A selective laser melting intelligent process design and simulation optimization platform and method

Abstract

The application relates to a laser selective melting intelligent process design and simulation optimization platform and method, and belongs to the technical field of laser selective melting processing, and solves the problem that the prior art cannot meet the requirements of a new generation of laser selective melting process and rapid response, high efficiency and high quality manufacturing. The platform comprises a three-dimensional model construction module, a model feature identification module, a process database, a knowledge base, a search similar laser selective melting process parameter of a target part to obtain laser selective melting initial recommended process parameters of the target part, a simulation interaction module, a three-dimensional model is imported into finite element analysis software, the finite element analysis software is driven to calculate and analyze under different parameter iterations, and calculation results are extracted for quality result data analysis. Laser selective melting process parameters can be quickly obtained without relying on manual experience trial and error and manual simulation analysis, and the digitization and intelligent level of process design is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of laser selective melting, in particular to a laser selective melting intelligent process design and simulation optimization platform and method. BACKGROUND

[0002] At present, the manufacturing process informationization and intelligentization capability of laser selective melting production process is seriously insufficient, and the process design still relies on manual experience trial and error and manual setting simulation analysis process. In order to meet the design index requirements, trial and error iteration needs to be repeated, which leads to problems such as high cost, long cycle and low quality reliability in product development, and it is difficult to meet the demand of rapid response, high efficiency and high quality of new generation laser selective melting process and manufacturing. The digitalization and intelligentization level of process design needs to be improved. SUMMARY

[0003] In view of the above analysis, the embodiments of the present application aim to provide a laser selective melting intelligent process design and simulation optimization platform and method to solve the problem that the existing process cannot meet the demand of rapid response, high efficiency and high quality of new generation laser selective melting process and manufacturing.

[0004] In one aspect, the embodiments of the present application provide a laser selective melting intelligent process design and simulation optimization platform, comprising:

[0005] A three-dimensional model construction module is used to simplify the forming target part and support structure, and save it as XT format to construct the three-dimensional model of the target part and tooling;

[0006] A model feature recognition module is used to compare the straight lines, fillets, curved surfaces, holes, planes, stretched and rotated entities and feature relationships in the recognized three-dimensional model with the feature parameters in the process database to determine the to-be-processed features of the target part;

[0007] A process database is used to store the basic information of laser selective melting process;

[0008] A knowledge base is used to obtain the laser selective melting process of the target part based on the process characteristics of the target part; obtain the technical requirements of the to-be-processed surface accuracy and forming quality of the target part based on the laser selective melting process of the target part and the to-be-processed surface features of the target part; search for similar laser selective melting process parameters of the target part in the process database based on the to-be-processed surface features of the target part and the technical requirements of the to-be-processed surface accuracy and forming quality of the target part to obtain the initial recommended process parameters of the laser selective melting of the target part;

[0009] The simulation interaction module is used for importing the built three-dimensional model into a finite element analysis software, driving the finite element analysis software to perform calculation and analysis under different parameter iterations, and extracting calculation results for quality result data analysis.

[0010] The finite element analysis module performs finite element analysis on the built three-dimensional model and laser selective melting initial recommended process parameters of the target part to obtain target part forming surface and stress-strain distribution-time data.

[0011] Based on further improvement of the above platform, the laser selective melting process basic information stored in the process database includes: laser selective melting process based on laser selective melting process characteristics of the target part, laser selective melting process parameters based on technical requirements of the precision and forming quality of the surface to be processed of the target part based on the characteristics of the surface to be processed of the target part, and technical requirements of the precision and forming quality of the surface to be processed based on the characteristics of the surface to be processed of the target part, and feature parameter data of straight lines, fillets, curved surfaces, holes, planes, stretching and rotating entities and feature relationships.

[0012] Based on further improvement of the above platform, the knowledge base is provided with a model-process-quality relationship prediction model and a rule base to search and calculate laser selective melting initial recommended process parameters.

[0013] Based on further improvement of the above platform, the model-process-quality relationship prediction model includes: an input layer, a hidden layer and an output layer.

[0014] The input layer satisfies:

[0015]

[0016] The hidden layer satisfies:

[0017]

[0018] The output layer satisfies:

[0019]

[0020] The number of nodes of the input layer, the hidden layer and the output layer is m0, m1 and m2 respectively.

[0021] Y (0) represents an input layer process parameter condition matrix.

[0022] Y (1) represents an output vector of the hidden layer.

[0023] Y (2) represents a mathematical relationship between model quality and process.

[0024] Based on the further improvement of the above platform, the hidden layer satisfies:

[0025]

[0026] Wherein, f (1) is an activation function, and for a laser selective melting process, the following is true:

[0027] f(x)=max(xα,x)

[0028] Wherein, alpha is a hyperparameter, obtained by data training, net (1) is a weighted quantity.

[0029] Based on the further improvement of the above platform, the rule base is the boundary condition of the process, used to build the process value range of laser selective melting, the quality requirement of initial input and the material fracture strain value.

[0030] Based on the further improvement of the above platform, the knowledge base is provided with a simulated annealing optimization model to optimize the process parameter data that does not meet the requirements;

[0031] The simulated annealing optimization model is used to optimize and modify the support structure position and size, cladding path, initial starting point, environmental temperature, laser power, welding gun diameter and cladding speed.

[0032] In addition, the embodiment of the present application also provides a laser selective melting intelligent process design and simulation optimization method, comprising obtaining the optimal laser selective melting process parameters of the target part by using the above laser selective melting intelligent process design and simulation optimization platform.

[0033] Based on the further improvement of the above method, comprising: step 1: using a three-dimensional model construction module to simplify the forming target part and the support structure, and saving it as XT format to construct a three-dimensional model of the target part and the tooling;

[0034] Step 2: using a model feature recognition module to recognize the straight line, fillet, arc surface, hole, plane, stretching and rotating entity and the feature relationship in the three-dimensional model, to obtain the surface feature to be processed of the target part;

[0035] Step 3: based on the surface feature to be processed of the target part, using the knowledge base to obtain the laser selective melting initial recommended process parameters of the target part from the process database;

[0036] Step 4: passing the obtained laser selective melting initial recommended process parameters and the three-dimensional model to the finite element analysis software through the simulation interaction module, to perform finite element analysis on the laser selective melting initial process parameters, to obtain the target part forming surface and stress-strain distribution-time data;

[0037] Step 5: deliver all limit element analysis results to the knowledge base through the simulation interaction module to compare and analyze the target part forming surface and stress-strain distribution-time data with the technical requirements in the process database through the knowledge base;

[0038] Step 6: use the simulated annealing algorithm to optimize the laser selective melting process parameter data that does not meet the requirements, and return to step 4 until the laser selective melting process parameters that meet the requirements are obtained.

[0039] Further improvement based on the above method also includes step 7: saving the process of obtaining the final result and the final result to the process database, and optimizing the model-process-quality relationship prediction model and the simulated annealing optimization model and the related rule base and the simulated annealing algorithm based on the updated tool database.

[0040] Compared with the prior art, the present application can at least achieve the following beneficial effects:

[0041] The present application automatically obtains the surface features to be processed of the three-dimensional model of the target part and the support structure through the model feature recognition module, and based on the surface features to be processed and the technical requirements, the initial recommended laser selective melting process parameters are obtained through automatic searching in the process database based on the knowledge base, and the initial laser selective melting process parameters are analyzed by finite element analysis, so as to realize the optimization of the laser selective melting process parameters, until the laser selective melting process parameters that meet the requirements are obtained, so as to improve the efficiency of obtaining the laser selective melting process parameters under the condition of ensuring the quality of process design, overcome the problem that the process design is too dependent on manual experience and trial and error and manual simulation analysis, and improve the efficiency of simulation iteration.

[0042] In the present application, the above technical solutions can also be combined with each other to realize more preferred combination solutions. Other features and advantages of the present application will be described in the subsequent specification, and some advantages will become apparent from the specification, or will be understood by implementing the present application. The purpose and other advantages of the present application can be realized and obtained from the contents specifically pointed out in the specification and the drawings. BRIEF DESCRIPTION OF DRAWINGS

[0043] The accompanying drawings are included to provide a further understanding of the application and are incorporated herein and constitute a part of the detailed description. The drawings illustrate embodiments of the application and, together with the description, serve to explain the principles of the application. In the drawings:

[0044] Figure 1 It is a structural block diagram of the intelligent process design and simulation optimization platform for laser selective melting in the present application;

[0045] Figure 2 It is a flow chart of the intelligent process design and simulation optimization method for laser selective melting in the present application;

[0046] Figure 3 The logical diagram of the searching and calculating method of the knowledge base in the present application. DETAILED DESCRIPTION

[0047] The preferred embodiments of the present application will be described in detail below with reference to the drawings, wherein the drawings constitute a part of this application and serve to explain the principles of the embodiments of the present application, but are not used to limit the scope of the present application.

[0048] The additive structure parts in aerospace and automobile manufacturing industry have the significant features of complex shape, high precision, and structure and function integration. With the continuous improvement of the technical requirements of laser selective melting manufacturing process, the laser selective melting target part manufacturing process presents the characteristics of high dispersion, multi-variety, small batch, and short development cycle. The control requirements for the size precision of the components are more and more strict, and the quality and efficiency of the process design directly affect the overall machining quality. Therefore, the high-quality and high-efficiency design of the machining process is extremely critical. However, at present, the informatization and intelligentization ability of the laser selective melting production process is seriously insufficient, and the process design still depends on manual experience trial and error and manual setting simulation analysis process. In order to meet the design index requirements, trial and error iteration needs to be repeatedly performed, which leads to the problems of high cost, long cycle, and low quality reliability in product development, and it is difficult to meet the demand of rapid response, high efficiency, and high quality of the new generation of laser selective melting process and manufacturing. The digitalization and intelligentization level of the process design needs to be improved.

[0049] To solve the above problems, the present application provides a laser selective melting intelligent process design and simulation optimization platform, comprising:

[0050] A three-dimensional model construction module is used to construct the three-dimensional model of the shaped target part and the support structure;

[0051] A model feature recognition module is used to compare the straight lines, fillets, curved surfaces, holes, planes, stretched and rotated entities and the feature relationships of each feature in the recognized three-dimensional model with the feature parameters in the process database, so as to determine the to-be-processed features of the target part;

[0052] A process database is used to store the basic information of the laser selective melting process;

[0053] A knowledge base is used to obtain the laser selective melting process of the target part based on the laser selective melting process characteristics of the target part, obtain the technical requirements of the to-be-processed surface precision and forming quality of the target part based on the laser selective melting process of the target part and the to-be-processed surface features of the target part, and search for and obtain the similar laser selective melting process parameters of the target part in the process database based on the to-be-processed surface features of the target part and the technical requirements of the to-be-processed surface precision and forming quality of the target part, so as to obtain the initial recommended process parameters of the laser selective melting of the target part;

[0054] An emulation interaction module is used to import the built three-dimensional model into finite element analysis software, drive the finite element analysis software to perform calculation and analysis under different parameter iterations, and meanwhile extract the calculation results and transmit them to the knowledge base for quality result data analysis.

[0055] The laser selective melting process based on the laser selective melting process characteristics of the target part, the laser selective melting process parameters based on the technical requirements of the machining surface precision and forming quality of the machining surface features of the target part, and the technical requirements of the machining surface precision and forming quality of the machining surface features of the target part further include feature parameter data of straight lines, fillets, curved surfaces, holes, planes, stretching and rotating entities and feature relationships.

[0056] That is, by recognizing the built three-dimensional model, the straight lines, fillets, curved surfaces, holes, planes, stretching and rotating entities and feature relationships of the target part are obtained, and the similar laser selective melting process parameters of the target part are searched in the process data by combining the technical requirements of the machining surface precision and forming quality.

[0057] The built three-dimensional model and the laser selective melting initial recommended process parameters are subjected to finite element analysis to obtain the target part forming surface and stress-strain distribution-time data, and the forming surface and stress-strain distribution-time data are compared with the technical requirements in the process library.

[0058] Compared with the prior art, the model feature recognition module automatically obtains the machining surface features of the three-dimensional model of the target part and the support structure, and based on the machining surface features and the technical requirements, the laser selective melting initial recommended process parameters are automatically searched in the process database through the knowledge base, and the laser selective melting initial process parameters are subjected to finite element analysis to realize the optimization of the laser selective melting process parameters until the laser selective melting process parameters meeting the requirements are obtained.

[0059] Specifically, the three-dimensional model construction module is used to simplify the forming target part and the support structure in the three-dimensional model software and save it in XT format to construct the three-dimensional model of the target part and the tooling.

[0060] The model feature recognition module is configured to compare the identified straight lines, rounded corners, curved surfaces, holes, planes, stretched and rotated entities and feature relationships in the three-dimensional model with feature parameters in the process database to determine the features to be processed of the target part.

[0061] The process database is configured to store basic information of the laser selective melting process input by the business terminal automatically and by the process personnel.

[0062] Exemplarily, the basic information of the laser selective melting process includes an ambient temperature, a laser power, a welding gun diameter and a cladding speed.

[0063] The knowledge base is configured to acquire the laser selective melting process of the target part based on the characteristics of the laser selective melting process of the target part, acquire technical requirements of the surface to be processed of the target part based on the laser selective melting process of the target part and the features of the surface to be processed of the target part, and search for similar laser selective melting process parameters of the target part in the process database based on the features of the surface to be processed of the target part and the technical requirements of the surface to be processed of the target part to obtain initial recommended process parameters of the laser selective melting of the target part.

[0064] The model-process-quality relationship prediction model and the simulated annealing optimization model and the related rule base are provided in the knowledge base to search for and calculate the initial recommended process parameters of the laser selective melting.

[0065] Specifically, the model-process-quality relationship prediction model is configured to set a three-layer neural network including an input layer, a hidden layer and an output layer.

[0066] The number of nodes of the input layer, the hidden layer and the output layer is m0, m1 and m2 respectively.

[0067] The input layer satisfies:

[0068]

[0069] The hidden layer satisfies:

[0070]

[0071] The output layer satisfies:

[0072]

[0073] For the hidden layer:

[0074]

[0075] Y (0) represents an input layer process parameter condition matrix.

[0076] Y(1) an output vector representing the hidden layer;

[0077] Y (2) a mathematical relationship representing the model quality and process.

[0078] wherein f (1) is an activation function, and for the laser selective melting process, f(x) = max(xa, x).

[0079] f(x) = max(xa, x)

[0080] wherein a is a hyperparameter obtained by data training, net (1) is a weighted quantity.

[0081] The rule base is a boundary condition of the process, used to construct the process value range of laser selective melting, the quality requirement of initial input, and the material fracture strain value.

[0082] Specifically, the search and calculation method is: after determining the model characteristics and quality requirements, the knowledge base queries the saved model-process-quality relationship to see if it matches the known facts.

[0083] After all IF clauses (relationships, i.e. hidden layers) in the rule match, backtrack to the THEN clause, enable the rule, and add the new facts generated by the rule to the database. At this time, all the facts in the database are the reasoning results, and the reasoning ends.

[0084] wherein the laser selective melting initial recommended process parameters include: ambient temperature, laser power, welding gun diameter, and cladding speed.

[0085] The simulation interaction module is used to import the constructed three-dimensional model into the finite element analysis software, drive the finite element analysis software to perform calculation and analysis under different parameter iterations, and extract the calculation results for quality result data analysis.

[0086] Specifically, the finite element analysis software is used to perform finite element analysis on the constructed three-dimensional model and the laser selective melting initial recommended process parameters of the target part, to obtain the target part forming surface and stress-strain distribution-time data.

[0087] wherein after obtaining the target part forming surface and stress-strain distribution-time data, the knowledge base is used to compare and analyze the obtained data with the technical requirements in the process database.

[0088] wherein if the comparison meets the requirements, the search obtains the similar laser selective melting process parameters of the target part as the final result.

[0089] If the comparison does not meet the requirements, the knowledge base is used to optimize the laser selective melting process until the process parameters meeting the requirements are obtained.

[0090] Among them, the simulated annealing optimization model is set in the knowledge base to optimize the process parameter data that does not meet the requirements until the process parameter meeting the requirements is obtained.

[0091] Among them, the simulated annealing algorithm is used to optimize and modify the support structure position and size, cladding path, initial starting point, environment temperature, laser power, welding gun diameter and cladding speed, and the simulation software is driven again for analysis until the result meeting the requirements is obtained.

[0092] Specifically, the probability p that the process parameter is accepted from x1 to x2 satisfies:

[0093]

[0094] Wherein, △f=f1-f2 is the difference between the values of the objective function f(x1), f(x2);

[0095] f(x) is a quality function, representing the maximum strain.

[0096] T is a temperature constant;

[0097] Slowly decrease with time, when the temperature is very high, take the second probability high, with the temperature decreasing, take the second kind of situation slowly reduces, until almost no second kind of situation is selected.

[0098] Wherein, the relationship between the iteration number n and the temperature constant T satisfies:

[0099] T(n+1)=K*T(n)

[0100] Wherein, K is a constant less than 1.

[0101] Therefore, the simulated annealing optimization model is used to optimize the laser selective melting process parameters until the laser selective melting process parameters meeting the requirements are obtained, finally, the process and the final result of obtaining the final result are saved to the process database, and the model-process-quality relationship prediction model and the simulated annealing optimization model and the related rule base and the simulated annealing algorithm are optimized based on the updated tool database.

[0102] Wherein, after obtaining the final result, the process and the final result of obtaining the final result are saved to the process database and the model-process-quality relationship prediction model and the simulated annealing optimization model and the related rule base and the simulated annealing algorithm are optimized based on the updated tool database.

[0103] In addition, the present application provides a laser selective melting intelligent process design and simulation optimization method, comprising:

[0104] Step 1: Simplify the forming target part and support structure by using a three-dimensional model building module, save it in XT format, and build a three-dimensional model of the target part and tooling to obtain the target part and tooling;

[0105] Step 2: Identify the straight lines, fillets, curved surfaces, holes, planes, stretched and rotated entities and their feature relationships in the three-dimensional model using a model feature recognition module to obtain the surface features to be machined of the target part;

[0106] Step 3: Based on the surface features to be machined of the target part, use the knowledge base to obtain the initial recommended process parameters of laser selective melting of the target part from the process database;

[0107] Step 4: Transfer the obtained laser selective melting initial recommended process parameters and three-dimensional model to the finite element analysis software through the simulation interaction module to perform finite element analysis on the laser selective melting initial process parameters to obtain the target part forming surface and stress-strain distribution-time data;

[0108] Step 5: Transfer all finite element analysis results to the knowledge base through the simulation interaction module to compare and analyze the target part forming surface and stress-strain distribution-time data with the technical requirements in the process database through the knowledge base;

[0109] Step 6: Optimize the laser selective melting process parameter data that does not meet the requirements using the simulated annealing algorithm and return to step 4 until the laser selective melting process parameters that meet the requirements are obtained.

[0110] Specifically, in step 1, a three-dimensional model of the forming target part and support structure is built using a three-dimensional model software.

[0111] Specifically, in step 2, the surface features to be machined of the three-dimensional model identified by the model feature recognition module are compared with the parameters in the process database to determine the surface features to be machined of the target part.

[0112] Among them, the feature parameters in the process database include: straight lines, fillets, curved surfaces, holes, planes, stretched and rotated entities and their feature relationship data.

[0113] Specifically, in step 3, it includes:

[0114] S31: Based on the laser selective melting process characteristics of the target part, obtain the laser selective melting process of the target part from the process knowledge base;

[0115] Specifically, specify the target part and support structure, select the cladding path and starting position, and select the target part material, including titanium alloy and aluminum alloy;

[0116] Wherein, the path and the starting position are determined by the coplanar collinear identified by the model features, and the starting and ending points of each path are determined by the shortest path, and the starting and ending points have no effect on a single path.

[0117] Wherein, the material of the target part is selected according to the actual processing part material type.

[0118] S32: Based on the laser selective melting process of the target part and the features of the surface to be processed of the target part, the technical requirements of the surface to be processed of the target part are obtained from the process database.

[0119] Wherein, the technical requirements of the surface to be processed of the target part are standard parameters stored in the process database.

[0120] S33: Based on the features of the surface to be processed of the target part and the technical requirements of the surface to be processed of the target part, the search similar laser selective melting process parameters of the target part are searched and obtained from the process database, so as to obtain the initial recommended process parameters of the laser selective melting of the target part.

[0121] Specifically, according to the model-process-quality relationship prediction model and the related rule base in the knowledge base, the initial recommended process parameters are searched and calculated;

[0122] Wherein, a three-layer neural network is set, including an input layer, a hidden layer and an output layer.

[0123] Wherein, the number of nodes of the input layer, the hidden layer and the output layer are m0, m1 and m2, respectively.

[0124] Wherein, the input layer satisfies:

[0125]

[0126] The hidden layer satisfies:

[0127]

[0128] The output layer satisfies:

[0129]

[0130] For the hidden layer:

[0131]

[0132] Wherein, Y (0) represents the input layer process parameter condition matrix.

[0133] Y (1) represents the output vector of the hidden layer.

[0134] Y(2) The mathematical relationship between the model quality and the process is represented.

[0135] wherein f (1) is an activation function, and for the laser selective melting process, it has:

[0136] f(x) = max(xa, x)

[0137] wherein a is a hyperparameter obtained by data training, net (1) is a weighted quantity.

[0138] The rule base is the boundary condition of the process, which is used to construct the process value range of the laser selective melting, the quality requirement of the initial input, and the material fracture strain value.

[0139] Specifically, the searching and calculating method is as follows: after the model characteristics and quality requirements are determined, the knowledge base queries the saved model-process-quality relationship to see whether it matches the known facts.

[0140] After all the IF clauses (relationships, i.e., hidden layers) in the rule match, the THEN clause is traced back to enable the rule, and the new facts generated by the rule are added to the database. At this time, all the facts in the database are the reasoning results, and the reasoning ends.

[0141] The initial recommended process parameters of the laser selective melting include: ambient temperature, laser power, welding gun diameter, and cladding speed.

[0142] Specifically, in step 5, if the quality technical requirements are met, the comparison is successful, and the final process parameters are obtained; if the quality technical requirements are not met, the comparison is unsuccessful.

[0143] Specifically, in step 6, the simulated annealing algorithm is used to optimize the laser selective melting process parameter data that does not meet the requirements, and the process returns to step 4 until the process parameters that meet the requirements are obtained.

[0144] Specifically, the probability p that the process parameter changes from x1 to x2 is accepted satisfies:

[0145]

[0146] wherein Af = f1-f2, which is the difference between the values of the objective functions f(x1), f(x2);

[0147] f(x) is a quality function, which represents the maximum strain.

[0148] T is a temperature constant.

[0149] As time slowly reduces, when the temperature is high, the second probability is high, and as the temperature decreases, the second case slowly reduces, until almost no second case is selected.

[0150] Wherein the relationship between the number of iterations n and the temperature constant T satisfies:

[0151] T(n+1) = K*T(n)

[0152] Wherein K is a constant less than 1.

[0153] Wherein the process parameters are laser power, scanning speed, scanning spacing, powder thickness.

[0154] Wherein the optimization modifies the support structure position and size, cladding path, initial starting point, ambient temperature, laser power, welding gun diameter and cladding speed, and drives the simulation software to analyze again until the required result is obtained.

[0155] Further, it further includes step 7: save the process of obtaining the final result and the final result to the process database, and based on the updated tool database, optimize the model-process-quality relationship prediction model and simulated annealing optimization model and related rule library and simulated annealing algorithm.

[0156] Compared with the prior art, the model feature recognition module of the present application automatically obtains the machined surface features of the three-dimensional model of the target part and the support structure, and based on the machined surface features and technical requirements, the initial recommended process parameters of laser selective melting are obtained by searching in the process database through the knowledge base, and the initial process parameters of laser selective melting are analyzed by finite element analysis, so as to realize the optimization of laser selective melting process parameters, until the laser selective melting process parameters meeting the requirements are obtained. In this way, under the condition of ensuring the quality of process design, the efficiency of obtaining laser selective melting process parameters is improved, the problem of over-reliance on manual experience trial and error and manual simulation analysis in process design is overcome, and the efficiency of simulation iteration is improved.

[0157] Example 1

[0158] A laser selective melting intelligent process design and simulation optimization platform comprises:

[0159] A three-dimensional model construction module is used to construct a three-dimensional model of a formed target part and a support structure.

[0160] Wherein, the formed target part and the support structure are simplified in the three-dimensional model software and saved as XT format to construct the three-dimensional model of the target part and the tooling.

[0161] A model feature recognition module is configured to compare the identified straight lines, rounded corners, curved surfaces, holes, planes, stretched and rotated entities and feature relationships in the three-dimensional model with feature parameters in the process database to determine the features to be processed of the target part.

[0162] A process database is configured to store laser selective melting process basic information input by the business terminal automatically iterated and process personnel;

[0163] The laser selective melting process basic information includes laser selective melting process based on the laser selective melting process characteristics of the target part, laser selective melting process parameters based on the technical requirements of the surface precision and forming quality of the surface to be processed of the target part based on the surface feature to be processed of the target part, and the technical requirements of the surface precision and forming quality of the surface to be processed of the target part, and the feature parameter data of straight lines, rounded corners, curved surfaces, holes, planes, stretched and rotated entities and feature relationships.

[0164] Exemplarily, the ambient temperature, laser power, welding gun diameter and cladding speed.

[0165] A knowledge base is configured to obtain the laser selective melting process of the target part based on the laser selective melting process characteristics of the target part, obtain the technical requirements of the surface precision and forming quality of the target part based on the laser selective melting process of the target part and the surface feature to be processed of the target part, and search for similar laser selective melting process parameters of the target part in the process database based on the surface feature to be processed of the target part and the technical requirements of the surface precision and forming quality of the target part to obtain the initial recommended process parameters of the laser selective melting of the target part.

[0166] The model-process-quality relationship prediction model and the simulated annealing optimization model and the related rule base are provided in the knowledge base to search and calculate the initial recommended process parameters of the laser selective melting.

[0167] Specifically, the model-process-quality relationship prediction model is configured to set a 3-layer neural network including an input layer, a hidden layer and an output layer.

[0168] The number of nodes of the input layer, the hidden layer and the output layer is m0, m1 and m2 respectively.

[0169] The input layer satisfies:

[0170]

[0171] The hidden layer satisfies:

[0172]

[0173] The output layer satisfies:

[0174]

[0175] wherein Y (0) represents the input layer process parameter condition matrix;

[0176] Y (1) represents the output vector of the hidden layer;

[0177] Y (2) represents the mathematical relationship between the model quality and the process.

[0178] For the hidden layer:

[0179]

[0180] wherein f (1) is an activation function, and for the laser selective melting process, f(x) = max(xa, x).

[0181] f(x) = max(xa, x)

[0182] wherein a is a hyperparameter obtained by data training, net (1) is a weighted quantity.

[0183] The rule base is the boundary condition of the process, which is used to construct the process value range of laser selective melting, the initial input quality requirement, and the material fracture strain value.

[0184] Specifically, the searching and calculating method is as follows: after the model characteristics and quality requirements are determined, the knowledge base queries the saved model-process-quality relationship to see whether it matches the known facts.

[0185] After all the IF clauses (relationships, i.e., hidden layers) in the rule match, the THEN clause is traced back to enable the rule, and the new facts generated by the rule are added to the database. At this time, all the facts in the database are the reasoning results, and the reasoning ends.

[0186] wherein the laser selective melting initial recommended process parameters include: ambient temperature, laser power, welding gun diameter, and cladding speed.

[0187] The simulation interaction module is used to import the constructed three-dimensional model into the finite element analysis software, drive the finite element analysis software to perform calculation and analysis under different parameter iterations, and extract the calculation results for quality result data analysis.

[0188] The finite element analysis module performs finite element analysis on the constructed three-dimensional model and the laser selective melting initial recommended process parameters of the target part to obtain the target part forming surface and stress-strain distribution-time data;

[0189] Wherein, after obtaining the target part forming surface and stress-strain distribution-time data, the knowledge base is utilized and the obtained data is compared and analyzed with the technical requirements in the process database;

[0190] Wherein, if the comparison meets the requirements, the search for the similar laser selective melting process parameters of the target part is output as the final result.

[0191] If the comparison does not meet the requirements, the knowledge base is utilized to optimize the laser selective melting process until the process parameters meeting the requirements are obtained.

[0192] Wherein, an annealing simulation optimization model is provided in the knowledge base to optimize the process parameter data not meeting the requirements until the process parameters meeting the requirements are obtained.

[0193] Wherein, the support structure position and size, cladding path, initial starting point, environmental temperature, laser power, welding gun diameter and cladding speed are optimized and modified according to the simulated annealing algorithm, and the simulation software is driven for analysis again until the result meeting the requirements is obtained.

[0194] Specifically, the probability p that the process parameter is accepted from x1 to x2 satisfies:

[0195]

[0196] Wherein, △f = f1-f2 is the difference between the values of the objective functions f(x1), f(x2);

[0197] f(x) is a quality function, representing the maximum strain.

[0198] T is a temperature constant;

[0199] As time decreases, when the temperature is high, the second probability is high, and as the temperature decreases, the second probability decreases slowly until almost no second probability is selected.

[0200] Wherein, the relationship between the iteration number n and the temperature constant T satisfies:

[0201] T(n+1) = K*T(n)

[0202] Wherein, K is a constant less than 1.

[0203] In this way, the laser selective melting process parameters are optimized through the simulated annealing optimization model until the laser selective melting process parameters meeting the requirements are obtained, finally, the process of obtaining the final result and the final result are saved to the process database, and based on the updated tool database, the model-process-quality relationship prediction model and the simulated annealing optimization model and the related rule base and simulated annealing algorithm are optimized.

[0204] Also included is a business terminal for process personnel to operate and interact with the interface;

[0205] In which, the business terminal process personnel can query, save, modify, read and use process information on the station.

[0206] The management terminal is used for account maintenance and permission setting for system management users, and for information query and management, statistics and analysis.

[0207] In which, the management terminal also includes establishing, querying, saving and modifying process personnel platform usage permissions.

[0208] Embodiment 2:

[0209] A laser selective melting intelligent process design and simulation optimization method, comprising:

[0210] Step 1: Simplify the forming target part and support structure using a three-dimensional model construction module, save it as an XT format, and build a three-dimensional model of the target part and the tooling to obtain the target part and the tooling;

[0211] Step 2: Use the model feature recognition module to identify straight lines, fillets, curved surfaces, holes, planes, stretched and rotated entities and their feature relationships in the three-dimensional model to obtain the surface features of the target part to be machined;

[0212] Step 3: Based on the surface features of the target part to be machined, use the knowledge base to obtain the laser selective melting initial recommended process parameters of the target part from the process database;

[0213] Step 4: Transfer the obtained laser selective melting initial recommended process parameters and three-dimensional model to the finite element analysis software through the simulation interaction module to perform finite element analysis on the laser selective melting initial process parameters to obtain the target part forming surface and stress-strain distribution-time data;

[0214] Step 5: Transfer all finite element analysis results to the knowledge base through the simulation interaction module to compare and analyze the target part forming surface and stress-strain distribution-time data with the technical requirements in the process database through the knowledge base;

[0215] Step 6: Use the simulated annealing algorithm to optimize the laser selective melting process parameter data that does not meet the requirements and return to step 4 until the laser selective melting process parameters that meet the requirements are obtained.

[0216] Specifically, in step 1, a three-dimensional model of the forming target part and the support structure is constructed using a three-dimensional model software.

[0217] Specifically, in step 2, the following is included: comparing the features of the surface to be processed of the three-dimensional model identified by the model feature recognition module with the parameters in the process database to determine the features of the target part to be processed.

[0218] The feature parameters in the process database include straight lines, fillets, curved surfaces, holes, planes, stretched and rotated entities, and feature relationship data.

[0219] Specifically, in step 3, the following is included:

[0220] S31: Based on the characteristics of the laser selective melting process of the target part, the laser selective melting process of the target part is obtained from the process knowledge base;

[0221] Specifically, the target part and the support structure are specified, the cladding path and the starting position are selected, and the target part material is selected, including titanium alloy and aluminum alloy.

[0222] The path and the starting position are determined by the collinearity on the coplanar surface identified by the model feature recognition, and the starting and ending points of each path are determined by the shortest path.

[0223] The material of the target part is selected according to the actual processing part material type.

[0224] S32: Based on the laser selective melting process of the target part and the features of the surface to be processed of the target part, the technical requirements for the precision and forming quality of the surface to be processed of the target part are obtained from the process database;

[0225] The technical requirements for the precision and forming quality of the surface to be processed of the target part are standard parameters stored in the process database.

[0226] S33: Based on the features of the surface to be processed of the target part and the technical requirements for the precision and forming quality of the surface to be processed of the target part, the similar laser selective melting process parameters of the target part are searched and obtained from the process database to obtain the initial recommended process parameters of the laser selective melting of the target part.

[0227] Specifically, according to the model-process-quality relationship prediction model and the related rule base in the knowledge base, the initial recommended process parameters are searched and calculated;

[0228] The three-layer neural network is set, including the input layer, the hidden layer and the output layer.

[0229] The number of nodes in the input layer, the hidden layer and the output layer is m0, m1 and m2, respectively.

[0230] The input layer satisfies:

[0231]

[0232] The hidden layer satisfies:

[0233]

[0234] The output layer satisfies:

[0235]

[0236] Among them, Y (0) Represents the input layer process parameter condition matrix;

[0237] Y (1) represents the output vector of the hidden layer;

[0238] Y (2) A mathematical relationship that represents the quality and workmanship of a model.

[0239] For the hidden layer:

[0240]

[0241] Among them, f (1) is the activation function, and for the selective laser melting process, we have:

[0242] f(x)=max(xα,x)

[0243] Among them, α is a hyperparameter obtained by data training, net (1) is the weighted amount.

[0244] The rule base is the boundary condition of the process, which is used to construct the process value range of laser selected melting, the quality requirements of the initial input, and the material fracture strain value.

[0245] Specifically, the search and calculation method is as follows: after determining the model characteristics and quality requirements, the knowledge base queries the saved model-process-quality relationship to see if it matches the known facts.

[0246] After all the IF clauses (relations, i.e., hidden layers) in the rule are matched, backtrack to the THEN clause, enable this rule, and add the new facts generated by the rule to the database. At this point, all the facts in the database are the inference results, and the inference ends.

[0247] The initial recommended process parameters for selective laser melting include: ambient temperature, laser power, welding gun diameter and cladding speed.

[0248] Specifically, in step 5, if the quality technical requirements are met, the comparison is successful and the final process parameters are obtained; if the quality technical requirements are not met, the comparison is unsuccessful.

[0249] Specifically, in step 6, the simulated annealing algorithm is used to optimize the laser selective melting process parameter data that does not meet the requirements, and returns to step 4 until the process parameters that meet the requirements are obtained.

[0250] Specifically, the probability p that the process parameter is accepted from x1 to x2 satisfies:

[0251]

[0252] Where, △f = f1-f2, is the difference between the values of the objective function f(x1), f(x2);

[0253] f(x) is a quality function, indicating the maximum strain.

[0254] T is a temperature constant;

[0255] Slowly decrease with time, when the temperature is very high, take the second probability, with the decrease of temperature, take the second case slowly, until almost no second case is selected.

[0256] Where, the relationship between the number of iterations n and the temperature constant T satisfies:

[0257] T(n+1) = K*T(n)

[0258] Where, K is a constant less than 1.

[0259] Wherein, the process parameters are laser power, scanning speed, scanning spacing, powder thickness.

[0260] Wherein, the support structure position and size, cladding path, initial starting point, environmental temperature, laser power, welding gun diameter and cladding speed are optimized and modified, and the simulation software is driven again for analysis until the required result is obtained.

[0261] Further, it also includes step 7: saving the process of obtaining the final result and the final result to the process database, and based on the updated tool database, optimizing the model-process-quality relationship prediction model and simulated annealing optimization model as well as the related rule base and simulated annealing algorithm.

[0262] Compared with the traditional process simulation design:

[0263] The traditional process simulation design needs to use orthogonal design when designing parameters. Taking the ribbed cylindrical part of the present application as an example, the wall thickness is 3mm, the ribs are longitudinal and transverse, arranged every 30mm with a height of 25mm, the cylindrical radius is 200mm, the depth is 400mm, and the process parameters are selected as follows:

[0264] Laser power Scan speed Scan spacing Powder thickness W1 (800 Watts) V1 (0.5 mm / s) T1 (0.05 mm) W1 (0.5 mm) W2 (850 Watts) V2 (1 mm / s) T2 (0.1 mm / s) W1 (1 mm) W3 (900 Watts) V3 (1.5 mm / s) T3 (0.15 mm / s) W3 (1.5 mmG) W4 (950 Watts) V4 (2 mm / s) T4 (0.22 mm / s) W4 (2 mmG)

[0265] According to the parameter expansion orthogonal design in the above, 4*4*4*4=256 groups of simulation need to be carried out to obtain the relationship between the process parameters and the quality, and the process parameters meeting the quality requirements are still unclear.

[0266] However, after the first group of simulation is completed, only 3-4 iterations are needed to obtain the required process parameters by using the method of the present application, and the design efficiency is greatly improved.

[0267] It can be seen that, compared with the traditional trial-and-error method and manual adjustment of simulation parameters and models, the intelligent process design and simulation optimization method of laser selective melting of the present application adopts a database automatic retrieval method, a knowledge base self-driven optimization method, and an automatic calculation of simulation software, and a network cloud platform sharing mode, realizes the intellectualization of process design and optimization, the digitization of process knowledge, the self-learning of knowledge base, and greatly improves the laser selective melting process design efficiency and the process test manufacturing cost.

[0268] Those skilled in the art can understand that all or part of the processes of the above-mentioned embodiments can be completed by a computer program instructing related hardware, and the program can be stored in a computer readable storage medium. The computer readable storage medium is a disk, an optical disk, a read-only memory or a random access memory, etc.

[0269] The above is only a preferred specific embodiment of the present application, but the protection scope of the present application is not limited thereto, and any person skilled in the art can easily think of changes or replacements within the technical range disclosed by the present application, which should be covered within the protection scope of the present application.

Claims

1. A laser selective melting intelligent process design and simulation optimization platform, characterized by: include: 3D model building module, used to simplify the forming target parts and supporting structures, save them in XT format, and build a 3D model of the target parts and tooling; The model feature recognition module is used to compare the straight lines, fillets, arcs, holes, planes, extruded and rotated entities and their feature relationships in the identified 3D model with the feature parameters in the process database to determine the features to be processed of the target part; Process database, used to store basic information of laser selective melting process; Knowledge base, based on the characteristics of the laser selective melting process of the target part, to obtain the laser selective melting process of the target part; Based on the laser selective melting process of the target part and the characteristics of the target part's surface to be processed, obtain the technical requirements for the target part's surface accuracy and forming quality; Based on the characteristics of the target part's surface to be processed and the technical requirements of the target part's surface accuracy and forming quality, similar laser selective melting process parameters of the target part are searched and obtained in the process database to obtain the initial recommended laser selective melting process parameters of the target part; The simulation interaction module is used to import the constructed three-dimensional model into the finite element analysis software, drive the finite element analysis software to perform calculation and analysis under different parameter iterations, and extract the calculation results for quality result data analysis.

2. The platform according to claim 1, characterized in that: The basic information of the laser selective melting process stored in the process database includes: the laser selective melting process based on the laser selective melting process characteristics of the target part, the laser selective melting process parameters based on the technical requirements of the precision and forming quality of the surface to be processed of the target part, and the technical requirements of the precision and forming quality of the surface to be processed based on the characteristics of the surface to be processed of the target part, and also includes feature parameter data of straight lines, fillets, arc surfaces, holes, planes, stretching and rotational entities and the relationship between each feature.

3. The platform according to claim 1, characterized in that: The knowledge base is provided with a model-process-quality relationship prediction model and a rule base for searching and calculating the initial recommended process parameters for selective laser melting.

4. The platform according to claim 3, characterized in that: The model-process-quality relationship prediction model includes: an input layer, a hidden layer and an output layer; Among them, the input layer satisfies: The hidden layer satisfies: The output layer satisfies: Among them, the number of nodes in the input layer, hidden layer and output layer are m0, m1 and m2 respectively; Among them, Y (0) Represents the input layer process parameter condition matrix; Y (1) represents the output vector of the hidden layer; Y (2) A mathematical relationship that represents the quality and workmanship of a model.

5. The platform according to claim 4, characterized in that: The hidden layer satisfies: Among them, f (1) is the activation function, and for the selective laser melting process, we have: f(x)=max(xα,x) Among them, α is a hyperparameter obtained by data training, net (1) is the weighted amount.

6. The platform according to claim 3, characterized in that: The rule base is the boundary condition of the process, which is used to construct the process value range of laser selective melting, the quality requirements of the initial input, and the material fracture strain value.

7. The platform according to claim 1, characterized in that: The knowledge base is provided with a simulated annealing optimization model to optimize process parameter data that does not meet the requirements; The simulated annealing optimization model is used to optimize and modify the support structure position and size, cladding path, initial starting point, ambient temperature, laser power, welding gun diameter and cladding speed.

8. A method for intelligent process design and simulation optimization of selective laser melting, characterized by: The method comprises using the laser selective melting intelligent process design and simulation optimization platform described in any one of claims 1 to 7 to obtain optimal laser selective melting process parameters of a target part.

9. The method according to claim 8, characterized in that include: Step 1: Use the 3D model building module to simplify the forming target part and supporting structure, save them in XT format, and build a 3D model of the target part and tooling; Step 2: Use the model feature recognition module to identify the straight lines, fillets, arcs, holes, planes, extrusions, and rotational entities and their feature relationships in the 3D model to obtain the surface features to be machined of the target part; Step 3: Based on the features of the target part's surface to be machined, the knowledge base is used to obtain the initial recommended process parameters for the target part's selective laser melting from the process database; Step 4: The obtained initial recommended process parameters and 3D model of the selective laser melting are transferred to the finite element analysis software through the simulation interaction module to perform finite element analysis on the initial process parameters of the selective laser melting to obtain the target part forming surface and stress-strain distribution-time data; Step 5: All finite element analysis results are transferred to the knowledge base through the simulation interaction module, so that the target part forming surface and stress-strain distribution-time data can be compared and analyzed with the technical requirements in the process database through the knowledge base; Step 6: Use the simulated annealing algorithm to optimize the laser selective melting process parameter data that does not meet the requirements, and return to step 4 until the laser selective melting process parameters that meet the requirements are obtained.

10. The method according to claim 9, characterized in that: The method also includes step 7: saving the process of obtaining the final result and the final result to the process database, and optimizing the model-process-quality relationship prediction model and the simulated annealing optimization model as well as the related rule base and the simulated annealing algorithm based on the updated tool database.

Images