, They represent the qth precision surface roughness of the target part and the typical part under the pth transformation feature respectively; M(p), N(p), and L(p) represent the number of sizes, the number of precision grades, and the number of surface roughness under the pth transformation feature respectively; α D , α AC , α R Respectively represent the influence weights of size parameters, accuracy level and surface roughness on the similarity of feature parameters, and satisfy α D +α AC +α R =1.

[0132] Since the probability of adding, deleting, and transforming features in the target part is basically the same, the degree of influence on the feature change similarity is the same, so the feature change similarity is calculated as shown in formula (14):

[0133]

[0134] Where S ΙΙΙ Indicates the similarity of feature changes, N add 、N del 、N alt They respectively represent the number of new features, the number of deleted features, and the number of transformed features of the target part relative to the typical part.

[0135] 1.4 Comprehensive Similarity

[0136] The comprehensive similarity between the target part and the typical part is calculated by integrating the part attribute similarity, part feature structure similarity and feature change similarity, as shown in formula (15):

[0137] S=α Ι S Ι +α ΙΙ S ΙΙ +α ΙΙΙ SΙΙΙ (15)

[0138] Where S represents the comprehensive similarity between the target part and the typical part. Ι , α ΙΙ , α ΙΙΙ Respectively represent the influence weights of part attributes, feature structures and feature parameters on the comprehensive similarity, and satisfy α Ι +α ΙΙ +α ΙΙΙ = 1. In subsequent calculations, the above-mentioned influence weights can be determined through the hierarchical analysis method.

[0139] After extensive case testing in the machining workshop, the comprehensive similarity of the vast majority of highly similar parts is distributed between 0.7 and 0.85. Therefore, the present invention believes that during the design change process, if the comprehensive similarity of the target part to the typical part meets the requirements, the target part and the typical part have high similarity, and the target part process route can be further obtained through design changes based on the existing process route. Otherwise, the similarity between the target part and the typical part is considered insufficient, and the target part process route needs to be replanned.

[0140] 2. Feature-oriented part process routing update process:

[0141] For aerospace manufacturing companies, the processing sequences, processing plans, and manufacturing resources for different parts of the same type are generally highly similar and important during process planning and production. Traditional process route planning involves planning individual parts one by one, without considering the relevance and similarity between parts. This leads to high repetitive functions, a large amount of mechanical operations, and is time-consuming, labor-intensive, and inefficient. Therefore, based on the repetitive processes of typical parts, the present invention implements feature-oriented process changes based on a process step priority graph and a 0-1 process step matrix, thereby efficiently and quickly updating the process routes of similar parts.

[0142] First, the purpose of performing feature-oriented design changes based on typical parts is to quickly and efficiently plan the process route for the target part. Based on the part information model constructed above, design changes can be divided into three categories: feature addition, feature deletion, and feature transformation. The difficulty and change process of each type of change vary, but the overall design concept is consistent and can be divided into the following three steps:

[0143] (1) Determine the processing plan for the changed features;

[0144] (2) Update the work step priority diagram and work step matrix of the target part;

[0145] (3) Solve the optimal process route based on the re-optimization genetic algorithm.

[0146] Specifically, for feature addition, deletion, and transformation, the first step is to determine the processing plan, work step sequence, machine tool, tool, and TAD for each changed feature. Secondly, determine the relative position of the work step corresponding to the changed feature in the work step priority diagram. If the changed feature is a deleted feature, the work step of the deleted feature must be deleted from the work step priority diagram of the typical part; if the changed feature is a transformed feature, the work step corresponding to the feature before the transformation must be deleted from the work step priority diagram of the typical part, and then the work step of the feature after the transformation must be added. After completing the above steps, the work step priority diagram of the target part is formed.

[0147] Based on the work step priority diagram of the target part, a 0-1 work step matrix can be obtained. By inputting relevant conditions, the process route of the target part can be updated according to the re-optimization genetic algorithm.

[0148] In summary, when updating a process route, the primary changes are the feature's machining plan, the sequence of work steps, and various manufacturing resources. For established machining shops, the selection of manufacturing resources like machine tools and cutting tools is relatively stable, while the machining plan for each feature is relatively variable. Therefore, determining the specific machining plan for each feature is crucial for feature changes.

[0149] Furthermore, the part process route updating process of the present invention is further described in detail.

[0150] 2.1 Determination of the change process processing plan

[0151] In a stable machining workshop, each type of machining feature has a certain number of machining solutions, each with its own optimal process parameter range. To determine the optimal machining solution for a modified feature, this paper establishes a similarity formula between the modified feature's process parameters and the available process parameters for the machining solution, thereby matching the optimal machining solution.

[0152] It is known that there are many factors that affect the selection of feature processing solutions, among which the most important are the feature's dimensional parameters, precision level, and surface roughness. Therefore, dimensional parameters, precision level, and surface roughness will serve as the main basis for calculating the matching degree of the feature processing solution. Common similarity calculation methods include Euclidean distance, cosine similarity, and Pearson correlation coefficient. Since the process parameters of the processing solution come from different dimensions, such as different numerical ranges and unit types of dimensional parameters and precision levels, it is obviously inappropriate to use traditional Euclidean distance and Pearson correlation coefficient. Therefore, this paper chooses cosine similarity for calculation, as shown in formula (16).

[0153]

[0154] In the formula, X represents the vector composed of the changed characteristic size parameters, accuracy level and surface roughness. For the convenience of expression, it is called the changed characteristic parameter vector. Y represents the optimal parameter vector of the processing solution. i Indicates the value of the i-th parameter in X, y i represents the i-th parameter value in Y. It can be seen that the smaller the difference between the changed feature parameter vector and the optimal parameter vector of the processing plan, the greater the cosine similarity.

[0155] However, not every parameter in the processing scheme process parameters is equally important. It is known that the degree of influence on the selection of processing schemes is ranked from large to small as follows: dimensional parameters, precision level and surface roughness. Therefore, this paper introduces the process parameter influence weight α on the basis of cosine similarity to distinguish the importance of each parameter and more accurately calculate the similarity of processing schemes. For processing schemes, the number of times each processing scheme is used varies depending on the factors such as processing difficulty, manufacturing resources, personal preferences, and historical data that the process personnel take into account in the long-term selection process. Therefore, compared with the traditional similarity formula, the present invention also considers the frequency of process personnel's selection of processing schemes and introduces the processing scheme usage coefficient β to make the processing scheme similarity formula more reasonable. The final processing scheme similarity formula is shown in formula (17).

[0156]

[0157] Where, β represents the utilization coefficient of the i-th processing scheme; α represents the influence weight of the i-th process parameter;

[0158] The process parameter influence weight α is calculated by AHP and satisfies formula (18):

[0159]

[0160] The utilization coefficient β of the processing scheme is determined by the historical selection probability of the scheme, and the calculation method is shown in (19):

[0161]

[0162] Where n i represents the number of times the i-th processing scheme is selected, N represents the total number of times similar features are processed, and k represents the production coefficient, which is determined by the specific manufacturing enterprise. The relationship is shown in formula (20). The β value can be calculated based on the production situation of the specific processing and manufacturing enterprise.

[0163]

[0164] Where m represents the number of all similar processing plans of a manufacturing enterprise.

[0165] 2.2 Work step priority diagram change process

[0166] After determining the machining plan for the changed feature, update the step priority diagram. The feature's machining plan corresponds to the machining steps in the step priority diagram. Steps that need to be deleted can generally be directly deleted from the step priority diagram. For new or modified steps, it's crucial to accurately locate their location in the typical part's step priority diagram before adding or modifying them.

[0167] It's important to note that the location and sequence of work steps during feature changes must adhere to common machining practices. Therefore, it's important to determine whether the changed feature is a datum feature. If it's not, then when updating the work step priority diagram, the work steps corresponding to the changed feature can be added, deleted, or modified directly. However, if the changed feature is a datum feature, the change will impact the machining of other features based on it. Therefore, when updating the work step priority diagram, appropriate adjustments to the newly added work steps must be made based on common machining practices.

[0168] like Figure 2 As shown in the example, Part 2 can be considered as a result of a design change to Part 1 based on features F2 and F3. Features F2 and F3 are modified to form features F2' and F3'. Feature F2's hole B is a non-datum feature in Part 1, and its modification to the inner contour of feature F2' has no direct impact on other features. Therefore, after the machining plan for feature F2' is determined, the machining plan for the original feature F2 can be directly replaced in the process step priority diagram. However, a change to feature F3 requires appropriate adjustments to the newly added process steps. Since feature F3 is a datum feature in Part 1, it will affect the subsequent machining of features F1 and F2. The rough surface of feature F3 is transformed into the fine surface of feature F3', which will add a new process step. According to common machining knowledge and prior knowledge, fine machining should not be performed directly after rough machining. Instead, fine machining should be performed only after both features F1 and F2' are machined to ensure the overall machining quality of Part 2.

[0169] 2.3 Solving the optimal process route based on the existing re-optimization genetic algorithm

[0170] In a mature production shop, the available manufacturing resources for various parts, including machine tools, cutting tools, and fixtures, are generally relatively stable, and it's rare for them to be completely unavailable. Therefore, the shop's existing infrastructure is used to provide various manufacturing resources for the features after the design change. The updated step priority diagram is used to generate a new 0-1 step matrix, which is then solved using a reoptimization genetic algorithm to further determine the optimal new process route for the part after the design change.

[0171] 3. Feasibility evaluation:

[0172] Since the new process routes created after a design change involve issues such as processing plan selection, process step sequencing, and resource matching, they are inevitably subject to irrationality. If these process routes are directly applied to the manufacturing of aerospace parts, they can lead to unfeasible process solutions, unusable parts, and even serious problems such as product scrapping. Therefore, to avoid these situations, it is necessary to effectively evaluate the quality of the process routes after the design change. Only when the new process route meets the set requirements can it be applied to actual part production. To this end, this paper comprehensively considers five factors: process cost, processing time, processing difficulty, the degree of technician requirements, and process route feasibility. A fuzzy comprehensive evaluation method is proposed to evaluate the process routes after the design change and further determine the quality of the process routes.

[0173] The construction of fuzzy comprehensive evaluation method includes the following aspects:

[0174] 3.1 Factor Set

[0175] For aerospace parts, there are many factors that influence the quality and feasibility of the process route. The most important factors are: process cost, processing time, processing difficulty, the degree of requirements for the technician, and the feasibility of the process route. The specific meanings are as follows:

[0176] (1) Process cost: mainly includes labor operation cost, machine tool use cost, tool use cost, and energy consumption cost.

[0177] (2) Processing time: mainly includes loading and unloading time, machine operation time, tool change time, cutting time and rest time.

[0178] (3) Processing difficulty: mainly reflected in the material properties of parts, complexity of geometric shapes, dimensional accuracy requirements, surface quality requirements, manufacturing resource requirements, etc.

[0179] (4) The degree of requirements for process engineers: mainly reflected in the design difficulty of process engineers, the proficiency of manual operation, the number of workers required for processing, etc.

[0180] (5) Feasibility of process route: It is mainly reflected in whether the process conforms to common sense of processing and whether the selection of manufacturing resources is standardized.

[0181] The five factors of process cost, processing time, processing difficulty, degree of process engineer requirements and process route feasibility are combined into a set, as shown in formula (21), and a comprehensive evaluation factor set of the process route is established, as shown in Figure 3 shown.

[0182] U={u1,u2,u3,u4,u5} (21)

[0183] Where U represents the factor set of fuzzy comprehensive evaluation of process route, u1, u2, u3, u4, and u5 represent five different factors affecting the process route.

[0184] 3.2 Evaluation Set

[0185] The evaluation set is a collection of evaluation results for each indicator in the factor set. When evaluating the factors affecting the process route after a design change, some factors cannot be quantitatively described with specific values, such as processing difficulty and the degree of requirements for the technician. Therefore, qualitative evaluation is often used in the evaluation.

[0186] Taking the qualitative value as the evaluation alternative set, the fuzzy evaluation set of the process route is established, as shown in formula (22).

[0187] V=(v1,v2,v3,v4,v5) (22)

[0188] In the formula, V represents the fuzzy evaluation set for the process route, and v1, v2, v3, v4, and v5 represent excellent, good, fair, poor, and poor, respectively. If the evaluation result of the process route after the design change reaches excellent or good, the process route is accepted and can be subsequently applied to actual processing and manufacturing. When evaluating different objects or indicators, the specific meanings represented by the evaluation set vary, but the degree of decrease is consistent. For example, when evaluating processing time, v1, v2, v3, v4, and v5 represent short, relatively short, average, relatively long, and long, respectively.

[0189] right Figure 3 The specific evaluation results are shown in Table 1.

[0190] Table 1 Description of evaluation levels

[0191]

[0192] 4.3 Weight Set

[0193] Because each factor in the fuzzy comprehensive evaluation factor set for a process route has varying degrees of influence on the process route, it is necessary to further determine the weight coefficient for each factor. Commonly used weight determination methods include the Analytic Hierarchy Process (AHP) and the Entropy Weight Method (EWM). However, since most factors influencing the process route are non-numerical or non-deterministic, the evaluation process is subject to excessive human subjectivity, which can easily lead to deviations from reality. Therefore, the present invention combines literature and expert evaluation to determine the weight coefficient for each factor.

[0194] The weight coefficient affecting the process route must meet specific conditions, as shown in formula (23).

[0195]

[0196] Where λ iRepresents the weight coefficient of the i-th influencing factor, i=1,2,3,4,5.

[0197] Combine the literature and expert opinions to determine each weight coefficient λ i , forming a weight set A, the specific values are shown in formula (24).

[0198] A=(0.1,0.1,0.3,0.2,0.3) (24)

[0199] 3.4 Judgment Matrix

[0200] In order to obtain an accurate and reliable judgment matrix, the present invention selects experienced workshop technicians, senior process engineers, researchers from the machining research institute, and senior professors in the field of machining to form a review panel. They conduct a single factor evaluation on each factor in the factor set U, further complete the evaluation of all factors, and determine the membership degree of the evaluation set V. The single factor evaluation set is shown in formula (25).

[0201] r i =(r i1 ,r i2 ,r i3 ,r i4 ,r i5 ) (25)

[0202] Where r i represents the single factor evaluation set of the i-th factor in the factor set, r i1 ,r i2 ,r i3 ,r i4 ,r i5 They respectively represent the membership of the i-th factor to the evaluation set V.

[0203] Furthermore, a single factor evaluation is performed on the factor set to generate a fuzzy judgment matrix, as shown in formula (26):

[0204]

[0205] In the formula, R represents the fuzzy judgment matrix, r ij It represents the membership value of the i-th factor in the factor set to the j-th element in the evaluation set. Usually, after normalization, r ij Satisfaction relationship.

[0206]

[0207] 3.5 Comprehensive evaluation model

[0208] The weight set and judgment matrix obtained above are subjected to fuzzy operation, as shown in formula (28), and the fuzzy comprehensive evaluation result can be obtained after normalization.

[0209] B=A·R (28)

[0210] Where B represents the evaluation result vector.

[0211] According to the maximum membership principle, the process route evaluation result is obtained, and then the feasibility of the part process route update is judged based on the evaluation result.

[0212] [Example]

[0213] The method for updating the process route of spacecraft cabin parts taking into account design changes disclosed in the present invention will be further explained below with reference to specific application cases.

[0214] In order to verify the effectiveness of the process route update method considering design changes, a case study of target cabin parts and typical parts was conducted based on the process route of typical cabin parts. Figure 4 As shown in the figure, the two parts have certain similarities in terms of materials, structure, and features. Both parts are spacecraft servo cabin parts, and both are made of aluminum tubes. The typical part includes 7 hole features, 1 window feature, 2 slot features, 3 surface features, and 1 end face feature. Target part A includes 4 hole features, 3 slot features, 2 window features, 3 surface features, and 1 end face feature. The feature information of the typical cabin part is shown in Table 2, and the feature comparison of the two parts is shown in Table 3. The specific feature changes of target part A relative to the typical part are shown in Table 4.

[0215] Table 2 Typical cabin parts feature information

[0216]

[0217]

[0218] Table 3 Comparison of part features

[0219]

[0220] Table 4 Target part A feature change table

[0221]

[0222] Design change process:

[0223] (1) Similarity calculation

[0224] Since the target part A and the typical part have the same type and material, they have the same strength and hardness, so the part type similarity and part material similarity of the target part A are shown in formula (29):

[0225]

[0226] According to formula (3), the part attribute similarity of target part A is calculated as shown in formula (30).

[0227]

[0228] Table 5 Statistics of feature changes

[0229]

[0230] Next, according to formula (6), the feature structure similarity is calculated. First, the target part and the typical part are compared to determine the number of features under different main types and subtypes, as shown in Table 5. Then, according to the data in Table 5, N is obtained. M 、N S (i) and N F (ij) value, further calculate N FS Finally, the feature structure similarity S is calculated by formula (6). ΙΙ , as shown in formula (31).

[0231]

[0232] We further calculated the feature change similarity. Table 4 shows that the similarity of the newly added feature F13' is 0.5, the similarities of the deleted features F9 and F10 are 0.5 and 0.5, respectively, and the similarities of the transformed features are 0.691, 0.791, and 0, respectively. Thus, we can calculate the feature change similarity, as shown in Equation (32).

[0233]

[0234] Finally, based on the obtained part attribute similarity, feature structure similarity, and feature change similarity, combined with their respective influence weights, the comprehensive similarity of the two parts can be calculated. Therefore, based on AHP and combined with the evaluation of the expert review panel, the corresponding weights of the three similarities are calculated as shown in Equation (33).

[0235] α Ι =0.427,α ΙΙ =0.498,α ΙΙΙ =0.073 (33)

[0236] Furthermore, the comprehensive similarity is calculated according to formula (15) as shown in formula (34).

[0237] S=α Ι S Ι +α ΙΙ S ΙΙ +α ΙΙΙ S ΙΙΙ =0.795 (34)

[0238] After calculation, since the comprehensive similarity in the target part A satisfies 0.795>0.7, the process route of the target part A can be further obtained by making design changes based on the existing typical process route.

[0239] (2) Design change process

[0240] The feature-solution similarity values are calculated for each changed feature in target part A. If a feature corresponds to multiple processing solutions, the optimal processing solution is matched to that feature based on the maximum similarity value. If a feature corresponds to only one processing solution, the selected solution is selected directly, as shown in Table 6. Next, appropriate manufacturing resources are selected for the changed feature based on the selected processing solution, forming a process step information table for the target part, as shown in Table 7.

[0241] Table 6 Calculation of change feature similarity and determination of processing plan

[0242]

[0243] Table 7 Target Part A Process Step Information Table

[0244]

[0245]

[0246] Furthermore, based on the above diagram, the work step priority diagram is changed. The specific process is as follows: Figure 5 As shown. Take the typical part step priority diagram as the initial step priority diagram, as shown Figure 5 a) Delete the steps to be deleted involved in the feature deletion and feature transformation process one by one, and complete the step deletion process in the priority graph, such as Figure 5 b); Secondly, the new steps involved in the feature addition and feature transformation process are added to the appropriate position according to the constraint rules to complete the step addition process, such as Figure 5 c); Finally, the process step priority diagram of the target part A is formed, such as Figure 5 d) Obtain the 0-1 step matrix of target part A based on the generated step priority graph, as shown in Figure 6 shown.

[0247] After completing the above steps, the process route of target part A is obtained by using the re-optimization genetic algorithm. The iterative image is as follows: Figure 7 The lowest processing cost is 3376, and one of the corresponding optimal process routes is shown in Table 8.

[0248] Table 8 Process route of target part A

[0249]

[0250] (3) Comprehensive evaluation

[0251] To effectively evaluate the quality of the changed process route, we first conducted a process analysis of target part A based on the actual process route after the design change, as shown in Table 9. A fuzzy evaluation of each factor u was performed by a review panel composed of experienced workshop technicians, senior process engineers, researchers at the Machining Research Institute, and senior professors in the field of machining, to obtain the evaluation results.

[0252] Table 9 Process analysis of target part A

[0253]

[0254]

[0255] Comprehensively considering the five factors that affect the quality of the process route after the design change, Equation (36) shows that the evaluation has the highest membership, 0.32. According to the maximum membership principle, the quality of the process route after the design change is good. The evaluation results show that the process route is practical and also demonstrate the feasibility of the process route update method considering design changes.

[0256] In summary, in order to solve the problem of updating the process route considering design changes, the present invention, based on the comprehensive similarity, changes the step priority diagram and the 0-1 step matrix based on the typical process route, and then optimizes the changed process route, and uses the fuzzy comprehensive evaluation method to evaluate the quality of the new process route. First, a comprehensive similarity formula for part similarity calculation is constructed considering the three aspects of part attribute similarity, feature structure similarity, and feature change similarity. Secondly, when the similarity reaches a certain threshold, the process route can be updated based on the typical part to consider the design changes. Furthermore, based on the feature differences (addition, deletion, transformation) between the target part and the typical part, the step priority diagram and the 0-1 step matrix are dynamically adjusted, and the optimal process route after the change is solved by re-optimizing the genetic algorithm. Finally, a fuzzy comprehensive evaluation is performed considering five factors: process cost, processing time, processing difficulty, manufacturing resource requirements, and process route feasibility, and the feasibility of the changed process route is effectively evaluated.

[0257] The above embodiments are not limitations of the present invention, and the present invention is not limited to the above examples. Any changes, modifications, additions or substitutions made by technicians in this technical field within the scope of the technical solution of the present invention also fall within the scope of protection of the present invention.

Claims

1. A method for updating the process route of spacecraft cabin parts taking into account design changes, characterized by: The following steps are involved: Step 1: Build a part information model and design a multi-dimensional comprehensive similarity calculation formula between two parts based on the part information model. The multi-dimensional calculation formula integrates part attribute similarity, part feature structure similarity, and feature change similarity; Step 2: When the similarity between two parts reaches a set threshold, a feature-oriented part process route update plan is designed based on the repetitive process of typical parts; The process of updating the part process route includes: dynamically adjusting the step priority diagram and 0-1 step matrix based on the feature differences between the target part and the typical part, including feature addition, deletion, and transformation, and using a re-optimization genetic algorithm to solve the optimal process route; Step 3: Based on the fuzzy comprehensive evaluation method, an evaluation mechanism for the updated part process route is constructed; Step 4: Evaluate the process route update plan designed in step 2 based on the evaluation mechanism. If the evaluation requirements are met, the process route update plan is determined to be feasible.

2. The method for updating the process route of spacecraft cabin parts considering design changes according to claim 1, characterized in that: In step 1, the constructed part information model includes three aspects and a total of 9 indicators, namely: Part attribute information: part type, part material; Part feature structure information: feature location, feature main type, feature subtype, and feature number; Characteristic parameters: dimensional parameters, accuracy grade, surface roughness.

3. The method for updating the process route of spacecraft cabin parts considering design changes according to claim 2, characterized in that: In step 1, the comprehensive similarity between the target part and the typical part is calculated by integrating the part attribute similarity, part feature structure similarity, and feature change similarity, as shown in formula (15): S=a Ι S Ι +a ΙΙ S ΙΙ +a ΙΙΙ S ΙΙΙ (15) In the formula, S represents the comprehensive similarity of the target part to the typical part, α Ι , α ΙΙ , α ΙΙΙ Respectively represent the influence weights of part attributes, part feature structures and feature parameters on the comprehensive similarity, and satisfy α Ι +α ΙΙ +α ΙΙΙ =1.

4. The method for updating the process route of spacecraft cabin parts considering design changes according to claim 3 is characterized in that: In step 1, the part attribute similarity is the part type similarity Similarity to part material The product of the two terms is shown in formula (3). Where S Ι Indicates the similarity of part attributes; similarity of part types As shown in formula (1), the part material similarity It is expressed as shown in formula (2); Where, and Represents the tensile strength limit of the target part and the typical part respectively; HBS obj and HBS typ Represents the Brinell hardness of the target part and the typical part respectively; α σ and α H Respectively represent the influence weights of strength and hardness on the similarity of part materials, and satisfy α σ +α H =1.

5. The method for updating the process route of spacecraft cabin parts considering design changes according to claim 3 is characterized in that: In step 1, the calculation of the part feature structure similarity is carried out in the following three steps: 1) Define the feature judgment factor and judge whether two features have high similarity based on the feature main type, subtype and position requirements, as shown in formula (4); 2) In the target part and the typical part, the number of parts with the same manufacturing features is counted based on the feature judgment factor, as shown in formula (5); 3) Calculate the feature structure similarity according to formula (6); Where, Represents the feature judgment factor, F x ,F y Respectively represent two manufacturing features that are compared with each other; represents the kth manufacturing feature under the jth subtype of the ith main type in the target part and the typical part respectively; N M Indicates the number of feature main types, N S (i) represents the number of characteristic subtypes under the i-th main type, N F (ij) represents the number of features under the jth subtype of the i-th main type, N FS Represents the total number of identical features between the target part and the typical part; N obj ,N tpy Represents the total number of features of target parts and typical parts respectively; S ΙΙ Indicates the similarity of part feature structure.

6. The method for updating the process route of spacecraft cabin parts considering design changes according to claim 3, characterized in that: In step 1, the calculation process of feature change similarity is as follows: Feature changes are divided into feature addition, feature deletion, and feature transformation; Among them, the similarity calculation of the newly added features is shown in formula (7): Where, Indicates the similarity of a single newly added feature of the target part relative to the typical part; The similarity calculation of feature deletion is shown in formula (8): Where, Indicates the similarity of the target feature to a single deleted feature of a typical part; The similarity calculation of feature transformation is shown in formula (9): Where, represents the similarity of a single transformed feature of the target part relative to the typical part; φ represents the transformation similarity operator, which is used to calculate the similarity of two features with the same main type and subtype; F obj ,F typ Represent the manufacturing features of target parts and typical parts respectively; φ(F obj ,F typ ) It is necessary to calculate the similarity of the size parameters, accuracy level and surface roughness, and then calculate the feature transformation similarity according to their respective weights, as shown in formulas (10)-(13): φ(F obj ,F typ )=a D S SIZ +a AC S ACC +a R S ROU (13) Where S SIZ 、S ACC 、S ROU They represent the similarity of dimensional parameters, accuracy level and surface roughness respectively; Represent the qth size of the target part and the typical part under the pth transformation feature respectively; Represent the qth accuracy level of the target part and the typical part under the pth transformation feature, They represent the qth precision surface roughness of the target part and the typical part under the pth transformation feature respectively; M(p), N(p), and L(p) represent the number of sizes, the number of precision grades, and the number of surface roughness under the pth transformation feature respectively; α D , α AC , α R Respectively represent the influence weights of size parameters, accuracy level and surface roughness on the similarity of feature parameters, and satisfy α D +α AC +α R =1; Since addition, deletion, and transformation have the same impact on feature change similarity, the feature change similarity is calculated as shown in formula (14): Where S ΙΙΙ Indicates the similarity of feature changes, N add 、N del 、N alt They respectively represent the number of new features, the number of deleted features, and the number of transformed features of the target part relative to the typical part.

7. The method for updating the process route of spacecraft cabin parts considering design changes according to claim 1, characterized in that: In step 2, the part process route update process specifically includes the following steps: 1) Determine the processing plan for each changed feature; 2) Determine the relative position of the work step corresponding to the changed feature in the work step priority graph, and form an updated work step priority graph of the target part based on the addition, deletion, and transformation operations of the feature differences; a 0-1 work step matrix can be obtained based on the work step priority graph of the target part; (3) Taking the 0-1 process step matrix as input, the reoptimization genetic algorithm is used to solve the optimal process route.

8. The method for updating the process route of spacecraft cabin parts considering design changes according to claim 7, characterized in that: In step 2, a similarity formula is established between the process parameters of the changed feature and the available process parameters of the processing plan to determine the optimal processing plan for the changed feature; Among them, the similarity formula for changing characteristic process parameters is shown in formula (16): Where X represents the vector composed of the changed characteristic size parameters, accuracy level and surface roughness; Y represents the optimal parameter vector of the processing solution; i Indicates the value of the i-th parameter in X, y i represents the value of the i-th parameter in Y; n represents the number of times the processing plan is selected; The similarity formula of the process parameters of the processing plan can be used, as shown in formula (17): Where β represents the utilization coefficient of the i-th processing scheme; α represents the influence weight of the i-th process parameter.

9. The method for updating the process route of spacecraft cabin parts considering design changes according to claim 1, characterized in that: In step 3, the comprehensive evaluation model is shown in formula (28): B=A·R (28) In the formula, B represents the evaluation result vector, and A represents the weight set; According to the maximum membership principle, the process route evaluation result is obtained, and then the feasibility of the part process route update is judged based on the evaluation result.

10. The method for updating the process route of spacecraft cabin parts considering design changes according to claim 9, characterized in that: In step 3, during the construction of the fuzzy comprehensive evaluation method: Establish the comprehensive evaluation factor set U of the process route, as shown in formula (21): U={u1,u2,u3,u4,u5} (21) Where U represents the factor set of fuzzy comprehensive evaluation of process route, u1, u2, u3, u4, and u5 represent the five factors affecting the process route, namely process cost, processing time, processing difficulty, degree of requirement of process engineer, and feasibility of process route; 2) Taking the qualitative value as the evaluation alternative set, the fuzzy evaluation set V of the process route is established, as shown in formula (22): V=(v1,v2,v3,v4,v5) (22) Where, represents the fuzzy evaluation set of the process route, v1, v2, v3, v4, and v5 represent excellent, good, medium, poor, and bad, respectively; The weight coefficients of the factors determined by combining literature and expert evaluation are required to meet specific conditions for the weight coefficients affecting the process route, as shown in formula (23): Where λ i represents the weight coefficient of the i-th influencing factor, i = 1, 2, 3, 4, 5; Combine the literature and expert opinions to determine each weight coefficient λ i , forming a weight set A, the specific values are shown in formula (24): A＝(0.1,0.1,0.3,0.2,0.3) (24) The membership degree of the evaluation set V is determined by the judgment matrix; the single factor evaluation set is shown in formula (25): r i ＝(r i1 ,r i2 ,r i3 ,r i4 ,r i5 ) (25) Where r i represents the single factor evaluation set of the i-th factor in the factor set, r i1 ,r i2 ,r i3 ,r i4 ,r i5 They respectively represent the membership of the i-th factor to the evaluation set V. Perform single factor evaluation on the factor set and generate a fuzzy judgment matrix, as shown in formula (26): In the formula, R represents the fuzzy judgment matrix, r ij It represents the membership value of the i-th factor in the factor set to the j-th element in the evaluation set; After normalization, r ij Satisfaction relationship:

Images