Template-based aeronautical part technological procedure reasoning method and system

Abstract

The invention relates to the technical field of intelligent manufacturing of aeronautical parts, and discloses an aeronautical part technological procedure reasoning method and system based on a template, and the method comprises the steps: firstly obtaining and storing technological characteristics and geometric structure characteristics of a to-be-machined part; secondly, encoding the feature data of the to-be-processed part, the feature data of the historical processing part, the technological procedure template data and the technological knowledge data which need to be used in a One-hot mode; then, according to the process characteristics and geometric structure characteristics of the to-be-machined part, a matched process procedure template is obtained through a characteristic-based template matching algorithm, and typical process parameters related to part process procedure compilation based on the process procedure template are obtained through a typical process parameter reasoning decision algorithm; and the matched technological procedure template and the typical technological parameters obtained through calculation are fused, and a part technological procedure is formed and output. According to the method, the required part technological procedure can be quickly and intelligently generated, and the design efficiency of the technological design link in the production process is greatly improved.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent manufacturing of aviation parts, and more specifically to a template-based aviation parts process specification reasoning method and system. Background Art

[0002] In the traditional process design process, the preparation of process specifications is too dependent on the personal experience and level of the process designers. The rationality, operability and preparation time of process documents mainly depend on the experience and proficiency of the process personnel; at the same time, the preparation cycle is long, there is a lot of repetitive work, and it is difficult to adapt to changes in products and manufacturing environments.

[0003] However, the internal structure of the aircraft is complex, the parts are of different shapes and in large quantities, and the development process is very complicated. Most of the products are accompanied by the characteristics of multiple varieties and small batches. Due to limited personal experience and different design levels, the designed process documents vary from person to person, and there are problems such as poor consistency, difficulty in optimization, and weak standardization. In addition, the inheritance of process design experience is also difficult. With the transformation and upgrading of the aerospace manufacturing industry, new requirements have been put forward for the preparation of process procedures for complex products such as modern aircraft products. Therefore, it is of great significance to push relevant knowledge to the preparation of process procedures and realize the rapid and intelligent preparation of part process procedures. Summary of the invention

[0004] The present invention proposes a template-based aviation part process specification reasoning method and system. The present invention quickly and intelligently generates the required part process specifications through the part features of the parts to be processed and the part features and processing data of the existing historically processed parts, thereby greatly improving the design efficiency of the process design link in the production process, improving the efficiency and intelligence of the production process, and further improving the quality and reliability of the part process design.

[0005] In order to achieve the above-mentioned object of the invention, the technical solution of the present invention is as follows: The present invention discloses a template-based aviation parts process specification reasoning method, comprising the following steps: Step S1. extracting the three-dimensional digital model information and features of the part to be processed, extracting the process features and geometric structure features of the part to be processed and storing them; Step S2. Encode the required feature data of the part to be processed, the feature data of the historical processed parts, the process specification template data and the process knowledge data in a one-hot manner; Step S3. According to the process features and geometric structure features of the part to be processed, a matching process specification template is obtained through a feature-based template matching algorithm; wherein the feature-based template matching algorithm includes: The Pearson correlation coefficient and mutual information are used to measure the correlation between part features and template identification values ​​respectively, and then the two measurement results are combined in a weighted manner to select the feature attributes with the highest correlation to establish a set; Define categories based on the template identification value data set, calculate the expected information required for sample classification, and divide the data set by feature attributes to calculate the entropy or expected information of the subsets; Convert the initial decision tree into a set of condition-action pairs and perform attribute pruning on the rules; Use the pruned rule set to classify and predict the data samples of the parts to be processed, and output the matching process specification template; Step S4. According to the process characteristics and geometric structure characteristics of the part to be processed, the typical process parameters involved in completing the part process specification compilation based on the process specification template are obtained through the typical process parameter reasoning decision algorithm; Step S5. Fusing the matched process specification template with the typical process parameters obtained by the typical process parameter reasoning decision algorithm to form a part process specification; Step S6: Edit the part process specification to finally obtain the part process specification that can be put into production.

[0006] Preferably, in step S2, the feature attributes of the part feature library and the template identification value used to uniquely identify the template in the process specification template library are encoded in a one-hot manner, each element is associated with a unique integer index, and is converted into a binary vector of the same size as the vocabulary length, thereby obtaining the corresponding feature attribute data set. and template identification value data set ,in, , X i Indicates the first i Part features; x ij Indicates i The corresponding j Class feature attributes, y j Indicates j The number of part instances that the class template identifies the value of the data set.

[0007] Preferably, in step S3, obtaining a matching process specification template through a feature-based template matching algorithm includes: Step S31. Use the Pearson correlation coefficient to measure the correlation between the feature attributes of the part feature library and the process specification template library. The calculation expression is as follows: Formula (1); in, Represents the Pearson correlation coefficient calculation expression; and Respectively represent i Part feature and template identification values Indicates i The covariance between the part features and the template identification values; represents the variance of the variable; Respectively represent i The average value of the part feature and template identification values; represents the expected value of a variable; Indicates the first i The covariance between the part features and the template identification value; Indicates the first i The variance between the part features and the template identification value; Step S32. Combine the mutual information to measure the correlation between the feature attributes of the part feature library and the process specification template library. The calculation expression is as follows: Formula (2); in, Represents the mutual information correlation calculation expression; Indicates i The information entropy of each part feature; Indicates the template identification value Y Under the conditions i The information entropy of each part feature; Indicates entropy calculation; represents the calculation of conditional entropy; Step S33. Combine the Pearson correlation analysis result and the mutual information correlation analysis result using a weighted method. The calculation expression is as follows: Formula (3); in, To measure the i Part Features Given this information, identify the value of the template the amount of information provided; Indicates that the mutual information calculation value is standardized; α and β is the weight, ; Step S34. Take the highest correlation in step S33 K c Feature attributes to create a collection ; in, Represents the corresponding t related part features, and there are ; Step S35. Assume that the template identification value data set have m Different values, defined m Different categories Assumptions Is a class C l The number of samples in , use formula (4) to calculate the expected information required for a given sample classification; (4); in Any sample belongs to C l The probability of estimate, Indicates the identification value corresponding to the actual sample; Represents the expected information required to calculate the given sample classification, represents a discrete variable; Step S36. Set feature attributes X i have v Different values , using feature attributes X i Will D c Divide into v Subset ;set up d lh Is a subset D h Medium C l The number of samples, according to X i The calculation expression of the entropy or expected information of the partitioned subset is shown as follows: Formula (5); in, Acting as the h The weight of the subset, h Subset representation Partitioned Subsets ; Represents the weighted value of the characteristic attribute set by the part design principle; Indicates based on Desired information for subset division; represents the normalization factor; Step S37. Convert the initial decision tree into a set of rules, each rule is represented as a condition-action pair, where the condition is a series of feature tests and the action is the corresponding category label; the generated rule set is , r i Indicates i rules, Representation Rules r i The condition part, Representation Rules r i The action part; Step S38: prune the generated rules, and the pruned rule set is ,in is the pruned rule, It's a rule The condition part, It's a rule The action part; Step S39. For the new part data sample to be processed, use the pruned rule set Perform classification prediction and finally output the process specification template corresponding to the parts to be processed.

[0008] Preferably, in step S4, the typical process parameter reasoning decision algorithm includes a process flow decision reasoning algorithm and a typical process parameter decision reasoning algorithm.

[0009] Preferably, the process flow decision reasoning algorithm is as follows: Step a. Use index encoding to encode each process flow method, and then use the embedding layer with word embedding size r to obtain the vector of each process flow: m Dimension; all process attributes are represented by character-level one-hot vectors, and each character is represented by n dimensional One-hot vector representation, forming the maximum number of characters ( N )× n Dimensional matrix; Step b. adopt the feature vector superposition into complex vector feature fusion method, encode the previous process sequence through the double-layer DNN-LSTM layer, take the process feature of each time step and the process attribute feature of the current input for vector superposition; Step c. Use a two-layer DNN-LSTM to maintain the sequence relationship between processes under process attributes, and generate the subsequent process flow by decoding the vector obtained by fusing the process attributes with the process flow sequence.

[0010] Preferably, step c specifically includes: first, passing the vector obtained by fusing the process attributes and the process flow sequence through a double-layer DNN-LSTM layer, then taking the vector output by the last hidden state of the double-layer DNN-LSTM as the vector generated by decoding, and finally normalizing the decoded vector through the softmax function of the fully connected layer to obtain the confidence distribution of the current generated process flow, and using a greedy search algorithm to select the process flow with the largest confidence in the confidence distribution of the current generated process flow as the generated process flow and output it.

[0011] Preferably, the typical process parameter decision reasoning algorithm is as follows: Step d. Based on the IF-THEN rule, define the selection decision rules around the process knowledge base, the tool selection knowledge base, and the process parameter decision knowledge base, so that each rule has a formal description of the IF condition and the THEN conclusion; Step e. Constructing a knowledge base including process decision knowledge, process parameter decision knowledge and tool selection knowledge; Step f. Search the knowledge base for matching rules according to the process flow and workpiece material, and finally output the typical process parameters corresponding to the matching rules.

[0012] Preferably, the typical process parameters include process flow, tooling and process parameters.

[0013] Based on the same inventive concept, the present invention also discloses a template-based aviation part process specification reasoning system on the other hand, the system is used to implement the above-mentioned template-based aviation part process specification reasoning method, including: The part feature library is used to store the part process features and part geometric structure features of the parts to be processed, as well as the part process features and part geometric structure feature data of the historically processed parts.

[0014] Process specification template library, used to store process specification template data; A process knowledge base is used to store process knowledge for making process selection decisions based on part features, store tool selection decision knowledge based on part features, and store process parameter decision knowledge based on part features; The process specification reasoning module is used to obtain a matching process specification template according to the process characteristics and geometric structure characteristics of the part to be processed through a feature-based template matching algorithm; and according to the matching process specification template, through a typical process parameter reasoning decision algorithm, obtain the process parameters involved in completing the part process specification compilation based on the process specification template; The data fusion module is used to fuse the matching process specification template with the typical process parameters obtained by the typical process parameter reasoning decision algorithm to form the part process specification; The human-computer interaction module is used to output the part process specification obtained by the data fusion module to the process specification generation interface, and allow the process designer to edit the obtained part process specification.

[0015] Preferably, the part feature library includes a part process feature library and a part geometry feature library; wherein the part features include part process features and part geometry features; the part process feature library is used to store part process feature information; and the part geometry feature library is used to store part geometry feature information.

[0016] Preferably, the process knowledge base includes a process flow knowledge base, a tool selection knowledge base and a process parameter decision knowledge base.

[0017] Preferably, the process procedure reasoning module includes a feature-based template matching module and a typical process parameter reasoning decision module.

[0018] Preferably, the typical process parameter reasoning and decision module includes a process flow reasoning and decision module and a process parameter reasoning and decision module.

[0019] On the other hand, the present invention also discloses a computer device, including a memory, a processor, and a computer program stored in the memory and executable in the processor, wherein when the processor executes the computer program, the above-mentioned template-based aviation part process procedure reasoning method is implemented.

[0020] A computer-readable storage medium stores a computer program, which, when executed in a computer processor, implements the above-mentioned template-based aviation part process specification reasoning method.

[0021] Beneficial effects of the present invention: The present invention analyzes the traditional industrial design process in the current manufacturing process and proposes a method and system for inferring the process specification of aviation parts based on templates, which is used to improve the efficiency and intelligence of the compilation of process specifications of parts. The process specification inference method of aviation parts based on templates can quickly and intelligently generate the required process specifications of parts through the part features of the parts to be processed and the part features and processing data of the existing historical processed parts, which greatly improves the design efficiency of the process design link in the production process, improves the efficiency and intelligence of the production process, and further improves the quality and reliability of the process design of parts. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] The foregoing and following detailed description of the present invention will become more apparent when read in conjunction with the following drawings, in which: Figure 1 is a flow chart of the method of the present invention; Figure 2 Schematic diagram of the system of the present invention. DETAILED DESCRIPTION

[0023] In order to enable those skilled in the art to better understand the technical solutions in the present invention, the technical solutions for achieving the purpose of the present invention will be further described below through specific embodiments. It should be noted that the technical solutions claimed for protection in the present invention include but are not limited to the following embodiments. Based on the embodiments in the present invention, all other embodiments obtained by those skilled in the art without creative work shall fall within the scope of protection of the present invention.

[0024] Due to the continuous development of the current aviation aircraft manufacturing industry, a high degree of informatization and automation has been achieved in the product design and product manufacturing links in the entire process. For traditional process design, it varies greatly depending on the company's resources and process habits. Under the same resources and constraints, different process designers may formulate different process regulations. This is a decision-making process with strong empiricism and many influencing factors. Therefore, traditional industrial design usually has the following defects: (1) Inconvenient for standardization and poor consistency. Due to limited personal experience and different design levels, the designed process documents vary from person to person, with poor consistency, difficult optimization, weak standardization, and difficulty in inheriting process design experience.

[0025] (2) The design cycle is long and cannot quickly respond to product changes. When a product is replaced, even if some of its parts are similar to the original parts, new process procedures must be redesigned, resulting in a large amount of duplication of work and difficulty in adapting to changes in products and manufacturing environments.

[0026] (3) Over-reliance on the personal experience and level of process designers. The rationality, operability and preparation time of process documents mainly depend on the experience and proficiency of process designers.

[0027] (4) It is difficult to ensure the accuracy of data. The process needs to process a large amount of graphic information, data information, and text information, and a large amount of process files and process data are generated through process design. The traditional process design method of processing data is very backward and cumbersome, and it is very easy to make mistakes.

[0028] (5) It is not convenient to concentrate process experts, experience and knowledge and make full use of them.

[0029] (6) Process personnel mainly perform repetitive and tedious work and lack research on innovative process work.

[0030] The internal structure of the aircraft is complex, the parts are of different shapes and in large quantities, and the development process is very complicated. Most of its products are accompanied by the characteristics of multiple varieties and small batches. With the transformation and upgrading of the aerospace manufacturing industry, new requirements have been put forward for the preparation of process specifications for complex products such as modern aircraft products. Therefore, it is of great significance to realize the rapid and intelligent preparation of part process specifications.

[0031] Based on this, an embodiment of the present invention proposes a template-based aviation part process specification reasoning method and system. The present invention first introduces and illustrates the process specification reasoning method. Figure 1 For the flow chart of the method of the present invention, please refer to the attached specification Figure 1 The template-based aviation parts process specification reasoning method specifically includes the following steps: Step S1. Extract the three-dimensional digital model information and features of the part to be processed, extract the process features and geometric structure features of the part to be processed and store them.

[0032] In the embodiments described in the present invention, a feature refers to a collection of all characteristic attributes of a part, and each characteristic attribute is a subdivided feature of a part. For example, a part has three characteristic attributes, such as part type, material name, and material specification, but these characteristic attributes are collectively referred to as the features of the part. The process features and geometric structure features of a part constitute part features.

[0033] Step S2: Encode the required feature data of the part to be processed, the feature data of the historically processed parts, the process specification template data and the process knowledge data in a one-hot manner.

[0034] In the embodiment described in the present invention, the specific steps of step S2 are: The characteristic attributes of the part feature library (part type, material name, material specification, material grade, material status, technical requirements and geometric features, etc.) and the template identification value used to uniquely identify the template in the process specification template library are encoded in a one-hot manner. Each element is associated with a unique integer index and converted into a binary vector of the same size as the vocabulary length, thereby obtaining the corresponding characteristic attribute data set. and template identification value data set ,in, , X i Represents the feature of the part in the part instance, that is, the i A set of characteristic attributes for a part; x ij Indicates i The corresponding j Class feature attributes, yj Indicates j The number of part instances that the class template identifies the value of the data set.

[0035] Step S3. According to the process features and geometric structure features of the part to be processed, a matching process specification template is obtained through a feature-based template matching algorithm.

[0036] In the embodiment described in the present invention, the step S3 specifically includes the following steps: Step S31. In order to measure the correlation between the feature attributes of the part feature library and the process specification template library, the Pearson correlation coefficient is first used to measure the correlation between the feature attributes of the part feature library and the process specification template library. The calculation expression is as follows: Formula (1); in, Represents the Pearson correlation coefficient calculation expression; and Respectively represent i Part feature and template identification values Indicates i The covariance between the part features and the template identification values; represents the variance of the variable; Respectively represent i The average value of the part feature and template identification values; represents the expected value of a variable; Indicates the first i The covariance between the part features and the template identification value; Indicates the first i The variance between the part features and the template identification value; Step S32. Further, the correlation between the feature attributes of the part feature library and the process specification template library is measured by mutual information, and the calculation expression is as follows: Formula (2); in, Represents the mutual information correlation calculation expression; Indicates i The information entropy of each part feature; Indicates the template identification value Y Under the conditions i The information entropy of each part feature; Indicates entropy calculation, used to measure the template identification value uncertainty; Represents the calculation of conditional entropy. Given the i Part Features In the case of There is still uncertainty; Step S33. In order to make up for the limitations of different correlation measures and make full use of multiple measures in correlation analysis, this embodiment uses a weighted method to combine the Pearson correlation analysis results and the mutual information correlation analysis results, and improves the reliability of the measurement through the mutual complementation of the two measures; the specific calculation expression is as follows: Formula (3); in, To measure the i Part Features Given this information, identify the value of the template The amount of information provided and weighting to adjust the calculation; Indicates that the mutual information calculation value is standardized; α and β is the weight, ; Step S34. Take the highest correlation in step S33 K c Feature attributes to create a collection , which is used for the subsequent decision tree method construction of gain value weighting and rule generation attribute pruning. After feature correlation analysis, a number of features with a high degree of correlation with the research object are selected as the calculation object of the subsequent steps. The calculation object is the selected K c feature set of feature attributes D c ;in, Represents the corresponding t related part features, and there are ; Step S35. From step S2, it can be seen that D c yes t A collection of part instance data samples. Assuming the template identification value dataset have m Different values, defined m Different categories Assumptions Is a class C l The number of samples in , use formula (4) to calculate the expected information required for a given sample classification; Formula (4); in Any sample belongs to C l The probability of estimate, Indicates the identification value corresponding to the actual sample; Represents the expected information required to calculate the given sample classification, Represents a discrete variable, that is, the sample classification identification value corresponding to template m; Step S36. Assume feature attributes X i have v Different values , using feature attributes X i Will D c Divide into v Subset ,in, D h Include D There are some samples like this, they are in X i Has value a h .if X i is chosen as the test attribute (i.e., the best partitioning attribute), then these subsets correspond to the subsets consisting of D The branch growing from the node of d lh Is a subset D h Medium C l The number of samples, according to X i The calculation expression of the entropy or expected information of the partitioned subset is shown as follows: Formula (5); in, Acting as the h The weight of the subset, h Subset representation Partitioned Subsets , used for information representation; Indicates based on The expected information of the subset, The smaller the value of, the better the subset division result; Represents the weighted value of the feature attribute set by the part design principle, which is used to adjust the influence of different features; represents the normalization factor, which is calculated by summing a set of samples to ensure that the calculation ratio is reasonable. D h , its expected information is: ,in , d h express Cl The number of samples in p lh Subset D h Any data sample belongs to the category C l The probability of Step S37. Convert the initial decision tree into a set of rules, each rule is represented as a condition-action pair, where the condition is a series of feature tests and the action is the corresponding category label; the generated rule set is , r i Indicates i rules, Representation Rules r i The condition part, Representation Rules r i The action part; Step S38. Attribute pruning is performed on the generated rules. The goal of attribute pruning is to remove unimportant features in the rules to simplify the rules and improve generalization performance. The pruned rule set is recorded as ,in is the pruned rule, It's a rule The condition part, It's a rule The action part; Step S39. For the new part data sample to be processed, use the pruned rule set For classification prediction, each rule in the rule set will treat the processed part data sample according to the conditional part x A test is performed. If the condition is partially met, the corresponding action part is returned as the predicted category label, and finally the process specification template corresponding to the part to be processed is output.

[0037] In the embodiment described in the present invention, the initial decision tree is constructed based on the part feature attributes and the template identification value data set. The specific process and principle are as follows: 1. Data preparation Input data: Feature attributes of the part feature library (such as part type, material name, technical requirements, etc.), and process specification template identification value; Each feature attribute is one-hot encoded to form a binary feature vector; The Template ID value is used to define a category label, indicating the template category that the part matches.

[0038] 2. Decision Tree Generation Algorithm Decision trees are usually generated based on the following principles: (1) Information gain Each split of the decision tree selects the feature with the largest information gain as the basis for splitting the current node The information gain calculation expression is as follows: ; in, For the current sample set D The entropy of D i It is a subset divided according to a certain feature; For the i The weight of the subset; For the i The entropy of the subset; (2) Calculation of expected information entropy In the process of generating a decision tree, the expected information entropy of each feature is calculated by the following formula: ; in, m is the number of categories; p k The sample belongs to k Probability of class; (3) Optimal partition attribute selection According to the information gain formula, the feature that reduces entropy the fastest is selected as the basis for division.

[0039] 3. Decision tree construction process Root node: Use the principle of maximizing information gain to select the first feature and divide the sample set D ; Child nodes: For each subset D i , repeat the above steps and continue to select the feature with the largest information gain until the stopping condition is met.

[0040] 4. Stop Condition (1) All samples in the sample set belong to the same category; (2) The remaining features cannot be further divided (or the maximum depth is reached); (3) The number of samples is too small to be effectively divided; 5. Initial form of decision tree The generated initial decision tree is a hierarchical tree structure; Node: Each node represents a feature test.

[0041] Branch: The branch corresponds to the feature value (such as the binary value of one-hot encoding).

[0042] Leaf node: A leaf node corresponds to a category label, that is, a template identification value.

[0043] Furthermore, in the embodiment described in the present invention, how the initial decision tree in step S37 is converted into a set of rules, the essence of the decision tree is to divide the data set into multiple subsets through testing, and finally make each subset belong to the same category as much as possible. Therefore, each path of the decision tree represents a series of characteristic conditions, and its end point (leaf node) corresponds to a category label. This path can naturally be converted into a condition-action rule. The specific conversion process and principle are as follows: 1. The correspondence between decision trees and rules (1) A decision tree is a tree structure that contains: Internal nodes: represent feature tests (such as ); Branch: represents the value of a feature (such as ); Leaf node: represents the category label (action part); (2) A rule is a condition-action pair of the form: ; : The conditional part of the rule, which represents the combination of feature tests from the root node to the leaf node; : The action part of the rule, indicating the category label (classification result) that the path finally corresponds to; Indicates i A specific rule, which comes from the rule set .

[0044] (3) Each path from the root node to the leaf node represents a rule, for example: path: ; Corresponding rules: .

[0045] 2. Conversion process Step (1) Traverse the decision tree Starting from the root node of the decision tree, traverse down along each path until you reach a leaf node; the path from the root node to the leaf node contains a series of feature tests, and each path corresponds to a rule; Step (2) Extract the condition part

[0046] Each internal node (feature test) in the path becomes the conditional part of the rule, and the conditional parts are combined by logical AND (∧). For example: Internal Node and .

[0047] The extracted condition part is ; Step (3) Extract the action part

[0048] The action part is taken from the leaf node (category label) at the end of the path, for example: The category label of the leaf node is C The action part is ; Step (4) Generate rules The conditional part and the action part Combined into a rule, the rule format is as follows: .

[0049] Furthermore, in the embodiment described in the present invention, for the attribute pruning in step S38, weighted information gain can be used to evaluate the importance of features, filter out key features, and delete unimportant features. This process not only optimizes the complexity of the rules, but also improves the generalization ability, but it is necessary to balance the degree of simplification and classification performance, and ensure the pruning effect through the verification data set. The specific pruning process is as follows: 1. Data preparation enter: Generated rule set R , where each rule R i Include condition part C i and the action part A i ; The conditional part of each rule consists of several features; Target: Remove unimportant features in the condition part and retain the most important features to improve the generalization performance of the rules.

[0050] 2. Feature Importance Assessment Use weighted information gain to evaluate the importance of each feature. The specific steps are as follows: (1) Information gain calculation formula: A feature in the calculation rule F j Information gain: ; in, Dataset covered by the rule D Entropy of D i Photo FeaturesF j The subset after partitioning; For the i The weight of the subset; It is i The entropy of the subset; The meaning is the number of subsets after division, that is, the data set D Divided into Subset ; (2) Weighted processing For each feature F j The information gain introduces weight w j : ; Weight w j Reflection characteristics F j Global importance in the rule set; (3) Normalization The weighted information gain is standardized to facilitate feature sorting and selection. The calculation formula is as follows: .

[0051] 3. Feature Screening According to the standardized information gain, the importance of features is determined and unimportant features are removed, including the following: (1) Importance threshold screening: Set a threshold θ to retain only features where the normalized information gain is greater than θ. The threshold θ can usually be set experimentally. (2) Before reservation k Features: If the rules are complex, you can directly retain the top ones in information gain ranking k Features simplify the conditional part of the rule.

[0052] 4. Update the rule set The pruned rule set is updated to: ; It is the conditional part after the construction, only the important features are retained; Represents the pruned rules The action part; Representation Rules The conditional part.

[0053] For example: Assumption Rules Ri The conditional part C i Included Features , the calculated weighted information gain is: F 1:0.4; F 2:0.3; F 3:0.2; F 4:0.1; According to the threshold θ=0.2, retain F 1, F 2, F 3. Delete F 4 , then the pruned rules are: .

[0054] Step S4. According to the process characteristics and geometric structure characteristics of the part to be processed, the process parameters involved in completing the part process specification compilation based on the process specification template are obtained through a typical process parameter reasoning decision algorithm.

[0055] In the embodiment described in the present invention, the typical process parameter reasoning decision algorithm includes a process flow decision reasoning algorithm and a typical process parameter decision reasoning algorithm.

[0056] Furthermore, the specific steps of the process flow decision reasoning algorithm are as follows: Step a. Build a two-layer DNN-LSTM network (deep neural network combined with long short-term memory network), integrate the advantages of DNN in feature extraction and transformation and the characteristics of LSTM in sequence structure analysis, and effectively combine the two; on this basis, adopt a two-layer DNN-LSTM structure, the upper layer can capture deeper context dependencies, and the lower layer can extract context hidden dependencies, and finally get the output as ,in, H m and U m Respectively represent the network fully connected weight matrix and bias vector between the lower layer DNN-LSTM and the upper layer DNN-LSTM; Represents the hidden state of the DNN-LSTM unit, that is, the direct output of the DNN-LSTM; It represents the final output after linear transformation, and its function is to be used for prediction results of downstream tasks. The input of the two-layer DNN-LSTM includes two categories: the previous process sequence and process attributes; For the process sequence, we first normalize the description of each process, use index encoding for each process method, and then embed the word r The embedding layer obtains the vector of each process flow as m dimension; all process attributes are represented by character-level one-hot vectors, and each character is represented by a one-hot vector of dimension, forming a maximum number of characters Dimensional matrix; Step b. Integrate the process attribute information into each process flow information, use the process attribute information to better guide the generation of the next process flow, and make the generated process flow more consistent with the process steps in the process described by the process attribute information. The specific operation is: adopt the feature vector superposition into complex vector feature fusion method, encode the previous process flow sequence through a double-layer DNN-LSTM layer, take the process flow features of each time step and the current input process attribute features for vector superposition; Step c. Use a double-layer DNN-LSTM to maintain the sequence relationship between processes under process attributes, and generate the subsequent process flow by decoding the vector obtained by fusing the process attributes with the process flow sequence. This process first passes the vector obtained by fusing the process attributes and the process flow sequence through a double-layer DNN-LSTM layer, then takes the vector output by the last hidden state of the double-layer DNN-LSTM as the vector generated by decoding, and finally normalizes the decoded vector through the softmax function of the fully connected layer to obtain the confidence distribution of the current generated process flow, and uses a greedy search algorithm to select the process flow with the highest confidence in the confidence distribution of the current generated process flow as the generated process flow and output it.

[0057] Furthermore, the specific steps of the typical process parameter decision reasoning algorithm are as follows: Step d. Based on the IF-THEN rule, define the selection decision rules around the process knowledge base, the tool selection knowledge base, and the process parameter decision knowledge base, so that each rule has a formal description of the IF condition and the THEN conclusion; Step e. Constructing a knowledge base including process decision knowledge, process parameter decision knowledge and tool selection knowledge; Step f. When the user provides information such as process flow and workpiece material as input, the system searches for matching rules in the knowledge base based on these conditions. Once a matching rule is found, the system will automatically push the corresponding typical process parameters to the user based on the THEN conclusion of the rule, without the need for the user to manually search and set them.

[0058] Step S5: The matched process specification template is integrated with the process parameters obtained by the typical process parameter reasoning decision algorithm to form a part process specification.

[0059] Step S6: Edit the part process specification to finally obtain the part process specification that can be put into production.

[0060] Based on the same inventive concept, an embodiment of the present invention also discloses a template-based aviation part process specification reasoning system. Since the principle of solving the problem of this system is similar to the template-based aviation part process specification reasoning method, the implementation of the system can refer to the implementation of the method, and the repeated parts will not be repeated. As used below, the term "unit" or "module" can be a combination of software and / or hardware that implements a predetermined function. Although the system described in the following embodiments is preferably implemented in software, the implementation of hardware, or a combination of software and hardware, is also possible and conceivable. Figure 2 The schematic diagram of the structure of the template-based aviation parts process specification reasoning system provided by the embodiment of the present invention is as follows: Figure 2 As shown, the system may include: a storage module, a process procedure reasoning module, a data fusion module and a human-computer interaction module; wherein, The storage module further includes a parts feature library, a process specification template library and a process knowledge library; wherein, The part feature library is used to store the features of the parts to be processed and the feature data of the parts processed in history, and is composed of a part process feature library and a part geometric structure feature library; the part process feature library is used to store the part process feature information; the part geometric structure feature library is used to store the part geometric structure feature information; The process specification template library is used to store process specification template data; the process specification templates are used to establish different types of typical parts process specifications according to the characteristic models of different typical parts to form different templates; wherein different process specification templates contain the technical requirements, process routes, process of the same type of typical parts, the names of the process steps used and other related information that should be standardized and normalized; The process knowledge base includes a process flow knowledge base, a tool selection knowledge base and a typical process parameter decision knowledge base; wherein, The process knowledge base is used to store process knowledge for making process selection decisions based on part features; the process knowledge is the process knowledge for selecting and making decisions on process flows that has been collected, sorted, classified and summarized in the past, and is the rules for selecting and making decisions on process flows formed after being standardized and formalized; The tool selection knowledge base is used to store tool selection decision knowledge based on part features; the tool selection knowledge is the process knowledge for selecting tool decisions that has been collected, sorted, classified and summarized in the past, and is the rules for selecting tool decisions that have been formed after being standardized and formalized; The process parameter decision knowledge base is used to store process parameter decision knowledge based on part features; the process parameter decision knowledge is the process knowledge for selecting decision process parameters collected, sorted, classified and summarized in the past, and is the rule for selecting decision tooling after being standardized and formalized; the process parameters are the process parameters required by each process in the process specification template; The process procedure reasoning module includes a feature-based template matching module and a typical process parameter reasoning decision module. Further, the typical process parameter reasoning decision module includes a process flow reasoning decision module and a typical process parameter reasoning decision module; wherein, The feature-based template matching module obtains a matching process specification template according to the process features and geometric structure features of the part to be processed through a feature-based template matching algorithm; The process flow reasoning and decision module is used to reason and decide the process flow of the part process specification according to the matched process specification template; The typical process parameter reasoning and decision module is used to reason and decide the tooling tools and process parameter data required for compiling the part process specification; the typical process parameters are the process parameters involved in completing the compilation of the part process specification based on the process specification template; The data fusion module is used to fuse the matched process specification template with typical process parameter data such as process flow, tooling, process parameters, etc. obtained by reasoning and decision making to form a part process specification; the process flow contained in the process specification template matched by the template is determined in combination with the process flow data reasoned and decided by the typical process parameter reasoning and decision module, so as to complete the correct process specification template; the tooling data required for each process in the process specification template is determined and filled in through the tooling data reasoned and decided by the typical process parameter reasoning and decision module; the specific process parameters required to be determined for each process in the process specification template are determined and filled in through the tooling data reasoned and decided by the typical process parameter reasoning and decision module; The human-computer interaction module is used to output the part process specification obtained by the data fusion module and display it on the system's process specification generation interface, and allows process designers to edit the obtained part process specification. After editing by the process designers, the part process specification that can be put into production is obtained.

[0061] It should be noted that the systems, devices, models or units described in the above embodiments can be implemented by computer chips or entities, or by products with certain functions. For the convenience of description, in this specification, the above systems are described in various units according to their functions. Of course, when implementing the present invention, the functions of each unit can be implemented in the same or multiple software and / or hardware.

[0062] Furthermore, in the present specification, adjectives such as first and second may be used merely to distinguish one element or action from another, without necessarily or implying any actual such relationship or order.

[0063] Furthermore, on the other hand, this embodiment also provides a computer device, which includes a processor, an input device, an output device and a memory, and the processor, the input device, the output device and the memory are interconnected; wherein the memory is used to store a computer program, the computer program includes program instructions, and the processor is configured to call the program instructions to execute the steps in the above embodiment.

[0064] Furthermore, another aspect of this embodiment also provides a computer-readable storage medium, characterized in that: the computer-readable storage medium stores a computer program, the computer program includes program instructions, and when the program instructions are executed by a processor, the processor executes the steps in the above embodiment.

[0065] In this embodiment, the processor may be a central processing unit (CPU). The processor may also be other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, or a combination of the above chips.

[0066] The memory, as a non-transitory computer-readable storage medium, can be used to store non-transitory software programs, non-transitory computer executable programs and units, such as the corresponding program units in the above method embodiments of the present invention. The processor executes various functional applications of the processor and works data processing by running the non-transitory software programs, instructions and modules stored in the memory, that is, implementing the method in the above method embodiments.

[0067] The memory may include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application required for at least one function; the data storage area may store data created by the processor, etc. In addition, the memory may include a high-speed random access memory, and may also include a non-volatile memory, such as at least one disk storage device, a flash memory device, or other non-volatile solid-state storage device. In some embodiments, the memory may optionally include a memory remotely arranged relative to the processor, and these remote memories may be connected to the processor via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0068] The one or more units are stored in the memory, and when executed by the processor, the method in the above embodiment is performed.

[0069] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware.

[0070] The above description is only a preferred embodiment of the present invention and does not limit the present invention in any form. Any simple modification or equivalent change made to the above embodiment based on the technical essence of the present invention shall fall within the protection scope of the present invention.

Claims

1. A template-based reasoning method for aviation parts process specification, characterized in that: The following steps are involved: Step S1. extracting the three-dimensional digital model information and features of the part to be processed, extracting the process features and geometric structure features of the part to be processed and storing them; Step S2. Encode the required feature data of the part to be processed, the feature data of the historical processed parts, the process specification template data and the process knowledge data in a one-hot manner; Step S3. According to the process features and geometric structure features of the part to be processed, a matching process specification template is obtained through a feature-based template matching algorithm; wherein the feature-based template matching algorithm includes: The Pearson correlation coefficient and mutual information are used to measure the correlation between part features and template identification values ​​respectively, and then the two measurement results are combined in a weighted manner to select the feature attributes with the highest correlation to establish a set; Define categories based on the template identification value data set, calculate the expected information required for sample classification, and divide the data set by feature attributes to calculate the entropy or expected information of the subsets; Convert the initial decision tree into a set of condition-action pairs and perform attribute pruning on the rules; Use the pruned rule set to classify and predict the data samples of the parts to be processed, and output the matching process specification template; Step S4. According to the process characteristics and geometric structure characteristics of the part to be processed, the typical process parameters involved in completing the part process specification compilation based on the process specification template are obtained through the typical process parameter reasoning decision algorithm; Step S5. Fusing the matched process specification template with the typical process parameters obtained by the typical process parameter reasoning decision algorithm to form a part process specification; Step S6: Edit the part process specification to finally obtain the part process specification that can be put into production.

2. The template-based aviation parts process specification reasoning method according to claim 1 is characterized in that: In step S2, the feature attributes of the part feature library and the template identification value used to uniquely identify the template in the process specification template library are encoded in a one-hot manner, each element is associated with a unique integer index, and converted into a binary vector of the size of the vocabulary length, thereby obtaining the corresponding feature attribute data set and template identification value data set ,in, , X i Indicates the first i Part features; x ij Indicates i The corresponding j Class feature attributes, y j Indicates j The number of part instances that the class template identifies the value of the data set.

3. The template-based aviation parts process specification reasoning method according to claim 1 is characterized in that: In step S3, a matching process specification template is obtained by a feature-based template matching algorithm, including: Step S31. Use the Pearson correlation coefficient to measure the correlation between the feature attributes of the part feature library and the process specification template library. The calculation expression is as follows: Formula (1); in, Represents the Pearson correlation coefficient calculation expression; and Respectively represent i Part feature and template identification values; Indicates i The covariance between the part features and the template identification values; represents the variance of the variable; Respectively represent i The average value of the part feature and template identification values; represents the expected value of a variable; Indicates the first i The variance between the part features and the template identification value; Indicates the first i The covariance between the part features and the template identification value; Step S32. Combine the mutual information to measure the correlation between the feature attributes of the part feature library and the process specification template library. The calculation expression is as follows: Formula (2); in, Represents the mutual information correlation calculation expression; Indicates i The information entropy of each part feature; Indicates the template identification value Y Under the conditions i The information entropy of each part feature; Indicates entropy calculation; represents the calculation of conditional entropy; Step S33. Combine the Pearson correlation analysis result and the mutual information correlation analysis result using a weighted method. The calculation expression is as follows: Formula (3); in, To measure the i Part Features Given this information, identify the value of the template the amount of information provided; Indicates that the mutual information calculation value is standardized; α and β is the weight, ; Step S34. Take the highest correlation in step S33 K c Feature attributes to create a collection ; in, Represents the corresponding t related part features, and there are ; Step S35. Assume that the template identification value data set have m Different values, defined m Different categories Assumptions Is a class C l The number of samples in , use formula (4) to calculate the expected information required for a given sample classification; （4）； in Any sample belongs to C l The probability of estimate, Indicates the identification value corresponding to the actual sample; Represents the expected information required to calculate the given sample classification, represents a discrete variable; Step S36. Set feature attributes X i have v Different values , using feature attributes X i Will D c Divide into v Subset ;set up d lh Is a subset D h Medium C l The number of samples, according to X i The calculation expression of the entropy or expected information of the partitioned subset is shown as follows: Formula (5); in, Acting as the h The weight of the subset, h Subset representation Partitioned Subsets ; Represents the weighted value of the characteristic attribute set by the part design principle; According to Desired information for subset division; represents the normalization factor; Step S37. Convert the initial decision tree into a set of rules, each rule is represented as a condition-action pair, where the condition is a series of feature tests and the action is the corresponding category label; the generated rule set is , r i Indicates i rules, Representation Rules r i The condition part, Representation Rules r i The action part; Step S38: prune the generated rules, and the pruned rule set is ,in is the pruned rule, It's a rule The condition part, It's a rule The action part; Step S39. For the new part data sample to be processed, use the pruned rule set Perform classification prediction and finally output the process specification template corresponding to the parts to be processed.

4. The template-based aviation parts process specification reasoning method according to claim 1 is characterized in that: In step S4, the typical process parameter reasoning decision algorithm includes a process flow decision reasoning algorithm and a typical process parameter decision reasoning algorithm.

5. The template-based aviation parts process specification reasoning method according to claim 4 is characterized in that: The process flow decision reasoning algorithm is as follows: Step a. Use index encoding to encode each process flow method, and then use the embedding layer with word embedding size r to obtain the vector of each process flow: m Dimension; all process attributes are represented by character-level one-hot vectors, and each character is represented by n dimensional One-hot vector representation, forming the maximum number of characters ( N )× n Dimensional matrix; Step b. adopt the feature vector superposition into complex vector feature fusion method, encode the previous process sequence through the double-layer DNN-LSTM layer, take the process feature of each time step and the process attribute feature of the current input for vector superposition; Step c. Use a two-layer DNN-LSTM to maintain the sequence relationship between processes under process attributes, and generate the subsequent process flow by decoding the vector obtained by fusing the process attributes with the process flow sequence.

6. The template-based aviation parts process specification reasoning method according to claim 5 is characterized in that: The step c specifically includes: first, passing the vector obtained by fusing the process attributes and the process flow sequence through a double-layer DNN-LSTM layer, then taking the vector output by the last hidden state of the double-layer DNN-LSTM as the vector generated by decoding, and finally normalizing the decoded vector through the softmax function of the fully connected layer to obtain the confidence distribution of the current generated process flow, and using a greedy search algorithm to select the process flow with the largest confidence in the confidence distribution of the current generated process flow as the generated process flow and output it.

7. The template-based aviation parts process specification reasoning method according to claim 4 is characterized in that: The typical process parameter decision reasoning algorithm is as follows: Step d. Based on the IF-THEN rule, define the selection decision rules around the process knowledge base, the tool selection knowledge base, and the process parameter decision knowledge base, so that each rule has a formal description of the IF condition and the THEN conclusion; Step e. Constructing a knowledge base including process decision knowledge, process parameter decision knowledge and tool selection knowledge; Step f. Search the knowledge base for matching rules according to the process flow and workpiece material, and finally output the typical process parameters corresponding to the matching rules.

8. The template-based aviation parts process specification thrust method according to claim 7, characterized in that: The typical process parameters include process flow, tooling and process parameters.

9. A template-based aviation part process specification reasoning system, the system is used to implement the template-based aviation part process specification reasoning method according to any one of claims 1 to 8, characterized in that: include: A part feature library is used to store part process features and part geometric structure features of parts to be processed, as well as part process features and part geometric structure feature data of historically processed parts; Process specification template library, used to store process specification template data; A process knowledge base is used to store process knowledge for making process selection decisions based on part features, store tool selection decision knowledge based on part features, and store process parameter decision knowledge based on part features; The process specification reasoning module is used to obtain the matching process specification template according to the process characteristics and geometric structure characteristics of the parts to be processed through a feature-based template matching algorithm; And according to the matched process specification template, through the typical process parameter reasoning decision algorithm, the typical process parameters involved in completing the part process specification compilation based on the process specification template are obtained; The data fusion module is used to fuse the matching process specification template with the typical process parameters obtained by the typical process parameter reasoning decision algorithm to form the part process specification; The human-computer interaction module is used to output the part process specification obtained by the data fusion module to the process specification generation interface, and allow the process designer to edit the obtained part process specification.

10. The template-based aviation parts process specification reasoning system according to claim 9, characterized in that: The part feature library includes a part process feature library and a part geometric structure feature library; wherein the part features include part process features and part geometric structure features; the part process feature library is used to store part process feature information; and the part geometric structure feature library is used to store part geometric structure feature information.

11. The template-based aviation parts process specification reasoning system according to claim 9, characterized in that: The process knowledge base includes a process flow knowledge base, a tool selection knowledge base and a process parameter decision knowledge base.

12. The template-based aviation parts process specification reasoning system according to claim 9, characterized in that: The process procedure reasoning module includes a feature-based template matching module and a typical process parameter reasoning decision module.

13. The template-based aviation parts process specification reasoning system according to claim 12, characterized in that: The typical process parameter reasoning and decision module includes a process flow reasoning and decision module and a process parameter reasoning and decision module.

Images