Part process generation method and system based on process knowledge base

Abstract

The invention discloses a part process generation method and system based on a process knowledge base, and the method comprises the steps: S1, connecting a process knowledge base through system software of a terminal, and carrying out the identification and screening of related process data in the process knowledge base; s2, according to the related process data screened and identified in the process knowledge base in the step S1, performing multi-domain similarity calculation and matching to obtain a processing feature process chain of the features of the to-be-processed part, and generating a step sequence set; and S3, according to the work step sequence set obtained in the step S2, optimizing the work step sequence set through an optimization algorithm, and generating a process route of the to-be-processed part. According to the invention, automatic generation of parts from characteristics to processes is realized, and intelligent and datamation automatic process generation is realized, so that the production efficiency of products is improved to a great extent.

Description

Technical Field

[0001] The present invention relates to the technical field of process design dataization and intelligentization, and in particular to a method and system for generating a part process based on a process knowledge base. Background Art

[0002] With the rapid development of information technology and artificial intelligence, the digitalization and intelligent transformation of traditional industries has become a key strategy to promote the high-quality development of the manufacturing industry. At present, intelligent manufacturing, as the main direction of the transformation and upgrading of the manufacturing industry, requires the combination of traditional manufacturing processes with artificial intelligence to achieve automation and intelligence in the entire manufacturing process. Today's product demand is showing trends such as diversification and personalization, and the increase in the frequency of product design updates has placed higher requirements on process design.

[0003] At present, the industrial design of most companies still relies on manual design based on experience or manuals. The level of automation and intelligence in the design process is low, and the generated CNC machining process solutions are difficult to match the actual process of the parts, which ultimately leads to low product production efficiency. Summary of the invention

[0004] In order to overcome the deficiencies of the above technologies, the present invention provides a method and system for generating part processes based on a process knowledge base.

[0005] The technical solution adopted by the present invention to overcome the technical problems is:

[0006] A method for generating a part process based on a process knowledge base comprises the following steps:

[0007] S1. Connecting to a process knowledge base through the system software of the terminal, and identifying and screening relevant process data in the process knowledge base, wherein the relevant process data at least includes processing process chain library data, processing tool library data, processing machine tool equipment library data, processing fixture library data, and measuring tool library data;

[0008] S2, according to the relevant process data in the process knowledge base screened and identified in step S1, obtain the processing feature process chain of the feature of the part to be processed by multi-domain similarity calculation and matching, and generate a process step sequence set;

[0009] S3. According to the process step sequence set obtained in step S2, the process step sequence set is optimized by using an optimization algorithm to generate a process route for the part to be processed.

[0010] Furthermore, in step S1, the data included in the process knowledge base includes at least part entity data, processing knowledge data, and manufacturing resource data; wherein the part entity data includes at least part process data, part feature processing chain, and part processing process route, the processing knowledge data includes at least processing feature process chain, work step matching manufacturing resources, and work step constraints, and the manufacturing resource data includes at least processing machine tool equipment information, processing tool information, and processing fixture information.

[0011] Furthermore, step S1 specifically includes:

[0012] S11. The terminal establishes a connection with the process knowledge base through an Internet API interface to obtain relevant process data stored in the process knowledge base;

[0013] S12, by setting process rules, classify and mark the relevant process data obtained in step S11 according to the type and attribute of the data, and filter the classified data to delete redundant information and erroneous information;

[0014] S13. Use structured rules to map the unstructured data in the process knowledge base into structured data, and combine them to generate a structured data matrix.

[0015] Furthermore, step S2 specifically includes:

[0016] S21, encoding the features of the part to be processed according to the rules, then structuring the unstructured data in the part to be processed, and finally converting the feature data of the part to be processed into a part feature matrix;

[0017] S22, multi-domain similarity calculation of part features, based on the commonly used processing process chain library of processing features, obtain the processing feature process chain of each feature of the part to be processed;

[0018] S23. Generate a set of part process step sequences, and match each process step with corresponding manufacturing resources based on the process chain of processing features, wherein the process steps at least include processing methods, processing features, processing machines, processing tools and processing fixtures, and the manufacturing resources at least include processing machines, processing tools and processing resources.

[0019] Furthermore, in step S22, the multi-domain similarity calculation step of part features includes:

[0020] Firstly, the similarity calculation of part feature type, part feature size parameter and part feature accuracy parameter are performed respectively;

[0021] Then, the weights of the part feature type, part feature size parameter and part feature accuracy parameter are determined respectively, so as to calculate the comprehensive similarity of the part features.

[0022] Furthermore, the multi-domain similarity calculation of part features is as follows:

[0023] (1) Calculation of similarity of part feature types: A classification tree is established based on the part feature types. The nodes represent the part feature types, and the edges represent the relationship between the part feature types. The distance between two nodes represents the similarity of the part feature types. The specific calculation formula is as follows:

[0024] (1)

[0025] In formula (1), Indicates the similarity of feature types, and For two given nodes, Represents the root node, Indicates the total number of nodes, express and The nearest common ancestor node, represents the depth of the nearest common ancestor node, is the edge weight between nodes;

[0026] (2) Calculation of similarity of part feature size parameters: For features of the same feature type, similarity is calculated and the Euclidean distance is used to characterize the similarity. The specific calculation formula is as follows:

[0027] (2)

[0028] In formula (2), Indicates the similarity of part feature size parameters, and Representing two features The value of the size parameter, The number of size parameters representing the feature;

[0029] (3) Calculation of similarity of part feature accuracy parameters: A similarity calculation method based on fuzzy mathematics is used to characterize the similarity using the set closeness in fuzzy mathematics. The specific calculation formula is as follows:

[0030] Sim p = σ 0 (A,B)= 1 2 [A∘B+(1-A⨀B)] (3)

[0031] In formula (3), Indicates the similarity of part feature accuracy parameters, and are respectively the precision fuzzy feature subsets, represents the inner product operation of the precision fuzzy feature subset, The outer product calculation of the fuzzy feature subset representing the accuracy;

[0032] (4) Calculation of comprehensive similarity of part features: Determine the weights of part feature type, part feature size parameters, and part feature accuracy parameters. The calculation formula for the comprehensive similarity of part features is as follows:

[0033] + + (4)

[0034] In formula (4), is the comprehensive similarity of part features, They are respectively the weight of the similarity of part feature type, the weight of the similarity of part feature size parameters, and the weight of the similarity of part feature accuracy parameters.

[0035] Furthermore, step S3 specifically includes:

[0036] S31, determining the process route planning objective function according to the process rules and manufacturing resource information, wherein the process rules at least include datum first, primary first then secondary, rough first then fine, and surface first then hole;

[0037] S32, following the basic process sequence constraints of the parts, treating the process steps as nodes, using a graph to characterize the constraints between the process steps, and generating a process step sequence constraint graph;

[0038] S33. Based on the objective function established in step S31 and the process sequence constraint diagram generated in step S32, the ant colony algorithm is improved using the topological sorting idea, so that the optimal process step sequence set can be found under the condition of satisfying the constraints, thereby generating a process route for the part to be processed.

[0039] Further, in step S31, the process route planning goal is that the optimized process route has the highest processing efficiency, so that the number of machine tool replacements, the number of processing clamping times, and the number of processing tool replacements when processing a single part are minimized, and a processing machine tool replacement cost function, a processing clamping replacement cost function, and a processing tool replacement cost function are established;

[0040] (1) The calculation formula of the cost function of machine tool replacement is:

[0041] (5)

[0042] In formula (5), and Indicates two processing steps, and Respectively represent the processing steps and processing steps The machine tools used;

[0043] (2) The calculation formula of the processing clamping replacement cost function is:

[0044] (6)

[0045] In formula (6), and Respectively represent the processing steps and processing steps The fixture used, and Respectively represent the processing steps and processing steps The clamping surface of the part;

[0046] (3) The calculation formula of the machining tool replacement cost function is:

[0047] (7)

[0048] In formula (7), and Respectively represent the processing steps and processing steps The machining tools used;

[0049] (4) Calculate the weighted sum of the cost function of machine tool replacement, the cost function of fixture replacement, and the cost function of tool replacement as the final optimization objective function. The expression of the objective function is as follows:

[0050] + + (8)

[0051] In formula (8), , , They are the cost of replacing the processing machine tool, the cost of replacing the processing fixture, and the cost of replacing the processing tool. , , are the weight of the cost of replacing the machining machine tool, the weight of the cost of replacing the machining fixture, and the weight of the cost of replacing the machining tool, respectively. , according to the length of replacement time, set the weight .

[0052] Furthermore, in step S33, the ant colony algorithm is improved by using the topological sorting idea, which specifically includes:

[0053] (1) When the ant selects the initial node, it obtains the nodes with in-degree 0 in the work step sequence constraint graph, randomly selects a node from the obtained nodes with in-degree 0 as the initial node of the ant path, and deletes the node from the work step sequence constraint graph;

[0054] (2) Each time the next node to be visited is selected for the ant, the nodes with in-degree 0 in the remaining nodes of the work step sequence constraint graph are obtained. For these nodes, a roulette wheel method is used to select a node as the next node visited by the ant, and the node is deleted from the work step sequence constraint graph;

[0055] (3) Repeat step (2) until all nodes have been visited.

[0056] The present invention also discloses a part process generation system based on a process knowledge base, which is used in the part process generation method based on the process knowledge base, and comprises:

[0057] A data screening module is configured to connect to a process knowledge base through the system software of a terminal, and screen and identify relevant process data in the process knowledge base, wherein the relevant process data at least includes processing process chain library data, processing tool library data, processing machine tool equipment library data, processing fixture library data, and measuring tool library data;

[0058] A processing feature process chain generation module is configured to obtain a processing feature process chain of a part feature to be processed by multi-domain similarity calculation and matching according to relevant process data in a process knowledge base after screening and identification, and generate a process step sequence set;

[0059] The process route generation module is configured to optimize the process step sequence set according to the process step sequence set through an optimization algorithm to generate a process route for the part to be processed.

[0060] The beneficial effects of the present invention are:

[0061] 1. The present invention realizes automatic connection with a process knowledge base and automatic acquisition of relevant process data in the process knowledge base through a software terminal.

[0062] 2. Different from the common fuzzy similarity calculation method, the present invention adopts a multi-domain similarity calculation method to calculate the part feature type similarity, the part feature size parameter similarity and the part feature accuracy parameter similarity respectively, and then integrates the three to obtain the comprehensive similarity of the part features, making the calculation result more reasonable and accurate.

[0063] 3. The present invention adopts the topological sorting idea to improve the ant colony algorithm, so that the process sequence optimization efficiency is higher and the accuracy is higher.

[0064] 4. The present invention realizes the automatic generation of parts from features to processes, and realizes the intelligent and data-based automatic generation of processes, thereby greatly improving the production efficiency of products. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] Figure 1 This is a flow chart of the method for generating a part process based on a process knowledge base according to Example 1 of the present invention.

[0066] Figure 2 This is a schematic diagram of the process flow of optimizing the step sequence of the ant colony algorithm using the topological sorting concept as described in Example 1 of the present invention. DETAILED DESCRIPTION

[0067] In order to facilitate those skilled in the art to better understand the present invention, the present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. The following is only exemplary and does not limit the protection scope of the present invention.

[0068] Example 1

[0069] The present embodiment describes a method for generating a part process based on a process knowledge base, such as Figure 1 As shown, the following steps are included:

[0070] S1. Connect to the process knowledge base through the system software of the terminal, and identify and filter the relevant process data in the process knowledge base.

[0071] Step S1 specifically includes the following steps:

[0072] S11. The terminal establishes a connection with the process knowledge base through an Internet API interface to obtain relevant process data stored in the process knowledge base. The terminal described in this embodiment mainly refers to the front-end page.

[0073] Specifically, the data included in the process knowledge base include at least part entity data, processing knowledge data, and manufacturing resource data; among them, the part entity data includes at least part process data, part feature processing chain, and part processing process route, the processing knowledge data includes at least processing feature process chain, process step matching manufacturing resources, and process step constraints, and the manufacturing resource data includes at least processing machine tool equipment information, processing tool information, and processing fixture information. The above data information is stored in the database in the form of a table, providing a basis for subsequent similarity calculation and process optimization.

[0074] In this embodiment, the relevant process data at least includes processing process chain library data, processing tool library data, processing machine tool equipment library data, processing fixture library data, and measuring tool library data.

[0075] S12. By setting process rules, the relevant process data obtained in step S11 are classified according to the type and attribute of the data, and marked. At the same time, the classified data are screened to delete redundant information and erroneous information.

[0076] S13. Use structured rules to map the unstructured data in the process knowledge base into structured data, and combine them to generate a structured data matrix.

[0077] S2. According to the relevant process data in the process knowledge base screened and identified in step S1, the processing feature process chain of the to-be-processed part features is obtained through multi-domain similarity calculation and matching, and a process step sequence set is generated.

[0078] In this embodiment, step S2 specifically includes the following steps:

[0079] S21. Encode the features of the parts to be processed according to relevant rules. Generally, there are structured data and unstructured data in the parts to be processed. The unstructured data in the parts to be processed are structured to make the data types in the parts to be processed unified. Finally, the feature data of the parts to be processed are converted into a part feature matrix.

[0080] S22. Calculate the multi-domain similarity of part features, and obtain the processing feature process chain of each feature of the part to be processed based on the commonly used processing process chain library of processing features.

[0081] In this embodiment, the multi-domain similarity calculation step of part features includes:

[0082] Firstly, the similarity calculation of part feature type, part feature size parameter and part feature accuracy parameter are performed respectively;

[0083] Then, the weights of the part feature type, part feature size parameter and part feature accuracy parameter are determined respectively, so as to calculate the comprehensive similarity of the part features.

[0084] Furthermore, the multi-domain similarity calculation of part features is as follows:

[0085] (1) Calculation of similarity of part feature types: A classification tree is established based on the part feature types. The nodes represent the part feature types, and the edges represent the relationship between the part feature types. The distance between two nodes represents the similarity of the part feature types. The specific calculation formula is as follows:

[0086] (1)

[0087] In formula (1), Indicates the similarity of feature types, and For two given nodes, Represents the root node, Indicates the total number of nodes, express and The nearest common ancestor node, represents the depth of the nearest common ancestor node, is the edge weight between nodes;

[0088] (2) Calculation of similarity of part feature size parameters: For features of the same feature type, similarity is calculated and the Euclidean distance is used to characterize the similarity. The specific calculation formula is as follows:

[0089] (2)

[0090] In formula (2), Indicates the similarity of part feature size parameters, and Representing two features The value of the size parameter, The number of size parameters representing the feature;

[0091] (3) Calculation of similarity of part feature accuracy parameters: A similarity calculation method based on fuzzy mathematics is used to characterize the similarity using the set closeness in fuzzy mathematics. The specific calculation formula is as follows:

[0092] Sim p = σ 0 (A,B)= 1 2 [A∘B+(1-A⨀B)] (3)

[0093] In formula (3), Indicates the similarity of part feature accuracy parameters, and are respectively the precision fuzzy feature subsets, represents the inner product operation of the precision fuzzy feature subset, The outer product calculation of the fuzzy feature subset representing the accuracy;

[0094] (4) Calculation of comprehensive similarity of part features: Determine the weights of part feature type, part feature size parameters, and part feature accuracy parameters. The calculation formula for the comprehensive similarity of part features is as follows:

[0095] + + (4)

[0096] In formula (4), is the comprehensive similarity of part features, They are the weights of the similarity of part feature types, the weights of the similarity of part feature size parameters, and the weights of the similarity of part feature precision parameters. The weights are determined by setting multiple groups of weights for comparative analysis to obtain the optimal weight value.

[0097] S23, generating a set of part process step sequences, and matching corresponding manufacturing resources for each process step by the process chain of processing features, wherein the process step at least includes processing methods, processing features, processing machines, processing tools and processing fixtures, and the manufacturing resources at least include processing machines, processing tools and processing resources. In this embodiment, the matching of manufacturing resources is mainly carried out through human-computer interaction. For processing machines and processing tools, the system recommends suitable processing machines and processing tool types to process personnel by querying the MySQL database, and then the process personnel select the processing machines and processing tools required for each process step during processing in combination with the specific production situation. For processing fixtures and clamping surfaces, the process personnel need to determine the clamping according to the characteristics of different parts and processing features.

[0098] S3. According to the process step sequence set obtained in step S2, the process step sequence set is optimized by using an optimization algorithm to generate a process route for the part to be processed.

[0099] In this embodiment, step S3 specifically includes the following steps:

[0100] S31. Determine the process route planning objective function according to the process rules and manufacturing resource information, wherein the process rules at least include datum first, primary first then secondary, rough first then fine, and surface first then hole.

[0101] Specifically, the datum comes first: before the start of processing, the machining surface of the precision datum needs to be machined first, so as to provide an accuracy reference for the machining of other machining surfaces.

[0102] First the main surface, then the secondary: give priority to processing the main surfaces of the parts.

[0103] Roughing before finishing: For each machining feature, the roughing operation should precede the finishing operation.

[0104] First the surface and then the hole: During the processing, the planes should be processed first, and then the holes attached to the planes should be processed. This can ensure the processing size and processing accuracy of the holes.

[0105] Among them, the process route planning goal is to optimize the process route with the highest processing efficiency, so that the number of machine tool replacements, processing clamping times and processing tool replacement times are minimized when processing a single part, and the cost function of machine tool replacement, processing clamping replacement cost function and processing tool replacement cost function are established;

[0106] (1) The calculation formula of the cost function of machine tool replacement is:

[0107] (5)

[0108] In formula (5), and Indicates two processing steps, and Respectively represent the processing steps and processing steps The machine tools used, Indicates the replacement of processing machine tools. It means that the processing machine tool has not been replaced;

[0109] (2) The calculation formula of the processing clamping replacement cost function is:

[0110] (6)

[0111] In formula (6), and Respectively represent the processing steps and processing steps The fixture used, and Respectively represent the processing steps and processing steps The clamping surface of the part, Indicates processing clamping replacement, Indicates that the processing fixture has not been changed;

[0112] (3) The calculation formula of the machining tool replacement cost function is:

[0113] (7)

[0114] In formula (7), and Respectively represent the processing steps and processing steps The processing tools used, Indicates the replacement of machining tools. Indicates that the machining tool has not been replaced;

[0115] (4) Calculate the weighted sum of the cost function of machine tool replacement, the cost function of fixture replacement, and the cost function of tool replacement as the final optimization objective function. The expression of the objective function is as follows:

[0116] + + (8)

[0117] In formula (8), , , They are the replacement cost of the processing machine tool, the replacement cost of the processing fixture, and the replacement cost of the processing tool, respectively. , , They are the weights of the replacement cost of the processing machine tool, the replacement cost of the processing fixture, and the replacement cost of the processing tool, respectively, and . According to the length of the replacement time, the weight is set.

[0118] S32. Follow the constraints of the basic process sequence of the part. Treat the machining step as a node, and use a graph to represent the constraints between machining steps to generate a machining step sequence constraint graph.

[0119] S33. Based on the objective function established in step S31 and the machining step sequence constraint graph generated in step S32, use the topological sorting idea to improve the ant colony algorithm, so as to find the optimal set of machining step sequences under the condition of meeting the constraints, and thus generate the process route of the part to be machined.

[0120] In this embodiment, the ant colony algorithm is improved using the topological sorting idea. As Figure 2 shown, let NC be the maximum number of iterations, nc be the current number of iterations, N be the number of ant colonies, n be the current ant serial number, P be the total number of machining step nodes, p be the current machining step serial number, use a taboo table to store the nodes visited by the ants, and obtain the final process route set. The specific process includes:

[0121] (1) Initialize the parameters of each part of the algorithm, set the number of ant colonies to N, the maximum number of iterations to NC, and the total number of nodes of the machining step to P. Obtain the nodes with an in-degree of 0 in the machining step sequence constraint graph, and randomly select a node from the nodes with an in-degree of 0 as the initial position of the ant path.

[0122] (2) Use the roulette wheel method to select the next node to visit, update the current machining step serial number p = p + 1, and judge p: if p < P, continue to execute step (2); otherwise, execute step (3).

[0123] (3) Update n = n + 1, and judge whether n < N holds: if it holds, execute step (2); if not, execute step (4).

[0124] (4) Calculate the optimization objective function value of the current path, and judge whether it is less than the optimal value: if it is less than the optimal value, update the optimization path; otherwise, directly jump to step (5).

[0125] (5) Judge the relationship between the current number of iterations nc and the maximum number of iterations NC: if nc < NC, the optimization ends and the optimal path is output; otherwise, jump to step (2) to continue execution.

[0126] Embodiment 2

[0127] The part process generation method based on the process knowledge base described in Example 1 is described below by taking a template-type part as an example, and the specific implementation is as follows.

[0128] As shown in Table 1, this is the processing feature set information of “template type parts”.

[0129] Table 1 “Template type parts” processing feature set information

[0130]

[0131] Implementation of step S1:

[0132] According to the enterprise and standard process data, the type parameters, size parameters and precision parameters of the part features are sorted out to establish a part feature processing chain library, as shown in Tables 2 and 3. Table 2 is the information (part) of the processing chain library to be selected for plane features, and Table 3 is the information (part) of the processing chain library to be selected for hole features.

[0133] Table 2 Information of the selected processing chain library of plane features (partial)

[0134]

[0135] Table 3 Information of the processing chain library to be selected for hole features (partial)

[0136]

[0137] Implementation of step S2:

[0138] Taking the hole feature with feature number "F02" in Table 1 as an example to perform multi-domain similarity calculation, it is necessary to calculate the similarity between the hole feature parameters with feature number "F02" and all the data in the processing process chain library, and then select the feature processing chain with higher similarity. Now, a feature processing chain in the processing chain library and the feature with feature number "F02" in Table 1 are used to perform similarity calculation demonstration, and the feature of the processing chain with serial number "327" is selected. The processing chain information with serial number "327" is shown in Table 4 below.

[0139] Table 4 Processing chain information with serial number "327"

[0140]

[0141] (1) Calculation of similarity of part feature types:

[0142] Since the feature number "F02" and the feature type "327" are both holes, their part feature type similarity is is 1.

[0143] (2) Calculation of similarity of part feature size parameters:

[0144] For hole features, the size parameters include diameter and depth, and min-max normalization is performed. The maximum diameter in the part processing chain library is 300mm and the minimum is 10mm, and the maximum depth is 400mm and the minimum is 10mm. The normalized matrix is:

[0145]

[0146] Substituting into the part feature similarity calculation formula, we can get:

[0147]

[0148] (3) Calculation of similarity of part feature accuracy parameters:

[0149] For the part feature accuracy parameter, set , the membership functions of the part features are:

[0150]

[0151]

[0152] in, and They represent the dimensional accuracy and roughness values ​​of part features respectively.

[0153] Substituting the dimensional accuracy and roughness of the part features into the above membership function, the fuzzy subset vector can be obtained:

[0154]

[0155] Calculate the closeness between the feature numbered "F02" and the feature numbered "327", and the similarity of the part feature accuracy parameters is:

[0156]

[0157]

[0158] in, =0.4651,

[0159] =0.0098.

[0160] (4) Calculation of comprehensive similarity of part features:

[0161] Pick , , The weight parameter is used to calculate the comprehensive similarity of part features:

[0162] + +

[0163] According to the above calculation process, the similarity of each feature of the part is calculated, and the calculation results are sorted by similarity to obtain the feature processing chain with the highest similarity, that is, the processing chain of part feature matching. As shown in Table 5, it is the process chain of each processing feature matching of template parts.

[0164] Table 5 Process chain matching of various processing features of template parts

[0165]

[0166] The process chain of each processing feature is disassembled to obtain processing steps, and the manufacturing resources such as processing machine tools, processing tools, processing clamps, etc. in the processing process are allocated to each processing step to obtain the process step sequence set of "template parts", as shown in Table 6.

[0167] Table 6 Process sequence set of “template parts”

[0168]

[0169] Implementation of step S3:

[0170] Set the step cost weight parameter to , , , where the ant colony population size N is set to 100 and the maximum number of iterations NC is set to 500.

[0171] With the minimum step cost as the goal, the step sequence is optimized by using the optimization algorithm, and the process is grouped by process to obtain the final processing procedures of the template parts as shown in Table 7, that is, the process route of the template parts described in this embodiment is generated.

[0172] Table 7 Final processing steps of template parts

[0173]

[0174] Example 3

[0175] This embodiment discloses a part process generation system based on a process knowledge base, which is used in the part process generation method based on a process knowledge base described in Embodiment 1, and includes:

[0176] A data screening module is configured to connect to a process knowledge base through the system software of a terminal, and screen and identify relevant process data in the process knowledge base, wherein the relevant process data at least includes processing process chain library data, processing tool library data, processing machine tool equipment library data, processing fixture library data, and measuring tool library data;

[0177] A processing feature process chain generation module is configured to obtain a processing feature process chain of a part feature to be processed by multi-domain similarity calculation and matching according to relevant process data in a process knowledge base after screening and identification, and generate a process step sequence set;

[0178] The process route generation module is configured to optimize the process step sequence set according to the process step sequence set through an optimization algorithm to generate a process route for the part to be processed.

[0179] In the above data screening module, processing feature process chain generation module and process route generation module, the implementation process of the functions and effects of each module is specifically described in Example 1 regarding the implementation process of the corresponding steps in the part process generation method based on the process knowledge base, which will not be repeated here.

[0180] As for the embodiment of the part process generation system of the process knowledge base, since it basically corresponds to the method embodiment, the relevant parts may refer to the partial description of the method embodiment.

[0181] The above only describes the basic principles and preferred embodiments of the present invention. Those skilled in the art may make many changes and improvements based on the above description, and these changes and improvements should fall within the protection scope of the present invention.

Claims

1. A method for generating a part process based on a process knowledge base, characterized in that: The steps include: S1. Connecting to a process knowledge base through the system software of the terminal, and identifying and screening relevant process data in the process knowledge base, wherein the relevant process data at least includes processing process chain library data, processing tool library data, processing machine tool equipment library data, processing fixture library data, and measuring tool library data; S2, according to the relevant process data in the process knowledge base screened and identified in step S1, obtain the processing feature process chain of the feature of the part to be processed by multi-domain similarity calculation and matching, and generate a process step sequence set; S3. According to the process step sequence set obtained in step S2, the process step sequence set is optimized by using an optimization algorithm to generate a process route for the part to be processed.

2. The method for generating part process based on process knowledge base according to claim 1, characterized in that: In step S1, the data included in the process knowledge base include at least part entity data, processing knowledge data, and manufacturing resource data; wherein, the part entity data includes at least part process data, part feature processing chain, and part processing process route, the processing knowledge data includes at least processing feature process chain, work step matching manufacturing resources, and work step constraints, and the manufacturing resource data includes at least processing machine tool equipment information, processing tool information, and processing fixture information.

3. The method for generating part process based on process knowledge base according to claim 1, characterized in that: Step S1 specifically includes: S11. The terminal establishes a connection with the process knowledge base through an Internet API interface to obtain relevant process data stored in the process knowledge base; S12, by setting process rules, classify and mark the relevant process data obtained in step S11 according to the type and attribute of the data, and filter the classified data to delete redundant information and erroneous information; S13. Use structured rules to map the unstructured data in the process knowledge base into structured data, and combine them to generate a structured data matrix.

4. The method for generating part process based on process knowledge base according to claim 1, characterized in that: Step S2 specifically includes: S21, encoding the features of the part to be processed according to the rules, then structuring the unstructured data in the part to be processed, and finally converting the feature data of the part to be processed into a part feature matrix; S22, multi-domain similarity calculation of part features, based on the commonly used processing process chain library of processing features, obtain the processing feature process chain of each feature of the part to be processed; S23. Generate a set of part process step sequences, and match each process step with corresponding manufacturing resources based on the process chain of processing features, wherein the process steps at least include processing methods, processing features, processing machines, processing tools and processing fixtures, and the manufacturing resources at least include processing machines, processing tools and processing resources.

5. The method for generating part process based on process knowledge base according to claim 4, characterized in that: In step S22, the multi-domain similarity calculation step of part features includes: Firstly, the similarity calculation of part feature type, part feature size parameter and part feature accuracy parameter are performed respectively; Then, the weights of the part feature type, part feature size parameter and part feature accuracy parameter are determined respectively, so as to calculate the comprehensive similarity of the part features.

6. The method for generating part process based on process knowledge base according to claim 5, characterized in that: The multi-domain similarity calculation of part features is as follows: (1) Calculation of similarity of part feature types: A classification tree is established based on the part feature types. The nodes represent the part feature types, and the edges represent the relationship between the part feature types. The distance between two nodes represents the similarity of the part feature types. The specific calculation formula is as follows: （1） In formula (1), Indicates the similarity of feature types, and For two given nodes, Represents the root node, Indicates the total number of nodes, express and The nearest common ancestor node, represents the depth of the nearest common ancestor node, is the edge weight between nodes; (2) Calculation of similarity of part feature size parameters: For features of the same feature type, similarity is calculated and the Euclidean distance is used to characterize the similarity. The specific calculation formula is as follows: （2） In formula (2), Indicates the similarity of part feature size parameters, and Representing two features The value of the size parameter, The number of size parameters representing the feature; (3) Calculation of similarity of part feature accuracy parameters: A similarity calculation method based on fuzzy mathematics is used to characterize the similarity using the set closeness in fuzzy mathematics. The specific calculation formula is as follows: （3） In formula (3), Indicates the similarity of part feature accuracy parameters, and are respectively the precision fuzzy feature subsets, represents the inner product operation of the precision fuzzy feature subset, The outer product calculation of the fuzzy feature subset representing the accuracy; (4) Calculation of comprehensive similarity of part features: Determine the weights of part feature type, part feature size parameters, and part feature accuracy parameters. The calculation formula for the comprehensive similarity of part features is as follows: + + （4） In formula (4), is the comprehensive similarity of part features, They are respectively the weight of the similarity of part feature type, the weight of the similarity of part feature size parameters, and the weight of the similarity of part feature accuracy parameters.

7. The method for generating part process based on process knowledge base according to claim 2, characterized in that: Step S3 specifically includes: S31, determining the process route planning objective function according to the process rules and manufacturing resource information, wherein the process rules at least include datum first, primary first then secondary, rough first then fine, and surface first then hole; S32, following the basic process sequence constraints of the parts, treating the process steps as nodes, using a graph to characterize the constraints between the process steps, and generating a process step sequence constraint graph; S33. Based on the objective function established in step S31 and the process sequence constraint diagram generated in step S32, the ant colony algorithm is improved using the topological sorting idea, so that the optimal process step sequence set can be found under the condition of satisfying the constraints, thereby generating a process route for the part to be processed.

8. The method for generating part process based on process knowledge base according to claim 7, characterized in that: In step S31, the process route planning goal is to optimize the process route with the highest processing efficiency, so that the number of machine tool replacements, processing clamping times and processing tool replacement times are minimized when processing a single part, and then a processing machine tool replacement cost function, a processing clamping replacement cost function and a processing tool replacement cost function are established; (1) The calculation formula of the cost function of machine tool replacement is: （5） In formula (5), and Indicates two processing steps, and Respectively represent the processing steps and processing steps The machine tools used; (2) The calculation formula of the processing clamping replacement cost function is: （6） In formula (6), and Respectively represent the processing steps and processing steps The fixture used, and Respectively represent the processing steps and processing steps The clamping surface of the part; (3) The calculation formula of the machining tool replacement cost function is: （7） In formula (7), and Respectively represent the processing steps and processing steps The machining tools used; (4) Calculate the weighted sum of the cost function of machine tool replacement, the cost function of fixture replacement, and the cost function of tool replacement as the final optimization objective function. The expression of the objective function is as follows: + + （8） In formula (8), , , They are the cost of replacing the processing machine tool, the cost of replacing the processing fixture, and the cost of replacing the processing tool. , , are the weight of the cost of replacing the machining machine tool, the weight of the cost of replacing the machining fixture, and the weight of the cost of replacing the machining tool, respectively. , according to the length of replacement time, set the weight .

9. The method for generating part process based on process knowledge base according to claim 7, characterized in that: In step S33, the ant colony algorithm is improved by using the topological sorting idea, which specifically includes: (1) When the ant selects the initial node, it obtains the nodes with in-degree 0 in the work step sequence constraint graph, randomly selects a node from the obtained nodes with in-degree 0 as the initial node of the ant path, and deletes the node from the work step sequence constraint graph; (2) Each time the next node to be visited is selected for the ant, the nodes with in-degree 0 in the remaining nodes of the work step sequence constraint graph are obtained. For these nodes, a roulette wheel method is used to select a node as the next node visited by the ant, and the node is deleted from the work step sequence constraint graph; (3) Repeat step (2) until all nodes have been visited.

10. A part process generation system based on a process knowledge base, used to implement the part process generation method based on a process knowledge base according to any one of claims 1 to 9, characterized in that: include: A data screening module is configured to connect to a process knowledge base through the system software of a terminal, and screen and identify relevant process data in the process knowledge base, wherein the relevant process data at least includes processing process chain library data, processing tool library data, processing machine tool equipment library data, processing fixture library data, and measuring tool library data; A processing feature process chain generation module is configured to obtain a processing feature process chain of a part feature to be processed by multi-domain similarity calculation and matching according to relevant process data in a process knowledge base after screening and identification, and generate a process step sequence set; The process route generation module is configured to optimize the process step sequence set according to the process step sequence set through an optimization algorithm to generate a process route for the part to be processed.

Images