阅读说明：本技术 一种基于实时三维检测的金属增材制造在线轨迹调整方法 (Metal additive manufacturing online track adjusting method based on real-time three-dimensional detection ) 是由 闵晓潮 党仁俊 于 2021-01-03 设计创作，主要内容包括：本发明提出一种基于实时三维检测的金属增材制造在线轨迹调整方法,属于增材制造检测技术领域。本方法首先在增材制造的设备上安装实时三维检测装置并对打印轨迹进行离线规划,然后在开始制造后,通过实时三维检测装置测量过渡熔池、局部新生面和局部基材面的三维形貌信息,通过计算将其转化为当前离线规划轨迹的修正量,叠加补偿至当前离线规划轨迹,不断调整当前离线规划轨迹,直至增材制造完成。本方法基于已经成熟的实时三维检测技术,同时考虑基材面、熔池和新生面的形貌等特征,建立双闭环控制系统在线调整打印轨迹,稳定性好、精度高,可以有效提高金属增材制造成型质量。(The invention provides a metal additive manufacturing online track adjusting method based on real-time three-dimensional detection, and belongs to the technical field of additive manufacturing detection. The method comprises the steps of firstly installing a real-time three-dimensional detection device on the additive manufacturing equipment, carrying out off-line planning on a printing track, then measuring three-dimensional shape information of a transition molten pool, a local newly-generated surface and a local base material surface through the real-time three-dimensional detection device after manufacturing is started, converting the three-dimensional shape information into a correction amount of a current off-line planning track through calculation, superposing and compensating the correction amount to the current off-line planning track, and continuously adjusting the current off-line planning track until additive manufacturing is completed. The method is based on the mature real-time three-dimensional detection technology, simultaneously considers the features of the substrate surface, the molten pool, the newly generated surface and the like, establishes a double closed-loop control system to adjust the printing track on line, has good stability and high precision, and can effectively improve the metal additive manufacturing and forming quality.)

1. A metal additive manufacturing online track adjusting method based on real-time three-dimensional detection is characterized by comprising the following steps:

1) the real-time three-dimensional detection device is arranged at a transmission device of a printing head of the metal additive manufacturing equipment or at a fixed position in the printing machine, so that the detection range of the real-time three-dimensional detection device covers the lower part and the peripheral area of the printing head, and the printing head is avoided;

2) carrying out three-dimensional modeling on a metal component to be manufactured to obtain a three-dimensional model of the metal component, then layering the three-dimensional model to obtain a corresponding layering result, setting a growth direction and printing parameters, planning to obtain an initial offline printing trajectory, and taking the initial offline printing trajectory as a current offline planning trajectory gamma_plan；

3) Planning a trajectory gamma according to the current off-line_planPerforming additive manufacturing on a metal component to be manufactured, and planning a trajectory gamma according to the current off-line_planDuring the manufacturing process, the corresponding transition weld pool P is formed in the printing head region_FTransition molten pool P_FLocal new surfaces S are respectively formed around₁And S₁Local substrate surface S formed after delay time tau₂；

4) Acquisition of transition molten pool P using real-time three-dimensional detection device_FLocal fresh noodle S₁And local substrate surface S₂Three-dimensional shape information of;

5) the transition molten pool P collected according to the step 4)_FAnd local substrate surface S₂Establishing and calculating a first adjustment quantity delta gamma₁Model1 of (1); the method comprises the following specific steps:

5-1) defining a current offline planning trajectory gamma_planOn the local substrate surface S₂Each point on the table is a current off-line planning track point p_i，p_iNeighborhood of (2)From point p_tConstitution p_tNormal to the surface of n_tThen p is_iNeighborhood of (2)The expression of (a) is as follows:

wherein, | | p_t-p_iI represents the point p_tAnd point p_iOn the local substrate surface S₂A distance of (d) is a point p_tAnd point p_iOn the local substrate surface S₂The maximum distance of (a);

the local substrate surface S acquired according to the step 4)₂Using a surface processing algorithm to calculate each p_iCorresponding p_tAnd n_t；

5-2) the transition zone molten pool P collected according to step 4)_FExtracting the molten pool P of the transition region according to a shape extraction algorithm_FIncluding the center of the molten poolLong shaftShort shaftAnd the angle of inclination

5-3) according to a single-droplet structure model M_sbModel M with multiple molten drops_mbAll of p obtained in step 5-1)_tCorresponding surface normal n_tMajor axis of molten pool obtained in step 5-2)Short shaftAnd the angle of inclinationCalculating a first attitude correction amount of the print headWhereinFirst rotation angle correction quantities of the printing head in the x, y and z directions respectively;

5-4) according to the single-droplet structure model M_sbModel M with multiple molten drops_mbAll p obtained in step 5-2)_tAnd the center of the molten pool obtained in the step 5-2)Calculating a first displacement correction amount Deltav of the print head₁＝[Δx₁,Δy₁,Δz₁]^TWherein Δ x₁,Δy₁,Δz₁First displacement correction amounts of the printing head in x, y and z directions respectively;

5-5) obtaining a first attitude correction quantity Δ ω according to step 5-3)₁And the first displacement correction amount Deltav obtained in step 5-4)₁Establishing and calculating a first adjustment quantity delta gamma₁Model1, and calculates a first adjustment amount Δ Γ₁Wherein the Model1 expression is:

Model1:

6) the local new surface S collected according to the step 4)₁Establishing and calculating a second adjustment quantity delta gamma₂Model2 of (1); the method comprises the following specific steps:

6-1) collecting the local new surface S according to the step 4)₁Three-dimensional topography information of S₁As the current local new surface, based on the Markov process, a probability model for estimating the local new surface at the next moment according to the current local new surface is established to obtain the local new surface at the next momentThree-dimensional topography information;

6-2) extracting and obtaining the expected local new surface of the current printing head position at the next moment according to the three-dimensional model and the layering result of the metal component obtained in the step 2)Three-dimensional shape information of;

6-3) the local new noodle at the next momentAnd the next moment when local new noodles are expectedComparing to obtain a difference value

6-4) according to Δ S₁Single-molten-drop structure model M_sbAnd a multiple droplet overlay model M_mbCalculating a second attitude correction amount of the print headAnd a second displacement correction amount Deltav₂＝[Δx₂,Δy₂,Δz₂]^TWhereinSecond angular corrections, Δ x, of the print head in the x, y, z directions, respectively₂,Δy₂,Δz₂Second displacement correction amounts of the printing head in x, y and z directions respectively;

6-5) according to Δ S₁、Δω₂And Δ v₂Establishing and calculating the second adjustment quantity delta gamma₂Model2, and calculating a second adjustment amount Δ Γ₂Wherein the Model2 expression is:

7) for the first adjustment quantity delta gamma₁And a second adjustment amount DeltaGamma₂And superposing to obtain a total track adjustment quantity delta gamma, and then compensating the delta gamma to the current offline planning track gamma_planObtaining the track gamma after on-line adjustment_onlineThe specific calculation method comprises the following steps:

wherein the content of the first and second substances,representing a superposition compensation operation;

8) gamma-gamma is formed_onlineAs new gamma_planAnd then returning to the step 3) again to perform a new round of off-line planning track adjustment until the whole metal component to be manufactured is manufactured.

2. The method of claim 1, wherein the three-dimensional topographic information of step 4) is in the form of any one of a three-dimensional depth map, a three-dimensional voxel volume, a three-dimensional polygon mesh, and a three-dimensional point cloud.

Technical Field

The invention belongs to the technical field of additive manufacturing detection, and particularly provides a metal additive manufacturing online track adjusting method based on real-time three-dimensional detection

Background

The metal additive manufacturing technology is a technology which takes powder or wire materials as raw materials, based on the discrete/accumulation principle, uses laser, electric arc or electron beam as a heat source to melt/solidify, and directly manufactures a metal solid three-dimensional component by a digital three-dimensional model. Compared with the traditional material reduction manufacturing, the technology effectively overcomes the defects of long manufacturing period, low material utilization rate and the like of the complex component, and is more suitable for manufacturing large complex components such as aerospace, automobiles, medical appliances and the like. However, in the existing additive manufacturing process, the process is very complex, especially in the metal melting process, the forming quality is very sensitive to the surface of the substrate and the molten pool, and meanwhile, the new surface is also an important reference for forming precision feedback, and the factors need to be considered at the same time to perform online adjustment on important parameters such as the feeding amount, the temperature, especially the printing track and the like.

However, in the existing research, the most mature online monitoring based on the visual imaging technology mainly focuses on online detection of the physical and dimensional parameters of the molten pool, namely the width and the height, establishment of feedback control, adjustment of the printing track and improvement of the forming quality. In other forms of online detection technologies, such as online detection technologies based on spectrum, temperature, sound or electrical signals, part of the application research of the technologies is still in the starting stage, and the other part of the technologies obtains less information and can only be used as an assistant.

Disclosure of Invention

The invention aims to overcome the defects of the prior art and provides a metal additive manufacturing online track adjusting method based on real-time three-dimensional detection. The method is based on the mature real-time three-dimensional detection technology, simultaneously considers the features of the substrate surface, the molten pool, the newly generated surface and the like, establishes a double closed-loop control system to adjust the printing track on line, has good stability and high precision, and can effectively improve the metal additive manufacturing and forming quality.

The invention provides a metal additive manufacturing online track adjusting method based on real-time three-dimensional detection, which is characterized by comprising the following steps of:

1) the real-time three-dimensional detection device is arranged at a transmission device of a printing head of the metal additive manufacturing equipment or at a fixed position in the printing machine, so that the detection range of the real-time three-dimensional detection device covers the lower part and the peripheral area of the printing head, and the printing head is avoided;

2) carrying out three-dimensional modeling on a metal component to be manufactured to obtain a three-dimensional model of the metal component, then layering the three-dimensional model to obtain a corresponding layering result, setting a growth direction and printing parameters, planning to obtain an initial offline printing trajectory, and taking the initial offline printing trajectory as a current offline planning trajectory gamma_plan；

3) Planning a trajectory gamma according to the current off-line_planPerforming additive manufacturing on a metal component to be manufactured, and planning a trajectory gamma according to the current off-line_planDuring the manufacturing process, the corresponding transition weld pool P is formed in the printing head region_FTransition molten pool P_FLocal new surfaces S are respectively formed around₁And S₁Local substrate surface S formed after delay time tau₂；

4) Acquisition of transition molten pool P using real-time three-dimensional detection device_FLocal fresh noodle S₁And local substrate surface S₂Three-dimensional shape information of;

5) the transition molten pool P collected according to the step 4)_FAnd local substrate surface S₂Establishing and calculating a first adjustment quantity delta gamma₁Model1 of (1); the method comprises the following specific steps:

5-1) defining a current offline planning trajectory gamma_planOn the local substrate surface S₂Each point on the table is a current off-line planning track point p_i，p_iNeighborhood of (2)From point p_tConstitution p_tNormal to the surface of n_tThen p is_iNeighborhood of (2)The expression of (a) is as follows:

wherein, | | p_t-p_iI represents the point p_tAnd point p_iOn the local substrate surface S₂A distance of (d) is a point p_tAnd point p_iOn the local substrate surface S₂The maximum distance of (a);

the local substrate surface S acquired according to the step 4)₂Using a surface processing algorithm to calculate each p_iCorresponding p_tAnd n_t；

5-2) the transition zone molten pool P collected according to step 4)_FExtracting the molten pool P of the transition region according to a shape extraction algorithm_FIncluding the center of the molten poolLong shaftShort shaftAnd the angle of inclination

5-3) according to a single-droplet structure model M_sbModel M with multiple molten drops_mbAll of p obtained in step 5-1)_tCorresponding surface normal n_tMajor axis of molten pool obtained in step 5-2)Short shaftAnd the angle of inclinationCalculating a first attitude correction amount of the print headWhereinFirst rotation angle correction quantities of the printing head in the x, y and z directions respectively;

5-4) according to the single-droplet structure model M_sbModel M with multiple molten drops_mbAll p obtained in step 5-2)_tAnd the center of the molten pool obtained in the step 5-2)Calculating a first displacement correction amount Deltav of the print head₁＝[Δx₁,Δy₁,Δz₁]^TWherein Δ x₁,Δy₁,Δz₁First displacement correction amounts of the printing head in x, y and z directions respectively;

5-5) obtaining a first attitude correction quantity Δ ω according to step 5-3)₁And the first displacement correction amount Deltav obtained in step 5-4)₁Establishing and calculating a first adjustment quantity delta gamma₁Model1, and calculates a first adjustment amount Δ Γ₁Wherein the Model1 expression is:

6) the local new surface S collected according to the step 4)₁Establishing and calculating a second adjustment quantity delta gamma₂Model2 of (1); the method comprises the following specific steps:

6-1) collecting the local new surface S according to the step 4)₁Three-dimensional topography information of S₁As the current local new surface, based on the Markov process, a probability model for estimating the local new surface at the next moment according to the current local new surface is established to obtain the local new surface at the next momentThree-dimensional topography information;

6-2) extracting and obtaining the expected local new surface of the current printing head position at the next moment according to the three-dimensional model and the layering result of the metal component obtained in the step 2)Three-dimensional shape information of;

6-3) the local new noodle at the next momentAnd the next moment when local new noodles are expectedComparing to obtain a difference value

6-4) according to Δ S₁Single-molten-drop structure model M_sbAnd a multiple droplet overlay model M_mbCalculating a second attitude correction amount of the print headAnd a second displacement correction amount Deltav₂＝[Δx₂,Δy₂,Δz₂]^TWhereinSecond angular corrections, Δ x, of the print head in the x, y, z directions, respectively₂,Δy₂,Δz₂Second displacement correction amounts of the printing head in x, y and z directions respectively;

6-5) according to Δ S₁、Δω₂And Δ v₂Establishing and calculating the second adjustment quantity delta gamma₂Model2, and calculating a second adjustment amount Δ Γ₂Wherein the Model2 expression is:

7) for the first adjustment quantity delta gamma₁And a second adjustment amount DeltaGamma₂And superposing to obtain a total track adjustment quantity delta gamma, and then compensating the delta gamma to the current offline planning track gamma_planGet on-lineAdjusted trajectory gamma_onlineThe specific calculation method comprises the following steps:

wherein the content of the first and second substances,representing a superposition compensation operation;

8) gamma-gamma is formed_onlineAs new gamma_planAnd then returning to the step 3) again to perform a new round of off-line planning track adjustment until the whole metal component to be manufactured is manufactured.

The invention has the characteristics and beneficial effects that:

according to the metal additive manufacturing online track adjusting method based on real-time three-dimensional detection, a double closed-loop control system is established to adjust the printing track online by utilizing a mature real-time three-dimensional detection technology and considering the features of the substrate surface, the molten pool, the newly generated surface and the like, so that the stability is good, the precision is high, and the metal additive manufacturing molding quality can be effectively improved.

When the real-time three-dimensional detection result is processed, the problem of sparse or singular noise and the like caused by light reflection of the forming surface and shielding of partial area is solved, the stability of online control is improved, and a tensor voting frame is adopted to process the point cloud result.

The invention simultaneously considers the influence of a new surface, a molten pool and a base material surface on additive manufacturing, adopts two models to calculate the online track adjustment amount, forms a double closed-loop control system, and effectively improves the stability and the precision of the printing track.

The invention can be used for online adjustment of printing tracks in metal additive manufacturing, can be used as an auxiliary means in the production and manufacturing process, and has an important effect on improving the product quality.

Drawings

Fig. 1 is an overall flowchart of a metal additive manufacturing online trajectory adjustment method based on real-time three-dimensional detection according to the present invention.

Fig. 2 is a schematic diagram of a real-time three-dimensional detection device and a metal additive manufacturing apparatus according to an embodiment of the invention.

Fig. 3 is a flowchart of calculating a first adjustment amount in the present invention.

Fig. 4 is a flowchart of calculating the second adjustment amount in the present invention.

In the figure, 1 is a projector, 2 is a camera, 3 is a print head, 4 is a console, 5 is a rigid frame, and 6 is a current off-line planned trajectory Γ_planAnd 7 is a transition molten pool P_F8 is a local fresh noodle S₁And 9 is a partial substrate surface S₂And 10 is an on-line adjusted trajectory gamma_online。

Detailed Description

The invention provides a metal additive manufacturing online track adjusting method based on real-time three-dimensional detection, and the invention is further described in detail below with reference to the accompanying drawings and specific embodiments. The following examples are intended to illustrate the invention, but are not intended to limit the scope of the invention.

The invention provides a metal additive manufacturing online track adjusting method based on real-time three-dimensional detection, and the overall flow is as shown in figure 1, and the method comprises the following steps:

s1: installing a real-time three-dimensional detection device;

the method comprises the steps that a real-time three-dimensional detection device is arranged at a reasonable position in a transmission device of a printing head of metal additive manufacturing equipment or a printer, so that the real-time three-dimensional detection device is ensured not to interfere with a normal additive manufacturing process, and the additive manufacturing process also does not interfere with the detection of the real-time three-dimensional detection device, on the basis, the detection range of the real-time three-dimensional detection device is ensured to cover the lower part and the peripheral area of the printing head, and the printing head is avoided;

in the present invention, the metal additive manufacturing equipment includes, but is not limited to, various types of additive manufacturing equipment based on existing mature technologies, such as laser, electron beam, plasma, electric arc, etc., or based on new technologies that may emerge in the future. The real-time three-dimensional detection device includes, but is not limited to, various real-time three-dimensional detection devices based on existing mature technologies, such as an extended stereo vision method, a passive stereo vision method, a time-of-flight method, a defocusing method, a structured light projection method, and the like, or based on new technologies that may appear in the future. The peripheral area generally refers to a rectangular area with a printing point of additive manufacturing equipment as a center, wherein the long edge of the rectangle is parallel to a printing track, the short edge of the rectangle is perpendicular to the printing track, the length of the short edge of the rectangle is generally 5-10 cm, the rest distances are also available, the length of the long edge depends on the balance of precision and speed, if the long edge is longer, the track adjustment precision is higher, but the printing time is slowed down due to more processing data, otherwise, the printing time is shortened, but the track adjustment precision is reduced, and from the balance of the two, the distance is generally 15-25 cm, and the rest distances are also available.

As shown in fig. 2, in this embodiment, the real-time three-dimensional detection device selects a detection device based on a structured light projection method, which includes: the projector comprises a projector 1, a camera 2 and a computer, wherein the projector 1 and the camera 2 are respectively connected with the computer. The metal additive manufacturing equipment adopts an electron beam fuse deposition equipment EBAM 110 system of Sciaky company, and comprises a printing head 3, an operating platform 4, an industrial personal computer and the like, wherein the printing head 3 and the operating platform 4 are respectively connected with the industrial personal computer. And the computer of the real-time three-dimensional detection device is connected with the industrial personal computer of the metal additive manufacturing equipment, and the computer is used for controlling the real-time three-dimensional detection device and the metal additive manufacturing equipment respectively. The projector 1 and the camera 2 are respectively installed on two sides of the printing head 3, the main shafts of the projector 1, the camera 2 and the printing head are fixedly connected through bolts by using a steel rigid frame 5, a rectangular area with a detection range of 10cm width and 20cm length covers the area below and nearby the printing head, and the long side of the rectangle is parallel to the advancing track of the printing head 3.

S2: planning a printing track off line;

carrying out three-dimensional modeling on a metal component to be manufactured to obtain a three-dimensional model of the metal component, then layering the three-dimensional model to obtain a corresponding layering result, setting a growth direction and printing parameters, planning to obtain an initial off-line printing track, and carrying out off-line printing on the initial off-line printing trackThe initial offline printing trajectory is taken as the current offline planning trajectory gamma_plan；

In this embodiment, the three-dimensional modeling software is Solidworks, and operations such as layering and planning are executed by software built in the EBAM 110 system industrial personal computer. Current offline planned trajectory Γ_plan6 as shown in fig. 2, depicted using dashed lines, the arrows point in the direction of travel.

S3: planning a trajectory gamma according to the current off-line_planPerforming additive manufacturing on a metal component to be manufactured, and planning a trajectory gamma according to the current off-line_planDuring the manufacturing process, the corresponding transition weld pool P is formed in the printing head region_FTransition molten pool P_FLocal new surfaces S are respectively formed around₁And S₁Local substrate surface S formed after delay time tau₂；

As shown in FIG. 2, in this embodiment, the printhead 3 follows a current offline planned trajectory Γ_planThe corresponding manufacturing process is run and executed 6, forming a transition melt pool 7 (as indicated by the circled portion below 3 in fig. 2), a local fresh surface 5 (as indicated by the diamond-shaped grid portion in fig. 2) and a local substrate surface 6 (as indicated by the square grid portion in fig. 2) on the operation table 4.

S4: acquisition of transition molten pool P using real-time three-dimensional detection device_FLocal fresh noodle S₁And local substrate surface S₂Three-dimensional shape information of;

in the present invention, three-dimensional topography information refers to all expression forms capable of representing information of an object in a three-dimensional space, including but not limited to three-dimensional depth maps, three-dimensional voxel volumes, three-dimensional polygonal meshes, and three-dimensional point clouds.

The embodiment uses a structured light four-step phase shift method and three-dimensional point cloud to transition molten pool P_FLocal fresh noodle S₁And local substrate surface S₂Measuring and expressing, projecting different coding patterns to the detection range by the projector 1, collecting the coding patterns by the camera 2, and processing to obtain a molten pool P containing transition_F7. Local new noodle S₁8 and partial substrate surface S₂9, further based on height and normal directionProcessing the segmentation point cloud to obtain a transition molten pool P_F7. Local new noodle S₁8 and partial substrate surface S₂9 independent three-dimensional point clouds.

S5: transition weld pool P collected according to S4_FAnd local substrate surface S₂Establishing and calculating a first adjustment quantity delta gamma₁Model1 of (1);

the overall flow of the steps is shown in fig. 3, and the specific steps are as follows:

s51, defining the current off-line planning trajectory gamma_planOn the local substrate surface S₂Each point on the table is a current off-line planning track point p_i，p_iNeighborhood of (2)From point p_tComposition of p_tNormal to the surface of n_tThen p is_iNeighborhood A of_piIs defined as:

wherein, | | p_t-p_iI represents the point p_tAnd point p_iOn the local substrate surface S₂A distance of (d) is a set point p_tAnd point p_iOn the local substrate surface S₂Should be less than p_tWith local substrate surface S₂The maximum distance of the boundary;

partial substrate surface S acquired in step S4₂Using a surface processing algorithm to calculate each p according to the specified delta_iCorresponding p_tAnd n_tWherein p is_tAnd n_tIn a one-to-one correspondence, and for repeated p_tOnly once calculated; s52: the transition zone molten pool P collected according to the step S4_FExtracting the molten pool P of the transition region according to a shape extraction algorithm_FIncluding the center of the molten poolLong shaftShort shaftAnd the angle of inclination

S53: according to a single droplet structure model M_sbModel M with multiple molten drops_mbAll p obtained in step S51_tCorresponding surface normal n_tMajor axis of molten pool obtained in step S52Short shaftAnd the angle of inclinationCalculating a first attitude correction amount of the print headWhereinFirst rotation angle correction quantities of the printing head in the x, y and z directions respectively;

s54: according to a single droplet structure model M_sbModel M with multiple molten drops_mbAll p obtained in step S52_tAnd the center of the molten pool obtained in step S52Calculating a first displacement correction amount Deltav of the print head₁＝[Δx₁,Δy₁,Δz₁]^TWherein Δ x₁,Δy₁,Δz₁In the x, y, z directions of the print head, respectivelyA first amount of displacement correction.

S55: the first posture correction amount Δ ω obtained in step S53₁And the first displacement correction amount Δ v obtained in step S54₁Establishing and calculating a first adjustment quantity delta gamma₁Model1, and calculates a first adjustment amount Δ Γ₁Wherein the Model1 is:

in the invention, a curved surface processing algorithm and a morphology extraction algorithm refer to algorithms capable of achieving the aim of extracting information, and the algorithms are influenced by the three-dimensional information expression form, for example, when a three-dimensional depth map is adopted to express morphology information, a shadow recovery method and a stereo matching method can be adopted, and when three-dimensional point cloud is adopted to express morphology information, an OPA algorithm, a neural network method and the like can be adopted.

In this embodiment, δ is defined to be 0.01m, and the local substrate surface S is processed using the tensor voting frame method in cooperation with the tensor space surface normal estimation method₂6 point cloud space, calculating point p_tAnd its surface normal n_t. Then from the transition bath P_FThe three-dimensional point cloud is calculated and obtained by using a contour interactive drawing algorithmLong shaftShort shaftAnd the angle of inclinationEach binding n_tAnd p_tAccording to a single droplet structure model M_sbAnd a multiple droplet overlay model M_mbCalculating a first attitude correctionAnd a first displacement correction amount Deltav₁＝[Δx₁,Δy₁,Δz₁]^T. Finally, a first adjustment amount delta gamma is calculated according to the first posture correction amount and the first displacement correction amount and the Model1₁。

S6: local fresh surface S collected according to S4₁Establishing and calculating a second adjustment quantity delta gamma₂Model2 of (1).

The overall flow of the steps is shown in fig. 4, and the specific steps are as follows:

s61: local new surface S collected according to step S4₁Three-dimensional topography information of S₁As the current local new surface, based on the Markov process, a probability model for estimating the local new surface at the next moment according to the current local new surface is established to obtain the local new surface at the next momentThree-dimensional topography information;

s62, extracting the expected local new surface at the next moment of the current printing head position according to the three-dimensional model of the metal member and the layering result obtained in the step S2Three-dimensional shape information of;

s63: the local new noodle at the next momentAnd the next moment when local new noodles are expectedComparing to obtain a difference value

S64: according to Δ S₁Single-molten-drop structure model M_sbAnd a multiple droplet overlay model M_mbCalculating a second attitude correction amount of the print headAnd a second displacement correction amount Deltav₂＝[Δx₂,Δy₂,Δz₂]^TWhereinSecond angular corrections, Δ x, of the print head in the x, y, z directions, respectively₂,Δy₂,Δz₂Second displacement correction amounts of the printing head in x, y and z directions respectively;

s65: according to Δ S₁、Δω₂And Δ v₂Establishing and calculating the second adjustment quantity delta gamma₂Model2, and calculating a second adjustment amount Δ Γ₂Wherein the Model2 is:

in this embodiment, a Markov process is first utilized, based on the local new surface S₁Estimating the local new surface of the next moment by the three-dimensional point cloudAnd the expectation of local new surfaceComparing the three-dimensional point clouds to obtain difference point cloudsAccording to a single droplet structure model M_sbAnd a multiple droplet overlay model M_mbIn combination with Δ S₁Calculating a second attitude correctionAnd a second displacement correction amount Deltav₂＝[Δx₂,Δy₂,Δz₂]^T. Finally, the second attitude correction and the second displacement correction are calculated according to the Model2Calculating to obtain a second adjustment quantity delta gamma₂。

S7: for the first adjustment quantity delta gamma₁And a second adjustment amount DeltaGamma₂And superposing to obtain a total track adjustment quantity delta gamma, and then compensating the delta gamma to the current offline planning track gamma_planObtaining the track gamma after on-line adjustment_onlineThe specific calculation method comprises the following steps:

wherein the content of the first and second substances,indicating a superposition compensation operation.

In this embodiment, first, the first adjustment amount Δ Γ obtained in step S5 is adjusted₁The second adjustment amount Δ Γ obtained in step S6 is compensated for by superposition₂Obtaining the total track adjustment quantity delta gamma, and then superposing and compensating the total track adjustment quantity delta gamma to the offline track gamma_plan6 (dotted line, arrow indicating print advance direction), the trajectory Γ after the in-line adjustment is obtained_online10 (indicated by a dot-dash line, an arrow indicating the direction of travel of the print), which is a line-off trajectory Γ_plan6 different print trajectories.

S8: gamma-gamma is formed_onlineAs new gamma_planAnd then returning to the step S3 again to perform a new round of off-line planned trajectory adjustment until the whole metal member to be manufactured is manufactured.