Aluminum alloy electric arc additive manufacturing path optimization method based on hypergeometric optimization

Abstract

The invention discloses an aluminum alloy electric arc additive manufacturing path optimization method based on hypergeometric optimization, which comprises the following steps: S1, acquiring three-dimensional part model data and manufacturing area geometric feature data in an aluminum alloy electric arc additive manufacturing device to form manufacturing data; s2, performing regional division on the manufacturing data to generate an initial deposition path; s3, constructing a multi-target path optimization target function based on the initial deposition path to form an optimization target; s4, performing global search on the initial deposition path by using a hypergeometric optimization algorithm to obtain a candidate global optimal deposition path; s5, based on the candidate global optimal deposition path, performing local explosion search on the critical path area by adopting a pyrotechnic algorithm to form a local optimized deposition path; and S6, updating the comprehensive optimization deposition path in real time based on the local optimization deposition path in combination with real-time manufacturing monitoring data. The welding speed and the welding power are dynamically adjusted, so that the deposition quality of different areas is more balanced, and the welding defects of cracks and air holes are reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of industrial production, and particularly to a path optimization method for aluminum alloy arc additive manufacturing based on hypergeometric optimization. Background Art

[0002] In recent years, aluminum alloy arc additive manufacturing technology has gradually become a research hotspot in the fields of aerospace, automotive manufacturing, and precision component processing due to its advantages of high efficiency, low cost, and high material utilization rate. However, there are many challenges in the process of aluminum alloy arc additive manufacturing, such as the fluctuation of forming quality caused by uneven heat input, the structural deformation caused by the accumulation of residual stress, and the influence of unreasonable path planning in the manufacturing process of complex structures on the quality of the final workpiece.

[0003] Currently, traditional path planning methods for aluminum alloy arc additive manufacturing usually generate paths based on regularization strategies or heuristic algorithms. Regularization strategies mainly rely on simple equidistant scanning or zoning strategies, dividing the manufacturing area into several parallel or staggered scanning paths. However, the heat input distribution characteristics of the manufacturing area are not considered, which easily leads to local heat accumulation, resulting in welding defects or dimensional accuracy deviations in the parts. In addition, regularization strategies are difficult to adapt to the geometric structures of complex parts, especially when manufacturing curved surfaces or multi-layer staggered structures, the path planning methods often lead to uneven material accumulation or a decrease in the interlayer bonding strength.

[0004] Another type of method is based on heuristic algorithms, such as genetic algorithms and particle swarm optimization algorithms. To a certain extent, the path planning quality has been improved and the manufacturing efficiency has been increased through search optimization techniques. However, the existing heuristic optimization methods still face many challenges in the path planning process. On the one hand, the existing algorithms are prone to falling into local optimal solutions during the optimization process, resulting in limited path planning effects and difficulty in achieving global optimal path optimization. On the other hand, the traditional optimization methods have relatively limited considerations in terms of heat input balance and residual stress control, and cannot perform refined optimization for key areas, resulting in problems such as heat concentration and structural deformation during the forming process. In addition, some optimization methods have a slow convergence speed during the calculation process, affecting the manufacturing efficiency and unable to meet the requirements of high-efficiency and precise manufacturing.

[0005] Therefore, there is an urgent need for a new path optimization method to break through the limitations of traditional path planning methods and achieve better manufacturing quality and higher manufacturing efficiency. Summary of the Invention

[0006] An object of the present invention is to propose a path optimization method for aluminum alloy arc additive manufacturing based on hypergeometric optimization. The present invention dynamically adjusts the welding speed and welding power to make the deposition quality of different regions more balanced, reducing welding defects such as cracks and pores.

[0007] A method for optimizing the path of arc additive manufacturing of aluminum alloy based on hypergeometric optimization according to an embodiment of the present invention includes the following steps:

[0008] S1. Collect three-dimensional part model data and manufacturing area geometric feature data in the aluminum alloy arc additive manufacturing device to form manufacturing data, and set initial manufacturing parameters according to the physical properties and manufacturing requirements of the aluminum alloy material;

[0009] S2. Divide the manufacturing data into regions, divide the manufacturing area into multiple manufacturing units according to the part structure and process requirements, and generate an initial deposition path;

[0010] S3. Construct a multi-objective path optimization objective function based on the initial deposition path, and set corresponding optimization constraint conditions to form an optimization objective;

[0011] S4. Use the hypergeometric optimization algorithm to globally search the initial deposition path, evaluate the initial deposition path according to the optimization objective, and obtain a candidate globally optimal deposition path;

[0012] S5. Based on the candidate globally optimal deposition path, use the fireworks algorithm to perform local explosion search on the critical path area, optimize the deposition path layout in the critical area, and form a locally optimized deposition path;

[0013] S6. Based on the locally optimized deposition path and combined with real-time manufacturing monitoring data, dynamically adjust the welding parameters and deposition speed to obtain dynamically adjusted parameters, and update the comprehensively optimized deposition path in real time.

[0014] Optionally, the S1 specifically includes the following steps:

[0015] S11. Collect three-dimensional part model data in the aluminum alloy arc additive manufacturing device, and obtain manufacturing area geometric feature data. The manufacturing area geometric feature data includes size parameters, curvature changes, structural complexity, and boundary constraint conditions of the manufacturing area;

[0016] S12. According to the physical properties of the aluminum alloy material, obtain the material density ρ, specific heat capacity C p , thermal conductivity k, and linear expansion coefficient α, and set initial manufacturing parameters in combination with manufacturing requirements. The initial manufacturing parameters include welding current I 0 , welding voltage U 0 , welding speed v 0 and layer thickness h 0 ;

[0017] S13. According to the three-dimensional part model data and the manufacturing area geometric feature data, calculate the local heat input distribution Q in the manufacturing path i :

[0018]

[0019] Among them, η is the welding heat efficiency, and A i is the cross-sectional area of the local area;

[0020] S14. Calculate the initial heat conduction characteristics of the manufacturing area, establish a temperature gradient model, and perform calculations using Fourier's law of heat conduction to obtain the transient temperature field distribution:

[0021]

[0022] Among them, T is the temperature distribution, t is the time, is the Laplace operator;

[0023] S15. Based on the transient temperature field of the manufacturing area, calculate the size L of the heat affected zone HAZ :

[0024]

[0025] Among them, t f is the moment when the key heat input action ends during the manufacturing process, λ is the temperature decay constant, reflecting the cooling rate, represents the gradient distribution of the temperature field, reflecting the local temperature change, and γ is the correction index reflecting the sensitivity of the material microstructure to the thermal gradient;

[0026] S16. Based on the manufacturing data and heat input evaluation, construct the initial manufacturing conditions and output the manufacturing dataset D init :

[0027]

[0028] Among them, N represents the total number of manufacturing units divided in the manufacturing area, and M i is the 3D part model data of the i-th manufacturing unit, reflecting the geometric shape of each area, and P i represents the set of initial manufacturing parameters corresponding to the i-th manufacturing unit, including the welding current I 0 , the welding voltage U 0 , the welding speed v 0 and the layer thickness h 0 , and T grad,i is the local temperature gradient of the i-th manufacturing unit:

[0029]

[0030] Among them, k is the thermal conductivity of the material, is the Laplace operator of the temperature field.

[0031] Optionally, the specific steps of S2 are as follows:

[0032] S21. Based on the manufacturing dataset D init Extract the 3D part model data and the geometric feature data of the manufacturing area, preliminarily divide the part manufacturing area, and obtain the manufacturing area division matrix R:

[0033] R = [R ij m×n ;

[0034] where m is the number of segmentation units of the part along the length direction, n is the number of segmentation units of the part along the width direction, and R ij is the i,j-th unit area;

[0035] S22. Refine the manufacturing area according to the material heat transfer characteristics and the local temperature gradient T grad,i to form the final manufacturing cell set Ω:

[0036]

[0037] where is the temperature gradient change rate of the k-th area, and ∈ is the preset temperature gradient change threshold;

[0038] S23. Calculate the center point coordinates C k of each manufacturing cell according to the final manufacturing cell set Ω, determine the deposition start and end points of each cell, and preliminarily generate the manufacturing path sequence P init = {C k | k ∈ Ω};

[0039] S24. Calculate the shortest connection method between paths according to the manufacturing path sequence P init and optimize the overall deposition order P opt , with the optimization goal of minimizing the total path length L:

[0040]

[0041] where d(C k , C k+1 ) is the Euclidean distance between adjacent manufacturing cells;

[0042] S25. Combine the optimized manufacturing path P opt , generate the initial deposition path, and adjust the deposition path according to the welding heat input model Q i and the heat affected zone L HAZ,i , and finally output the optimized initial deposition path P final :

[0043] P final = P opt +ΔP;

[0044] ​where ΔP is the path correction amount adjusted based on the heat input balance:

[0045]

[0046] where α is the path correction weight factor.

[0047] Optionally, step S3 specifically includes the following steps:

[0048] S31. Based on the initial deposition path P final Construct a multi-objective path optimization objective function. The multi-objective path optimization objective function F(P) consists of the forming quality F q , heat input balance F h , residual stress distribution F s and manufacturing efficiency F e :

[0049] F(P) = w 1 F q + w 2 F h + w 3 F s + w 4 F e ;

[0050] where w 1 , w 2 , w 3 and w 4 are the weight factors of each optimization objective respectively;

[0051] S32. The forming quality objective F q is evaluated based on the deposition path uniformity and the interlayer bonding quality:

[0052]

[0053] where h i is the actual deposition thickness of the i-th layer, and h 0 is the theoretical layer thickness;

[0054] S33. The heat input balance objective F h is calculated based on the local heat input model Q i and the heat affected zone size L HAZ,i , and the optimization objective is to minimize the heat input fluctuation:

[0055]

[0056] where is the mean value of the global heat input;

[0057] S34. The residual stress distribution objective F sTaking the temperature gradient in the manufacturing area as the influencing factor:

[0058]

[0059] Among them, is the temperature gradient of the i-th layer, is the average temperature gradient in the manufacturing area to ensure a balanced residual stress distribution;

[0060] S35. Manufacturing efficiency target F e Calculated based on the path length L and the deposition speed v i The optimization goal is to minimize the manufacturing time T total :

[0061]

[0062] Among them, L i is the path length of the i-th layer, and v i is the deposition speed of the corresponding layer;

[0063] S36. Set path optimization constraint conditions. The constraint conditions include heat input balance constraint, path continuity constraint, and manufacturing feasibility constraint. In the heat input balance constraint, the local heat input Q i does not exceed the set threshold δ Q , in the path continuity constraint, the connection between path segments is smooth, and the path turning angle θ does not exceed the set threshold θ max , in the manufacturing feasibility constraint, the deposition path conforms to the equipment motion range, and the path coordinates (x, y, z) need to satisfy:

[0064] (x, y, z) ∈ W workspace ;

[0065] Among them, W workspace is the manufacturing working space reachable by the equipment.

[0066] Optionally, the specific steps of S4 are as follows:

[0067] S41. Based on the initial deposition path P final As the starting input for global search, by introducing diversity perturbations to the initial deposition path P final Generate a candidate path set P init :

[0068]

[0069] Among them, represents the k-th candidate path generated based on P final after perturbation, and M is the total number of candidate paths;

[0070] S42. For the candidate path set Pinit each candidate path in Improve the multi-objective path optimization objective function and construct an improved multi-objective path optimization objective function

[0071]

[0072] wherein, is the value of the multi-objective path optimization objective function, is a regulation factor reflecting the deviation between local heat input and temperature gradient:

[0073]

[0074] wherein, represents the candidate path corresponding local heat input, is the global heat input mean value, represents the local temperature gradient corresponding to the candidate path, is the global average temperature gradient, and β and μ are regulation factors;

[0075] S43. Use the improved hypergeometric optimization algorithm for the candidate path Calculate its selection probability according to the improved multi-objective path optimization objective function

[0076]

[0077] wherein, M opt is the number of current relatively good candidate paths, and m k is the number of times the k-th candidate path is selected during the optimization process, represents the combination number;

[0078] S44. According to the selection probability Screen the candidate paths to form a set P of relatively good candidate paths best :

[0079]

[0080] wherein, P threshold is a preset selection threshold;

[0081] S45. Within the set P of relatively good candidate paths best Calculate the average value of the improved multi-objective path optimization objective function and introduce the global search convergence index

[0082]

[0083] wherein, N1 For P best The number of candidate paths in

[0084] S46. When the global search convergence index is less than the set convergence threshold δ D select the candidate path with the smallest improved multi-objective path optimization objective function as the candidate global optimal deposition path P global ;

[0085] S47. If the global search convergence index is greater than the set convergence threshold δ D return to S42, perform a new global search on the candidate path set P init and adaptively update the regulation factors β and μ according to the real-time manufacturing monitoring data:

[0086]

[0087] where ΔQ is the deviation between the actual local heat input and the average global heat input, and ΔT grad is the deviation between the actual local temperature gradient and the average global temperature gradient, and β new and μ new are the updated regulation factors;

[0088] S48. Repeat steps S42 to S47 until the convergence condition is met, and finally output the candidate global optimal deposition path P global .

[0089] Optionally, the S5 specifically includes the following steps:

[0090] S51. Based on the local heat input deviation ΔQ global and the local temperature gradient deviation ΔT i of the candidate global optimal deposition path P grad,i , construct the critical path area recognition function J(R i ):

[0091]

[0092] where ω 4 , ω 5 are weight factors. When the critical path area recognition function J(R i ) ≥ J threshold , where J threshold is the set critical area recognition threshold, define the area R i as the critical path area, and form the critical area set R key ;

[0093] S52. For the set R of critical path regions key , an improved fireworks algorithm is used to perform local explosion search to generate the number of fireworks individuals n key,i at the center position C of the critical region i :

[0094]

[0095] where n base is the basic number of fireworks, and n max is the upper limit of the increment of the number of fireworks individuals;

[0096] S53. The generation direction of the fireworks individuals is determined according to the local heat input change gradient and the temperature gradient change rate to generate the directional fireworks individual path layout

[0097]

[0098] where σ 1 is the perturbation amplitude coefficient, ε (i,j) is the local random perturbation amount, and ε 1 , ε 2 are very small constants to prevent the denominator from being zero;

[0099] S54. Based on the improved multi-objective path optimization objective function F * (P), the fitness of all fireworks individuals is evaluated, and the explosion radius r of the fireworks individuals is determined i,j :

[0100]

[0101] where r max is the maximum explosion radius, are respectively the maximum and minimum values of the fitness in the fireworks individuals, γ is the sensitivity factor for dynamically adjusting the explosion radius, and ∈ 0 is a very small constant to prevent the denominator from being zero;

[0102] S55. Perform a secondary directional explosion within the explosion radius r of the fireworks individuals i,j to generate a local optimized path layout with the local heat input fluctuation and the temperature gradient change rate as the reference directions

[0103]

[0104] where ξ (i,j,l)is a directional perturbation random variable, and l is the sequence number of the secondary explosion path layout;

[0105] S56. According to the fitness evaluation result of the improved multi-objective path optimization objective function F * (P), select the path layout with the lowest fitness among all the firework individuals and the secondary directional explosion path layouts as the local optimal path layout of the key area

[0106]

[0107] S57. Replace all the local optimal path layouts of the key areas to the corresponding areas in the candidate global optimal deposition path P global to finally form the locally optimized deposition path P local .

[0108] The beneficial effects of the present invention are as follows:

[0109] (1) The present invention uses the hypergeometric optimization algorithm for global search to obtain the candidate global optimal deposition path, and at the same time combines the fireworks algorithm to perform local explosion search optimization on the key areas, overcoming the problem that the traditional path planning methods based on genetic algorithms or particle swarm optimization are prone to fall into local optima. Through the probability search characteristics of the hypergeometric optimization algorithm, the search space is effectively expanded, and the global optimization ability is improved. At the same time, the fireworks algorithm performs local search on the areas with uneven heat input and key geometric areas, making the optimized deposition path more balanced as a whole and reducing the structural deformation problem caused by uneven heat input.

[0110] (2) The present invention constructs a multi-objective optimization function including forming quality, heat input balance, residual stress distribution and manufacturing efficiency, and sets heat input balance constraints, path continuity constraints and manufacturing feasibility constraints, so as to ensure that the optimized path not only reduces the total path length, but also optimizes the heat input distribution, improves the residual stress balance and manufacturing efficiency. Among them, the heat input balance objective reduces the heat accumulation problem and improves the interlayer bonding quality by constraining the local heat input; the manufacturing feasibility constraint ensures that the path conforms to the equipment movement range, so that the optimized path can be actually manufactured and executed.

[0111] (3) On the basis of path optimization, the present invention combines real-time manufacturing monitoring data and optimizes the deposition path by dynamically adjusting welding parameters. Most of the traditional additive manufacturing path optimization methods are based on static parameter settings and cannot adapt to the real-time environmental changes during the manufacturing process. The deformation or uneven welding caused by excessive local heat input is solved by dynamically adjusting the welding speed and welding power in combination with the heat input model and temperature gradient feedback, making the deposition quality in different areas more balanced and reducing welding defects such as cracks and pores. Brief Description of the Drawings

[0112] The drawings are used to provide a further understanding of the present invention, and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention, and do not constitute a limitation to the present invention. In the drawings:

[0113] Figure 1 is a flowchart of a method for optimizing the deposition path of aluminum alloy arc additive manufacturing based on hypergeometric optimization proposed by the present invention. Detailed Embodiments

[0114] Now, the present invention will be further described in detail with reference to the drawings. These drawings are all simplified schematic diagrams, only showing the basic structure of the present invention in a schematic way, so they only show the components related to the present invention.

[0115] Reference Figure 1 , a method for optimizing the deposition path of aluminum alloy arc additive manufacturing based on hypergeometric optimization, includes the following steps:

[0116] S1. Collect three-dimensional part model data and manufacturing area geometric feature data in the aluminum alloy arc additive manufacturing device to form manufacturing data, and set initial manufacturing parameters according to the physical properties and manufacturing requirements of the aluminum alloy material;

[0117] S2. Divide the manufacturing data into regions, divide the manufacturing area into multiple manufacturing units according to the part structure and process requirements, and generate an initial deposition path;

[0118] S3. Construct a multi-objective path optimization objective function based on the initial deposition path, and set corresponding optimization constraint conditions to form an optimization objective;

[0119] S4. Use the hypergeometric optimization algorithm to perform a global search on the initial deposition path, evaluate the initial deposition path according to the optimization objective, and obtain a candidate globally optimal deposition path;

[0120] S5. Based on the candidate globally optimal deposition path, use the fireworks algorithm to perform a local explosion search on the critical path area, optimize the deposition path layout in the critical area, and form a locally optimized deposition path;

[0121] S6. Based on the locally optimized deposition path and combined with real-time manufacturing monitoring data, dynamically adjust the welding parameters and deposition speed to obtain dynamically adjusted parameters, and update the comprehensively optimized deposition path in real time.

[0122] In this embodiment, S1 specifically includes the following steps:

[0123] S11. Collect the 3D part model data in the aluminum alloy arc additive manufacturing device, and obtain the geometric feature data of the manufacturing area. The geometric feature data of the manufacturing area includes the size parameters, curvature changes, structural complexity, and boundary constraint conditions of the manufacturing area;

[0124] S12. According to the physical properties of the aluminum alloy material, obtain the material density ρ, specific heat capacity C p , thermal conductivity k, and linear expansion coefficient α. Combine the manufacturing requirements to set the initial manufacturing parameters. The initial manufacturing parameters include welding current I 0 , welding voltage U 0 , welding speed v 0 and layer thickness h 0 ;

[0125] S13. According to the 3D part model data and the geometric feature data of the manufacturing area, calculate the local heat input distribution Q in the manufacturing path i :

[0126]

[0127] where η is the welding heat efficiency, and A i is the cross-sectional area of the local area;

[0128] S14. Calculate the initial heat conduction characteristics of the manufacturing area, establish a temperature gradient model, and use Fourier's law of heat conduction for calculation to obtain the transient temperature field distribution:

[0129]

[0130] where T is the temperature distribution, t is the time, is the Laplace operator;

[0131] S15. Based on the transient temperature field in the manufacturing area, calculate the heat affected zone size L HAZ :

[0132]

[0133] where t f is the moment when the key heat input action ends during the manufacturing process, λ is the temperature decay constant, reflecting the cooling speed, represents the gradient distribution of the temperature field, reflecting the local temperature change, and γ is the correction index reflecting the sensitivity of the material microstructure to the thermal gradient;

[0134] S16. Based on the manufacturing data and heat input evaluation, construct the initial manufacturing conditions and output the manufacturing dataset D init :

[0135]

[0136] Among them, N represents the total number of manufacturing units divided in the manufacturing area, and M i is the 3D part model data of the i-th manufacturing unit, reflecting the geometric shape of each area, and P i represents the set of initial manufacturing parameters corresponding to the i-th manufacturing unit, including the welding current I 0 , the welding voltage U 0 , the welding speed v 0 and the layer thickness h 0 , and T grad,i is the local temperature gradient of the i-th manufacturing unit:

[0137]

[0138] Among them, k is the thermal conductivity of the material, is the Laplace operator of the temperature field.

[0139] In this embodiment, S2 specifically includes the following steps:

[0140] S21. Based on the manufacturing dataset D init Extract the 3D part model data and the geometric feature data of the manufacturing area, and preliminarily divide the part manufacturing area to obtain the manufacturing area division matrix R:

[0141] R = [R ij m×n ;

[0142] Among them, m is the number of segmentation units of the part along the length direction, n is the number of segmentation units of the part along the width direction, and R ij is the i,j-th unit area;

[0143] S22. Refine the manufacturing area according to the material heat transfer characteristics and the local temperature gradient T grad,i to form the final manufacturing unit set Ω:

[0144]

[0145] Among them, is the temperature gradient change rate of the k-th area, and ∈ is the preset temperature gradient change threshold;

[0146] S23. Calculate the center point coordinates C of each manufacturing unit according to the final manufacturing unit set Ω k , determine the deposition start and end points of each unit, and preliminarily generate the manufacturing path sequence P init = {C k | k ∈ Ω};

[0147] S24. Calculate the shortest connection method between paths according to the manufacturing path sequence P init and optimize the overall deposition order P​opt , the optimization objective is to minimize the total path length L:

[0148]

[0149] where d(C k , C k+1 ) is the Euclidean distance between adjacent manufacturing cells;

[0150] S25. Combine the optimized manufacturing path P opt to generate an initial deposition path, and adjust the deposition path according to the welding heat input model Q i and the heat affected zone L HAZ,i to finally output the optimized initial deposition path P final :

[0151] P final = P opt + ΔP;

[0152] where ΔP is the path correction amount adjusted based on heat input balance:

[0153]

[0154] where α is the path correction weight factor.

[0155] In this embodiment, S3 specifically includes the following steps:

[0156] S31. Construct a multi-objective path optimization objective function based on the initial deposition path P final . The multi-objective path optimization objective function F(P) consists of the forming quality F q , the heat input balance F h , the residual stress distribution F s and the manufacturing efficiency F e :

[0157] F(P) = w 1 F q + w 2 F h + w 3 F s + w 4 F e ;

[0158] where w 1 , w 2 , w 3 and w 4 are the weight factors of each optimization objective respectively;

[0159] S32. The forming quality objective F q is evaluated based on the deposition path uniformity and the interlayer bonding quality:

[0160]

[0161] where h i is the actual deposition thickness of the i-th layer, and h 0 is the theoretical layer thickness;

[0162] S33. Thermal input balance objective F h is calculated based on the local thermal input model Q i and the size L of the heat affected zone HAZ,i with the optimization objective of minimizing the thermal input fluctuation:

[0163]

[0164] where is the mean value of the global thermal input;

[0165] S34. Residual stress distribution objective F s with the temperature gradient in the manufacturing area as the influencing factor:

[0166]

[0167] where is the temperature gradient of the i-th layer, is the average temperature gradient in the manufacturing area to ensure balanced residual stress distribution;

[0168] S35. Manufacturing efficiency objective F e is calculated based on the path length L and the deposition speed v i with the optimization objective of minimizing the manufacturing time T total :

[0169]

[0170] where L i is the path length of the i-th layer, and v i is the deposition speed of the corresponding layer;

[0171] S36. Set path optimization constraints, including thermal input balance constraint, path continuity constraint, and manufacturing feasibility constraint. In the thermal input balance constraint, the local thermal input Q i shall not exceed the set threshold δ Q . In the path continuity constraint, the connection between path segments is smooth, and the path turning angle θ shall not exceed the set threshold θ max . In the manufacturing feasibility constraint, the deposition path conforms to the equipment movement range, and the path coordinates (x, y, z) shall satisfy:

[0172] (x, y, z) ∈ W workspace ;

[0173] Among them, W workspace is the manufacturing workspace accessible to the device.

[0174] In this embodiment, S4 specifically includes the following steps:

[0175] S41. Based on the initial deposition path P final as the starting input for global search, by introducing diversity perturbations to the initial deposition path P final a candidate path set P init is generated:

[0176]

[0177] Among them, represents the k-th candidate path generated after perturbation based on P final and M is the total number of candidate paths;

[0178] S42. For each candidate path in the candidate path set P init the multi-objective path optimization objective function is improved to construct an improved multi-objective path optimization objective function

[0179]

[0180] Among them, is the multi-objective path optimization objective function value, is a regulation factor reflecting the deviation between local heat input and temperature gradient:

[0181]

[0182] Among them, represents the local heat input corresponding to the candidate path , is the global heat input mean value, represents the local temperature gradient corresponding to the candidate path, is the global average temperature gradient, and β and μ are regulation factors;

[0183] S43. Using the improved hypergeometric optimization algorithm to calculate the selection probability of the candidate path according to the improved multi-objective path optimization objective function

[0184]

[0185] Among them, M opt is the current number of relatively good candidate paths, m k is the number of times the k-th candidate path is selected during the optimization process, Denote the combination number;

[0186] S44. According to the selection probability Filter the candidate paths to form a set P of better candidate paths best :

[0187]

[0188] where P threshold is a preset selection threshold;

[0189] S45. Within the set P of better candidate paths best calculate the average value of the improved multi-objective path optimization objective function and introduce the global search convergence index

[0190]

[0191] where N 1 is the number of candidate paths in P best ;

[0192] S46. When the global search convergence index is less than the set convergence threshold δ D select the candidate path with the minimum improved multi-objective path optimization objective function as the candidate global optimal deposition path P global ;

[0193] S47. If the global search convergence index is greater than the set convergence threshold δ D return to S42 to perform a new global search on the set P of candidate paths init and adaptively update the regulation factors β and μ according to the real-time manufacturing monitoring data:

[0194]

[0195] where ΔQ is the deviation between the actual local heat input and the mean value of the global heat input, and ΔT grad is the deviation between the actual local temperature gradient and the mean value of the global temperature gradient, β new and μ new are the updated regulation factors;

[0196] S48. Repeat steps S42 to S47 until the convergence condition is met, and finally output the candidate global optimal deposition path P global .

[0197] In this embodiment, S5 specifically includes the following steps:

[0198] S51. Based on the local heat input deviation ΔQ global and the local temperature gradient deviation ΔT i along the candidate global optimal deposition path P grad,i , construct the critical path region recognition function J(R i ):

[0199]

[0200] where ω 4 , ω 5 are weight factors. When the critical path region recognition function J(R i ) ≥ J threshold , where J threshold is the set critical region recognition threshold, define the region R i as the critical path region, forming the critical region set R key ;

[0201] S52. For the critical path region set R key , use the improved fireworks algorithm to perform local explosion search to generate the number of fireworks individuals n key,i at the center position C i of the critical region:

[0202]

[0203] where n base is the basic number of fireworks, and n max is the upper limit of the increment of the number of fireworks individuals;

[0204] S53. The generation direction of the fireworks individuals is determined according to the local heat input change gradient and the temperature gradient change rate , generating a directional fireworks individual path layout

[0205]

[0206] where σ 1 is the perturbation amplitude coefficient, ε (i,j) is the local random perturbation amount, and ε 1 , ε 2 are extremely small constants to prevent the denominator from being zero;

[0207] S54. Based on the improved multi-objective path optimization objective function F * (P), evaluate the fitness of all fireworks individuals and determine the explosion radius r i,j of the fireworks individuals:

[0208]

[0209] Among them, r max is the maximum explosion radius, are respectively the maximum and minimum values of the fitness in the individual fireworks, γ is the sensitivity factor for dynamically adjusting the explosion radius, and ∈ 0 is a very small constant to prevent the denominator from being zero;

[0210] S55. Perform secondary directional explosion within the range of the explosion radius r i,j of the individual fireworks, with the local heat input fluctuation and the temperature gradient change rate as the reference directions to generate the local optimized path layout

[0211]

[0212] Among them, ξ (i,j,l) is the directional perturbation random variable, and l is the serial number of the secondary explosion path layout;

[0213] S56. According to the fitness evaluation result of the improved multi-objective path optimization objective function F * (P), select the path layout with the lowest fitness among all the individual fireworks and the secondary directional explosion path layouts as the local optimal path layout of the key area

[0214]

[0215] S57. Replace all the local optimal path layouts of the key areas to the corresponding areas in the candidate global optimal deposition path P global to finally form the locally optimized deposition path P local .

[0216] Example 1:

[0217] On June 12, 2023, at an aerospace manufacturing base, the technical team received a task to manufacture a batch of aluminum alloy parts for a spacecraft cabin connection bracket. The parts are designed with high precision and high strength. The material is selected as 7075 aluminum alloy, with dimensions of 600mm × 400mm × 300mm, large wall thickness variation, high curvature in some areas, and serious heat accumulation risk during the manufacturing process.

[0218] The team originally planned to use the traditional equidistant scanning path for arc additive manufacturing, but in the early trial production stage, the following problems were found:

[0219] In the path intersection area and the positions with large curvature changes, the local heat input is too high, resulting in thermal cracks on the surface of the parts, and the crack rate reaches 7.5%.

[0220] Due to the uneven heat input, the overall part is deformed, and the final measured dimension error is as high as 0.48 mm, far exceeding the standard range (±0.3 mm).

[0221] The tensile stress in the residual stress concentration area reaches 135 MPa, seriously affecting the mechanical properties of the part.

[0222] Facing these problems, the technical team decided to adopt the path optimization method based on hypergeometric optimization of the present invention, hoping to improve the manufacturing quality, increase the forming accuracy, and reduce welding defects.

[0223] On June 14, 2023, the team used a Leica AT960 laser scanner to perform 3D modeling on the part, and combined with the welding process requirements, key manufacturing parameters were extracted:

[0224] Material density: 2.7 g / cm 3 ; Welding current: 120 A; Welding voltage: 26 V; Welding speed: 5 mm / s; Heat input threshold: not exceeding 500 J / mm at most 2 .

[0225] For the geometric complexity and heat input requirements of this part, the team first used an equidistant scanning path for simulation testing. The simulation results showed that the heat input in two key areas of the part exceeded 530 J / mm 2 , which means that excessive local heat accumulation leads to cracking and deformation. Therefore, path optimization must be carried out to reduce the heat input peak and improve the heat balance.

[0226] The team decided to adopt the path optimization method of the present invention, combine the hypergeometric optimization algorithm for global search to generate candidate paths, and combine the fireworks algorithm for local optimization in key areas.

[0227] On June 15, 2023, the team ran the hypergeometric optimization algorithm to globally optimize the initial path based on the optimization objectives (minimum path length, balanced heat input, reduced residual stress).

[0228] Before optimization, the total length of the equidistant scanning path was 85.6 m, and the theoretically calculated manufacturing time was about 13.2 hours. After optimization, the length of the globally optimized path was reduced to 77.3 m, and the manufacturing time was expected to be shortened by 11.8%.

[0229] The optimization results show that only using hypergeometric optimization can significantly shorten the path length, but after combining the fireworks algorithm, the heat input fluctuation is reduced by 55%, making the manufacturing process more stable.

[0230] On June 16, 2023, in the central connecting rib and the left support structure area of the part, the team used thermal sensors and infrared thermal imagers to monitor the heat input and found that the temperature peaks in these areas still exceeded expectations, with the local temperature reaching as high as 420 °C (normally it should be below 380 °C). In order to further optimize the manufacturing path of these key areas, the team used the fireworks algorithm for local explosion search and adjusted the welding path.

[0231] Adjusted parameters:

[0232] Welding speed adjustment in the key area: decreased from 5.0 mm / s to 4.3 mm / s; Welding current adjustment: decreased from 120 A to 115 A; Path fine-tuning: adjusted the local interlayer spacing, reduced from 1.5 mm to 1.3 mm.

[0233] After optimization, the temperature peak in this area decreased to 385 °C, and the welding defect rate decreased from 7.5% to 2.4%, effectively improving the forming quality.

[0234] On June 18, 2023, the manufacturing team used the OTCDaihen WB-M350 welding equipment to carry out actual manufacturing according to the optimized path and compared it with the traditional method.

[0235] Comparison items Traditional method (equidistant scanning) Method of the present invention (optimized path) Improvement rate Manufacturing time (h) 13.2 11.6 -12.1％ Welding defect rate (%) 7.5 2.4 -68.0％ Coefficient of thermal input fluctuation (%) ±9.8 ±4.3 -55.1％ Forming accuracy error (mm) 0.48 0.27 -43.7％ Peak residual stress (MPa) 135 98 -27.4％

[0236] After manufacturing, the team carried out X-ray non-destructive testing on the finished parts, and the results showed that:

[0237] For the parts manufactured by the traditional method, 12 microcracks were detected in the high-stress area, among which 4 exceeded the standard size (0.3 mm); for the parts manufactured by the method of the present invention, only 3 microcracks were detected, and the sizes were all within the acceptable range (≤0.1 mm).

[0238] In addition, the team carried out dimensional measurement after the parts were completed. The maximum dimensional deviation of the parts manufactured by the traditional method reached 0.48 mm, while the optimized manufacturing method of the present invention reduced the deviation to 0.27 mm, ensuring the high-precision requirements of the parts.

[0239] This embodiment proves the feasibility and effectiveness of the method of the present invention through actual manufacturing cases. Compared with the traditional path optimization method, the path optimization method of the present invention can:

[0240] Reduce manufacturing time and improve production efficiency (manufacturing time shortened by 12.1%); Reduce the welding defect rate and improve the forming quality (defect rate reduced by 68%); Optimize the heat input distribution and reduce thermal deformation (heat input fluctuation reduced by 55%); Improve the mechanical properties of the parts and reduce the residual stress (maximum residual stress reduced by 27.4%).

[0241] Overall, the method of the present invention provides a more stable and efficient path optimization strategy for arc additive manufacturing of aluminum alloys, which can be widely applied to industrial manufacturing scenarios with high precision and high requirements.

[0242] The present invention uses a hypergeometric optimization algorithm for global search to obtain candidate global optimal deposition paths. At the same time, it combines a fireworks algorithm to perform local explosion search optimization on key regions, overcoming the problem that traditional path planning methods based on genetic algorithms or particle swarm optimization are prone to falling into local optima. Through the probability search characteristics of the hypergeometric optimization algorithm, the search space is effectively expanded, and the global optimization ability is improved. At the same time, the fireworks algorithm performs local search on regions with uneven heat input and key geometric regions, making the optimized deposition path more balanced as a whole and reducing the structural deformation problem caused by uneven heat input.

[0243] The present invention constructs a multi-objective optimization function including forming quality, heat input balance, residual stress distribution, and manufacturing efficiency, and sets heat input balance constraints, path continuity constraints, and manufacturing feasibility constraints, so as to ensure that the optimized path not only reduces the total path length, but also optimizes the heat input distribution, improves the residual stress balance, and manufacturing efficiency. Among them, the heat input balance objective reduces the heat accumulation problem and improves the interlayer bonding quality by constraining the local heat input; the manufacturing feasibility constraint ensures that the path conforms to the equipment movement range, enabling the optimized path to be actually manufactured and executed.

[0244] Based on path optimization, the present invention combines real-time manufacturing monitoring data to optimize the deposition path by dynamically adjusting welding parameters. Most traditional additive manufacturing path optimization methods are based on static parameter settings and cannot adapt to real-time environmental changes during the manufacturing process. Problems such as deformation or uneven welding caused by excessive local heat input are solved by dynamically adjusting the welding speed and welding power of the deposition path in real time, combining the heat input model and temperature gradient feedback, making the deposition quality in different regions more balanced and reducing welding defects such as cracks and pores.

[0245] The above is only a preferred specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention, according to the technical solution and inventive concept of the present invention, making equivalent replacements or changes should be covered within the protection scope of the present invention.

Claims

1. A method for optimizing the path of aluminum alloy arc additive manufacturing based on hypergeometric optimization, characterized in that: The steps include: S1. Collect the three-dimensional part model data and the geometric feature data of the manufacturing area in the aluminum alloy arc additive manufacturing device to form manufacturing data, and set the initial manufacturing parameters according to the physical properties and manufacturing requirements of the aluminum alloy material; S2. Divide the manufacturing data into regions, divide the manufacturing region into multiple manufacturing units according to the part structure and process requirements, and generate an initial deposition path; S3. construct a multi-objective path optimization objective function based on the initial deposition path, and set corresponding optimization constraints to form an optimization goal; S4. Using a hypergeometric optimization algorithm to perform a global search for the initial deposition path, evaluating the initial deposition path according to the optimization goal, and obtaining a candidate global optimal deposition path; S5. Based on the candidate global optimal deposition path, a pyrotechnics algorithm is used to perform a local explosion search on the key path area, optimize the deposition path layout in the key area, and form a local optimized deposition path; S6. Based on the local optimized deposition path combined with real-time manufacturing monitoring data, the welding parameters and deposition speed are dynamically adjusted to obtain dynamic adjustment parameters, and the comprehensive optimized deposition path is updated in real time.

2. The method for optimizing aluminum alloy arc additive manufacturing path based on hypergeometric optimization according to claim 1, characterized in that: The S1 specifically includes the following steps: S11. Collecting three-dimensional part model data in the aluminum alloy arc additive manufacturing device and obtaining geometric feature data of the manufacturing area, the geometric feature data of the manufacturing area including dimensional parameters, curvature changes, structural complexity and boundary constraints of the manufacturing area; S12. According to the physical properties of aluminum alloy materials, obtain the material density ρ and specific heat capacity C p , thermal conductivity k and linear expansion coefficient α, and set initial manufacturing parameters in combination with manufacturing requirements, wherein the initial manufacturing parameters include welding current I0, welding voltage U0, welding speed v0 and layer thickness h0; S13. Calculate the local heat input distribution Q in the manufacturing path according to the three-dimensional part model data and the manufacturing area geometric feature data i : Among them, η is the welding thermal efficiency, A i is the cross-sectional area of ​​the local region; S14. Calculate the initial heat conduction characteristics of the manufacturing area, establish a temperature gradient model, and use Fourier's heat conduction law to calculate and obtain the transient temperature field distribution: Where T is the temperature distribution, t is the time, is the Laplace operator; S15. Based on the transient temperature field of the manufacturing area, calculate the size L of the heat affected zone HAZ : Among them, t f is the moment when the critical heat input in the manufacturing process ends, λ is the temperature decay constant, reflecting the cooling rate, It represents the gradient distribution of the temperature field and reflects the local temperature change. γ is a correction index that reflects the sensitivity of the material microstructure to the thermal gradient. S16. Based on manufacturing data and heat input evaluation, construct manufacturing initial conditions and output manufacturing data set D init : Where N represents the total number of manufacturing units divided in the manufacturing area, M i is the 3D part model data of the i-th manufacturing unit, reflecting the geometric shape of each area, P i represents the initial manufacturing parameter set corresponding to the i-th manufacturing unit, including welding current I0, welding voltage U0, welding speed v0 and layer thickness h0, T grad,i is the local temperature gradient of the i-th manufacturing unit: Where k is the thermal conductivity of the material, is the Laplace operator of the temperature field.

3. The method for optimizing aluminum alloy arc additive manufacturing path based on hypergeometric optimization according to claim 1, characterized in that: The S2 specifically includes the following steps: S21. Based on manufacturing dataset D init Extract the 3D part model data and the manufacturing area geometric feature data, make a preliminary division of the part manufacturing area, and obtain the manufacturing area division matrix R: R=[R ij ] m×n ; Among them, m is the number of segmentation units along the length direction of the part, n is the number of segmentation units along the width direction of the part, R ij is the i,jth unit area; S22. Based on the heat transfer characteristics of the material and the local temperature gradient T grad,i The manufacturing area is refined to form the final manufacturing unit set Ω: in, is the temperature gradient change rate of the kth region, ∈ is the preset temperature gradient change threshold; S23. Calculate the center point coordinates C of each manufacturing unit based on the final manufacturing unit set Ω k , determine the deposition start and end points of each unit, and preliminarily generate the manufacturing path sequence P init ={C k |k∈Ω}; S24. According to the manufacturing path sequence P init Calculate the shortest connection between paths and optimize the overall deposition order P opt , the optimization goal is to minimize the total path length L: Among them, d(C k ,C k+1 ) is the Euclidean distance between adjacent manufacturing units; S25. Combined with the optimized manufacturing path P opt , generate the initial deposition path, and according to the welding heat input model Q i and heat affected zone L HAZ,i Adjust the deposition path and finally output the optimized initial deposition path P final : P final =P opt +ΔP; Where ΔP is the path correction amount adjusted based on heat input balance: Among them, α is the path correction weight factor.

4. The method for optimizing aluminum alloy arc additive manufacturing path based on hypergeometric optimization according to claim 1, characterized in that: The S3 specifically includes the following steps: S31. Based on the initial deposition path P final Construct a multi-objective path optimization objective function, and the multi-objective path optimization objective function F(P) is composed of the forming quality F q , Heat input balance F h , residual stress distribution F s and manufacturing efficiency F e composition: F(P)=w1F q +w2F h +w3F s +w4F e ; Among them, w1, w2, w3 and w4 are the weight factors of each optimization objective; S32. Forming quality target F q The uniformity of deposition paths and the quality of interlayer bonding are used as evaluation criteria: Among them, h i is the actual deposition thickness of the i-th layer, and h0 is the theoretical layer thickness; S33. Heat input balance target F h Based on the local heat input model Q i And the size of the heat affected zone L HAZ,i The calculation is performed with the optimization goal of minimizing the heat input fluctuation: in, is the mean value of global heat input; S34. Residual stress distribution target F s Taking the temperature gradient in the manufacturing area as the influencing factor: in, is the temperature gradient of the i-th layer, The average temperature gradient of the manufacturing area ensures a balanced residual stress distribution; S35. Manufacturing efficiency target F e According to the path length L and deposition velocity v i Calculation, the optimization goal is to minimize the manufacturing time T total : Among them, L i is the path length of the i-th layer, v i is the deposition rate of the corresponding layer; S36. Set path optimization constraints, including heat input balance constraints, path continuity constraints, and manufacturing feasibility constraints. The local heat input Q i Does not exceed the set threshold δ Q , the connection between the path segments in the path continuity constraint is smooth, and the path turning angle θ does not exceed the set threshold θ max In the manufacturing feasibility constraint, the deposition path complies with the equipment motion range, and the path coordinates (x, y, z) must satisfy: (x,y,z)∈W workspace ; Among them, W workspace Provides a manufacturing workspace that is accessible to equipment.

5. The method for optimizing aluminum alloy arc additive manufacturing path based on hypergeometric optimization according to claim 1, characterized in that: The S4 specifically comprises the following steps: S41. Based on the initial deposition path P final As the starting input of the global search, the initial deposition path P final Introduce diversity perturbation to generate candidate path set P init : in, Indicates that based on P final The kth candidate path generated after perturbation, M is the total number of candidate paths; S42. For the candidate path set P init Each candidate path in Improve the multi-objective path optimization objective function and construct an improved multi-objective path optimization objective function in, Optimize the objective function value for multiple objective paths, To reflect the control factor of the local heat input and temperature gradient deviation: in, Represents candidate paths The corresponding local heat input, is the global heat input mean, represents the local temperature gradient corresponding to the candidate path, is the global average temperature gradient, β and μ are the control factors; S43. Using the improved hypergeometric optimization algorithm to optimize the candidate paths Calculate the selection probability based on the improved multi-objective path optimization objective function Among them, M opt is the number of currently preferred candidate paths, m k is the number of times the kth candidate path is selected during the optimization process, Indicates the number of combinations; S44. Based on the selection probability Screen the candidate paths to form a better candidate path set P best : Among them, P threshold is the preset selection threshold; S45. In the optimal candidate path set P best Calculate the average value of the improved multi-objective path optimization objective function And introduce the global search convergence index Where N1 is P best The number of candidate paths in ; S46. When the global search converges to the index Less than the set convergence threshold δ D When selecting the improved multi-objective path optimization objective function The smallest candidate path is taken as the candidate global optimal deposition path P global ; S47. If the global search convergence index Greater than the set convergence threshold δ D , then return to S42 and select the candidate path set P init Perform a global search again and adaptively update the control factors β and μ based on real-time manufacturing monitoring data: Where ΔQ is the deviation between the actual local heat input and the global heat input mean, ΔT grad is the deviation between the actual local temperature gradient and the global average temperature gradient, β new and μ new is the updated regulatory factor; S48. Repeat steps S42 to S47 until the convergence condition is met, and finally output the candidate global optimal deposition path P global .

6. The method for optimizing aluminum alloy arc additive manufacturing path based on hypergeometric optimization according to claim 1, characterized in that: The S5 specifically includes the following steps: S51. Based on the candidate global optimal deposition path P global The local heat input deviation ΔQ i and local temperature gradient deviation ΔT grad,i , construct the critical path area identification function J(R i ): Among them, ω4 and ω5 are weight factors. When the critical path area identification function J(R i )≥J threshold When J threshold The key area recognition threshold is set, and the area R i Defined as the critical path area, forming the critical area set R key ; S52. For the critical path area set R key , the improved fireworks algorithm is used to search for local explosions at the center of the key area C key,i The number of fireworks individuals generated at n i : Among them, n base is the number of basic fireworks, n max The increment limit of the number of fireworks. S53. The generation direction of fireworks individuals depends on the local heat input gradient and the temperature gradient change rate Determine and generate directional fireworks individual path layout Among them, σ1 is the disturbance amplitude coefficient, ε (i,j) is the local random disturbance, ε1 and ε2 are extremely small constants to prevent the denominator from being zero; S54. Based on the improved multi-objective path optimization objective function F * (P) For all firework entities Perform fitness evaluation and determine the explosion radius r of individual fireworks i,j : Among them, r max is the maximum explosion radius, are the maximum and minimum values ​​of fitness in individual fireworks, respectively; γ is the sensitivity factor for dynamically adjusting the explosion radius; ∈0 is a very small constant to prevent the denominator from being zero; S55. In the explosion radius r of individual fireworks i,j A secondary directional explosion is carried out within the range to achieve local heat input fluctuations and the temperature gradient change rate Generate local optimized path layout as reference direction Among them, ξ (i,j,l) is the directional disturbance random variable, l is the secondary explosion path layout number; S56. Based on the improved multi-objective path optimization objective function F * The fitness evaluation result of (P) is used to select the path layout with the lowest fitness among all fireworks individuals and secondary directional explosion path layouts as the local optimal path layout of the key area. S57. Layout the local optimal path of all key areas Replace to the candidate global optimal deposition path P global The corresponding area in the middle finally forms the local optimized deposition path P local .

Images