Laser welding wire feeding control method for robot servo control device

Abstract

The invention discloses a laser welding wire feeding control method for a robot servo control device, which relates to the field of welding data management, and comprises the following steps of: acquiring the operation record of laser welding wire feeding equipment and the characteristics of a keyhole wall in a unit detection period, and forming a target period group by two adjacent unit detection periods; acquiring function features of operation parameters of the laser welding wire feeding equipment in the target period groups, acquiring change values of features of key hole walls in the target period groups, acquiring detection records of welding seams in the range of the target period groups, collecting the detection records corresponding to the plurality of target period groups to form a detection record set, acquiring image features of the surfaces of the welding seams, and displaying the image features. Calculating a surface evaluation value in the weld joint surface image, collecting detection records of abnormities in the weld joint, calculating an internal evaluation value of the weld joint according to the proportion of abnormal areas in the weld joint, extracting the surface evaluation value of the target period group and the characteristics of the internal evaluation value, and establishing a quality evaluation vector of the target period group.

Description

Technical Field

[0001] The invention relates to the field of welding data management, in particular to a laser welding wire feeding control method of a robot servo control device. Background Art

[0002] Laser welding is a welding method that uses a high-energy-density laser beam as a heat source. It is an efficient and precise welding technology. During the laser welding process, high-temperature, high-pressure gases are generated inside the molten pool. These gases solidify during solidification to form metal pores. If the pore density is too high, keyhole defects will form.

[0003] Keyhole defects are a manifestation of substandard welding quality, which will affect the overall quality of welded components. Keyhole defects reduce the effective working cross-section of the weld, causing the weld's bearing capacity to decrease. When subjected to external forces, stress concentration is likely to occur around the keyhole, causing the weld to more easily break.

[0004] When the wire feed speed is too low, the filled liquid metal may not be enough to fully wet and fill the weld, resulting in some areas in the weld not being completely melted or poorly fused. This poor fusion may increase the risk of keyhole defects because cavities or gaps may form in the unmelted metal areas; when the wire feed speed is too high, too much liquid metal may compress the laser keyhole, causing the keyhole to be blocked or deformed. In this case, the pressure inside the keyhole may increase, making keyhole defects more likely to form. Summary of the invention

[0005] The object of the present invention is to provide a laser welding wire feeding control method of a robot servo control device to solve the problems raised in the prior art.

[0006] To achieve the above object, the present invention provides the following technical solutions: Step S100: acquiring data features of operating parameters of the laser welding wire feeding device within a unit time period, recording a combination of one or at least two data features as a first feature value, and collecting a change value of a feature of a keyhole wall within a target period group, recording it as a second feature value; Step S200: obtaining the inspection records of the weld within a unit time period, collecting the abnormal records on the weld surface and the abnormal records inside the weld in the inspection records, and calculating the abnormal evaluation value according to the proportion of the abnormal area; Step S300: extracting the average value of the abnormality on the weld surface and the abnormality evaluation value inside the weld respectively, and establishing a quality evaluation vector within a unit time period; Step S400: The first eigenvalue, the second eigenvalue and the quality evaluation vector of the target period group are collected to form an evaluation sequence corresponding to the target period group, and several evaluation sequences are obtained. The evaluation vectors are clustered according to the locations where the weld anomalies occur, and the cluster core of each cluster is obtained. Step S500: collecting the operation records of the laser welding wire feeding device and the change values ​​of the characteristics of the keyhole wall in the current time period, calculating the predicted value of the welding result, and establishing a quality prediction vector corresponding to the predicted value; Step S600: Calculate the distance between the quality prediction vector and the cluster core of the cluster respectively, and sort the inspection items of the current weld according to the distance.

[0007] Furthermore, step S100 includes: Step S101: setting a unit detection cycle, obtaining the operation record of the laser welding wire feeding device and the characteristics of the keyhole wall in the unit detection cycle, forming a target cycle group with two adjacent unit detection cycles, obtaining the operation parameters of the laser welding wire feeding device in the target cycle group, plotting the operation function of the operation parameters changing with time, and recording the data characteristics of the operation function as the first characteristic value; Step S102: Representing the three-dimensional morphological features of the keyhole wall by i spatial feature vectors, obtaining i spatial feature vectors of the keyhole wall in the first unit detection cycle in the target cycle group, and grouping the spatial feature vectors into a first vector cluster; Step S103: Obtaining i spatial feature vectors of the keyhole wall in the second unit detection cycle in the target cycle group, and grouping the spatial feature vectors into a second vector cluster; Step S104: Record the difference between the second vector cluster and the first vector cluster as the second eigenvalue.

[0008] Further, step S200 includes: Step S201: obtaining an image feature extraction model, and training a recognition model for abnormal weld surface features through images of normal weld surfaces and images of abnormal weld surfaces; Step S202: Collecting the detection records with abnormalities on the weld surface into a first group, and identifying images with normal weld surfaces in each detection record in the first group; Step S203: obtaining an image of the weld surface corresponding to the k1th detection record in the first group in the target period group, recording the area of ​​the obtained image as Dk1, identifying the partial image of the weld surface abnormality in the image, recording the area of ​​the obtained partial image as dk1; Step S204: Calculate the surface evaluation value αk1 of the k1th detection record in the first group. Among them, αk1= dk1 / Dk1.

[0009] Further, step S200 includes: Step S205: normal weld surfaces are grouped into a second group, and abnormal features of abnormal areas inside the weld are obtained, where the abnormal features include a corresponding space of the abnormal area or a two-dimensional feature slice of the corresponding space of the abnormal area; Step S206: Obtain all areas of the weld surface corresponding to the k2th detection record in the second group in the target period group, record the area as Vk2, obtain the partial area with abnormality in the k2th detection record, and record the obtained partial area as vk2, wherein, when the abnormal feature adopts the corresponding space of the abnormal area, Vk2 represents the volume inside the weld in the target period group, and vk2 represents the volume of the abnormal partial area; when the abnormal feature adopts the two-dimensional feature slice, Vk2 represents the area of ​​the two-dimensional feature slice image in the target period group, and vk2 represents the area of ​​the image corresponding to the abnormal partial area; Step S207: Calculate the internal evaluation value βk2 of the k2th detection record in the second group. Among them, βk2= vk2 / Vk2.

[0010] Furthermore, step S300 includes: Step S301: Calculate the surface evaluation value and internal evaluation value of each test record in the test record set. When the weld surface or the weld interior of a test record in the test record set is normal, the corresponding surface evaluation value or internal evaluation value is recorded as 0; Step S302: Obtain the surface evaluation value α of the nth detection record in the detection record set n and the internal evaluation value β n , calculate the comprehensive evaluation value γ of the nth detection record n , ; Step S303: Construct the quality evaluation vector e of the nth detection record n , e n (α n , β n , γ n ).

[0011] Furthermore, step S400 includes: Step S401: setting a first coordinate axis for measuring a first eigenvalue and a second coordinate axis for measuring a second eigenvalue, and establishing a management plane for a first evaluation set; Step S402: The first eigenvalue and the second eigenvalue of the target cycle group are combined into an eigenvalue pair, the position of the eigenvalue pair is marked in the management plane, and the quality evaluation vector of the target cycle group is matched with the position; By collecting historical data, the characteristic data of the welding process and the characteristic data of the welding results in the historical data are obtained, and a corresponding relationship between the two types of data is established. When the data of the current welding process is collected, the possible welding results are predicted.

[0012] Furthermore, step S400 includes: Step S403: Divide all evaluation vectors in the target cycle groups into three categories: abnormality on the weld surface, abnormality inside the weld, and abnormality on both the weld surface and inside the weld; Step S404: clustering the three types of evaluation vectors according to the coordinates in the vectors by using a clustering algorithm, and recording the three types of cluster cores as P1, P2 and P3, respectively, where P1 represents the cluster core with abnormalities on the weld surface, P2 represents the cluster core with abnormalities inside the weld, and P3 represents the cluster core with abnormalities on both the weld surface and inside the weld; Combined with step S301, when a weld has only one defect, the value of αn or βn is 0. In the three-dimensional space, the quality evaluation vector will be in one plane, and the two vectors with only one defect exist in different planes and are separated to a large extent. Therefore, after clustering, it is easy to distinguish the clustering cores of the three situations.

[0013] Furthermore, step S500 includes: Step S501: two adjacent unit detection cycles in the current laser welding process are combined into a current cycle group, and a first characteristic value m1 of the laser welding wire feeding device and a second characteristic value m2 of the keyhole wall in the current cycle group are collected; Step S502: m1 and m2 are combined into a target index f(m1, m2).

[0014] Furthermore, step S500 includes: Step S503: Indicate the index position of the target index in the management plane of the first evaluation set, set a distance threshold r, obtain all evaluation sequences in the first evaluation set whose distance from the index position is less than r, and merge the evaluation sequences into the reference sequence set; Step S504: Obtain the number of evaluation sequences in the reference sequence set, record it as n, and obtain the coordinates of the jth evaluation sequence in the reference sequence set in the management plane (q1 j ,q2 j ); Step S505: Calculate the distance lmj ​​between the jth evaluation sequence and the target index. , calculate the sum of the distances between n evaluation sequences and the target index and record it as Ln; Step S506: Calculate the weight uj of the j-th evaluation sequence, uj=1-(lmj / Ln), and obtain the evaluation vector E in the j-th evaluation sequencej ; Step S507: Gather the evaluation vectors of the weights of all evaluation sequences in the reference sequence set, and calculate the quality prediction vector H corresponding to the target index. .

[0015] Furthermore, step S600 includes: Step S601: Calculate the distances between the quality prediction vector and P1, P2 and P3 respectively, wherein the distance between the quality prediction vector and P1 is recorded as ph1, the distance between the quality prediction vector and P2 is recorded as ph2, and the distance between the quality prediction vector and P3 is recorded as ph3; Step S602: setting a threshold value ω, when ph3 ≥ ω, prompting relevant staff to monitor the weld corresponding to the current cycle group; Step S603: When ph1>ph2, the weld surface corresponding to the current cycle group is monitored first; when ph2>ph1, the weld interior corresponding to the current cycle group is monitored first.

[0016] Compared with the prior art, the beneficial effects of the present invention are as follows: by collecting the historical operation records of the laser welding wire feeding equipment and the quality of the corresponding welding products, the association between the operation status of different equipment and the quality of welding products is summarized. First, a correlation model between the operation of the laser welding wire feeding equipment and the keyhole change and the quality of welding products is established. When the operation data of the laser welding wire feeding equipment is detected, the possible welding quality is predicted. Furthermore, a correlation model between welding quality and the order of delivery is established, which further explores the association between the operation characteristics of the laser welding wire feeding equipment and the types of specific quality problems generated, thereby improving the inspection efficiency of product quality problems. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 The present invention is a flowchart of a method for controlling wire feeding of laser welding by a robot servo control device. DETAILED DESCRIPTION

[0018] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0019] Example: Figure 1 As shown, the present invention provides a technical solution, a robot servo control device laser welding wire feeding control method: S1: setting a unit detection cycle, obtaining the operation record of the laser welding wire feeding device and the characteristics of the keyhole wall in the unit detection cycle, forming a target cycle group with two adjacent unit detection cycles, obtaining the function characteristics of the operation parameters of the laser welding wire feeding device in the target cycle group, recording them as the first characteristic value, and collecting the change value of the characteristics of the keyhole wall in the target cycle group, recording them as the second characteristic value; Laser welding wire feeder is a device used to automatically or manually control the feeding of welding wire into the weld during laser welding; In the embodiment, when obtaining the operating parameters of the laser welding wire feeder, the optional operating parameter items include: laser power, welding speed and wire feeding speed; In the embodiment, the characteristics of the function of the operating parameter changing with time are used as the first characteristic value, including the time series characteristics of the operating parameter, such as the changing trend characteristics of the operating parameter, such as the increase or decrease of the function, the periodic change area of ​​the function, such as the change frequency or the change period, the statistical characteristics of the operating parameter, such as the mean, variance or standard deviation, such as the waveform characteristics of the function, such as the waveform of the operating parameter changing with time, the correlation characteristics of the function, such as the autocorrelation characteristics of the function and the cross-correlation characteristics of the function; Among them, S1 includes: S11: obtaining operating parameters of the laser welding wire feeding device in the target cycle group, drawing an operating function of the operating parameters changing with time, and recording a function characteristic of the operating function as a first characteristic value; S12: Representing the three-dimensional morphological features of the keyhole wall by i spatial feature vectors, obtaining i spatial feature vectors of the keyhole wall in the first unit detection cycle in the target cycle group, and grouping the spatial feature vectors into a first vector cluster; S13: Obtaining i spatial feature vectors of the keyhole wall in the second unit detection cycle in the target cycle group, and grouping the spatial feature vectors into a second vector cluster; S14: Record the difference between the second vector cluster and the first vector cluster as the second eigenvalue; In an embodiment, the difference between two vector clusters may be calculated by comparing the vectors in the vector clusters one by one, or the first vector cluster and the second vector cluster may be respectively formed into i-order matrices, and the second eigenvalue may be obtained by comparing the difference values ​​of the two i-order matrices; Methods for comparing matrix differences include: Frobenius norm (F-norm), Manhattan distance and Euclidean distance.

[0020] S2: Obtain the inspection records of the weld within the target cycle group, collect the inspection records corresponding to several target cycle groups to form an inspection record set, obtain the image features of the weld surface, calculate the ratio of the abnormal part in the weld surface image to the normal image of the weld surface to obtain the surface evaluation value; Among them, S2 includes: S21: obtaining an image feature extraction model, and training a recognition model for abnormal weld surface features through images of normal weld surfaces and images of abnormal weld surfaces; S22: Collecting the detection records with abnormal weld surface into a first group, and identifying images with normal weld surface in each detection record in the first group; S23: obtaining an image of the weld surface corresponding to the k1th detection record in the first group in the target period group, recording the area of ​​the obtained image as Dk1, identifying a partial image of the weld surface abnormality in the image, recording the area of ​​the obtained partial image as dk1; S24: Calculate the surface evaluation value αk1 of the k1th detection record in the first group, Among them, αk1= dk1 / Dk1.

[0021] S3: Collect the detection records of abnormalities inside the weld, and calculate the internal evaluation value of the weld according to the proportion of the abnormal area inside the weld; Among them, S3 includes: S31: The normal weld surface is grouped into a second group, and abnormal features of the abnormal area inside the weld are obtained, where the abnormal features include a corresponding space of the abnormal area or a two-dimensional feature slice of the corresponding space of the abnormal area; S32: Obtain all areas of the weld surface corresponding to the k2th detection record in the second group in the target period group, record the area as Vk2, obtain the partial area with abnormality in the k2th detection record, and record the obtained partial area as vk2, wherein, when the abnormal feature adopts the corresponding space of the abnormal area, Vk2 represents the volume inside the weld in the target period group, and vk2 represents the volume of the abnormal partial area; when the abnormal feature adopts the two-dimensional feature slice, Vk2 represents the area of ​​the two-dimensional feature slice image in the target period group, and vk2 represents the area of ​​the image corresponding to the abnormal partial area; S33: Calculate the internal evaluation value βk2 of the k2th detection record in the second group, Among them, βk2= vk2 / Vk2.

[0022] S4: extracting the features of the surface evaluation value and the internal evaluation value of the target periodic group, and establishing the quality evaluation vector of the target periodic group; Among them, S4 includes: S41: Calculate the surface evaluation value and the internal evaluation value of each test record in the test record set. When the weld surface or the weld interior of a test record in the test record set is normal, the corresponding surface evaluation value or internal evaluation value is recorded as 0; S42: Obtain the surface evaluation value αn and the internal evaluation value βn of the nth detection record in the detection record set, and calculate the comprehensive evaluation value γn of the nth detection record. ; S43: Construct the quality evaluation vector en, en (αn, βn, γn) of the nth detection record.

[0023] S5: collecting the first eigenvalue, the second eigenvalue and the quality evaluation vector of the target period group to form an evaluation sequence corresponding to the target period group, and obtaining evaluation sequences corresponding to several target period groups to form a first evaluation set; Among them, S5 includes: S51: Setting a first coordinate axis for measuring a first eigenvalue and a second coordinate axis for measuring a second eigenvalue, and establishing a management plane for a first evaluation set; S52: The first eigenvalue and the second eigenvalue of the target periodic group are combined into an eigenvalue pair, the position of the eigenvalue pair is marked in the management plane, and the quality evaluation vector of the target periodic group is matched with the position.

[0024] S6: Obtain all evaluation vectors in several target cycle groups, and cluster the evaluation vectors according to the presence of abnormalities on the weld surface, the presence of abnormalities inside the weld, and the presence of abnormalities on both the weld surface and inside the weld, to obtain cluster cores of the three clusters; Among them, S6 includes: S61: Divide all evaluation vectors in the target cycle groups into three categories: abnormality on the weld surface, abnormality inside the weld, and abnormality on both the weld surface and inside the weld; S62: Through the clustering algorithm, the three types of evaluation vectors are clustered according to the coordinates in the vector, and the three types of cluster cores are recorded as P1, P2 and P3 respectively, among which P1 represents the cluster core with abnormalities on the weld surface, P2 represents the cluster core with abnormalities inside the weld, and P3 represents the cluster core with abnormalities on both the weld surface and the weld.

[0025] S7: collecting the operating parameters of the current laser welding wire feeding device and the characteristics of the keyhole wall, forming a current cycle group with two adjacent unit detection cycles, obtaining the first characteristic value and the second characteristic value in the current cycle group, and forming a target index with the first characteristic value and the second characteristic value; Among them, S7 includes: S71: two adjacent unit detection cycles in the current laser welding process are combined into a current cycle group, and a first characteristic value m1 of the laser welding wire feeding device and a second characteristic value m2 of the keyhole wall in the current cycle group are collected; S72: Combine m1 and m2 to form a target index f(m1, m2).

[0026] S8: Map the target index to the first evaluation set, obtain the evaluation sequence within the distance threshold range from the target index, calculate the weight of the target index and the evaluation sequence according to the distance, perform weighted operation on the quality evaluation vector in the evaluation sequence, and obtain the quality prediction vector; Among them, S8 includes: S81: Indicate the index position of the target index in the management plane of the first evaluation set, set a distance threshold r, obtain all evaluation sequences in the first evaluation set whose distance from the index position is less than r, and merge the evaluation sequences into the reference sequence set; S82: Obtain the number of evaluation sequences in the reference sequence set, record it as n, and obtain the coordinates (q1j, q2j) of the jth evaluation sequence in the reference sequence set in the management plane; S83: Calculate the distance lmj ​​between the jth evaluation sequence and the target index. , calculate the sum of the distances between n evaluation sequences and the target index and record it as Ln; S84: Calculate the weight uj of the j-th evaluation sequence, uj=1-(lmj / Ln), and obtain the evaluation vector Ej in the j-th evaluation sequence; S85: Gather the evaluation vectors of the weights of all evaluation sequences in the reference sequence set, and calculate the quality prediction vector H corresponding to the target index. .

[0027] S9: Calculate the distances between the quality prediction vector and the cluster cores of the three clusters respectively, and sort the inspection items of the current weld according to the distances.

[0028] Among them, S9 includes: S91: respectively calculating the distances between the quality prediction vector and P1, P2 and P3, wherein the distance between the quality prediction vector and P1 is recorded as ph1, the distance between the quality prediction vector and P2 is recorded as ph2, and the distance between the quality prediction vector and P3 is recorded as ph3; S92: Setting a threshold value ω, when ph3 ≥ ω, prompting relevant staff to monitor the weld corresponding to the current cycle group; S93: When ph1>ph2, the weld surface corresponding to the current cycle group is monitored first; when ph2>ph1, the weld interior corresponding to the current cycle group is monitored first.

[0029] It will be apparent to those skilled in the art that the invention is not limited to the details of the exemplary embodiments described above and that the invention can be implemented in other specific forms without departing from the spirit or essential features of the invention. Therefore, the embodiments should be considered exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description, and it is intended that all variations falling within the meaning and scope of the equivalent elements of the claims be included in the invention. Any reference numeral in a claim should not be considered as limiting the claim to which it relates.

Claims

1. A method for controlling wire feeding of laser welding by a robot servo control device, characterized in that: The method comprises the following steps: Step S100: acquiring data features of operating parameters of a laser welding wire feeding device within a unit time period, recording a combination of one or at least two data features as a first feature value, and collecting a change value of a feature of a keyhole wall within the target period group, recording it as a second feature value; Step S200: obtaining the inspection records of the weld within a unit time period, collecting the abnormal records on the weld surface and the abnormal records inside the weld in the inspection records, and calculating the abnormal evaluation value according to the proportion of the abnormal area; Step S300: extracting the average value of the abnormality on the weld surface and the abnormality evaluation value inside the weld respectively, and establishing a quality evaluation vector within a unit time period; Step S400: The first eigenvalue, the second eigenvalue and the quality evaluation vector of the target period group are collected to form an evaluation sequence corresponding to the target period group, and several evaluation sequences are obtained. The evaluation vectors are clustered according to the locations where the weld anomalies occur, and the cluster core of each cluster is obtained. Step S500: collecting the operation records of the laser welding wire feeding device and the change values ​​of the characteristics of the keyhole wall in the current time period, calculating the predicted value of the welding result, and establishing a quality prediction vector corresponding to the predicted value; Step S600: Calculate the distance between the quality prediction vector and the cluster core of the cluster respectively, and sort the inspection items of the current weld according to the distance.

2. The method for controlling laser welding wire feeding by a robot servo control device according to claim 1, characterized in that: Step S100 includes: Step S101: setting a unit detection cycle, obtaining the operation record of the laser welding wire feeding device and the characteristics of the keyhole wall in the unit detection cycle, forming a target cycle group with two adjacent unit detection cycles, obtaining the operation parameters of the laser welding wire feeding device in the target cycle group, plotting the operation function of the operation parameters changing with time, and recording the data characteristics of the operation function as the first characteristic value; Step S102: Representing the three-dimensional morphological features of the keyhole wall by i spatial feature vectors, obtaining i spatial feature vectors of the keyhole wall in the first unit detection cycle in the target cycle group, and grouping the spatial feature vectors into a first vector cluster; Step S103: Obtaining i spatial feature vectors of the keyhole wall in the second unit detection cycle in the target cycle group, and grouping the spatial feature vectors into a second vector cluster; Step S104: Record the difference between the second vector cluster and the first vector cluster as the second eigenvalue.

3. The method for controlling laser welding wire feeding of a robot servo control device according to claim 2, characterized in that: The method for calculating the abnormal evaluation value of the weld surface in step S200 includes: Step S201: obtaining an image feature extraction model, and training a recognition model for abnormal weld surface features through images of normal weld surfaces and images of abnormal weld surfaces; Step S202: Collecting the detection records with abnormalities on the weld surface into a first group, and identifying images with normal weld surfaces in each detection record in the first group; Step S203: obtaining an image of the weld surface corresponding to the k1th detection record in the first group in the target period group, obtaining an area of ​​the image as Dk1, identifying a partial image of the weld surface abnormality in the image, obtaining an area of ​​the partial image as dk1; Step S204: Calculate the surface evaluation value αk1 of the k1th detection record in the first group. Among them, αk1= dk1 / Dk1.

4. The method for controlling laser welding wire feeding by a robot servo control device according to claim 3, characterized in that: The method for calculating the abnormal evaluation value inside the weld in step S200 includes: Step S205: normal weld surfaces are grouped into a second group, and abnormal features of abnormal areas inside the weld are obtained, where the abnormal features include a corresponding space of the abnormal area or a two-dimensional feature slice of the corresponding space of the abnormal area; Step S206: Obtain all areas of the weld surface corresponding to the k2th detection record in the second group corresponding to the target period group, record the area as Vk2, obtain the partial area with abnormality in the k2th detection record, obtain the partial area and record it as vk2, wherein, when the abnormal feature adopts the corresponding space of the abnormal area, Vk2 represents the volume inside the weld in the target period group, and vk2 represents the volume of the abnormal partial area, and when the abnormal feature adopts the two-dimensional feature slice, Vk2 represents the area of ​​the two-dimensional feature slice image in the target period group, and vk2 represents the area of ​​the image corresponding to the abnormal partial area; Step S207: Calculate the internal evaluation value βk2 of the k2th detection record in the second group, Among them, βk2= vk2 / Vk2.

5. The method for controlling laser welding wire feeding of a robot servo control device according to claim 4, characterized in that: Step S300 includes: Step S301: Calculate the surface evaluation value and internal evaluation value of each test record in the test record set. When the weld surface or the weld interior of a test record in the test record set is normal, the corresponding surface evaluation value or internal evaluation value is recorded as 0; Step S302: Obtain the surface evaluation value α of the nth detection record in the detection record set n and the internal evaluation value β n , calculate the comprehensive evaluation value γ of the nth detection record n , ; Step S303: Construct the quality evaluation vector e of the nth detection record n , e n (α n , β n , γ n ).

6. A method for controlling laser welding wire feeding by a robot servo control device according to claim 5, characterized in that: The method of constructing the evaluation sequence in step S400 includes: Step S401: setting a first coordinate axis for measuring a first eigenvalue and a second coordinate axis for measuring a second eigenvalue, and establishing a management plane for a first evaluation set; Step S402: The first eigenvalue and the second eigenvalue of the target cycle group are combined into an eigenvalue pair, the position of the eigenvalue pair is marked in the management plane, and the quality evaluation vector of the target cycle group is matched with the position.

7. A method for controlling laser welding wire feeding by a robot servo control device according to claim 6, characterized in that: The cluster analysis method in step S400 includes: Step S403: Divide all evaluation vectors in the target period groups into three categories: abnormality on the weld surface, abnormality inside the weld, and abnormality on both the weld surface and inside the weld; Step S404: The three types of evaluation vectors are clustered according to the coordinates in the vectors through a clustering algorithm, and the three types of cluster cores are respectively recorded as P1, P2 and P3, where P1 represents a cluster core with abnormalities on the weld surface, P2 represents a cluster core with abnormalities inside the weld, and P3 represents a cluster core with abnormalities on both the weld surface and inside the weld.

8. The method for controlling laser welding wire feeding of a robot servo control device according to claim 7, characterized in that: The processing method for collecting data in the current time period in step S500 includes: Step S501: two adjacent unit detection cycles in the current laser welding process are combined into a current cycle group, and a first characteristic value m1 of the laser welding wire feeding device and a second characteristic value m2 of the keyhole wall in the current cycle group are collected; Step S502: m1 and m2 are combined into a target index f(m1, m2).

9. The method for controlling laser welding wire feeding of a robot servo control device according to claim 8, characterized in that: The method of establishing the quality prediction vector corresponding to the prediction value in step S500 includes: Step S503: Indicate the index position of the target index in the management plane of the first evaluation set, set a distance threshold r, obtain all evaluation sequences in the first evaluation set whose distance from the index position is less than r, and merge the evaluation sequences into the reference sequence set; Step S504: Obtain the number of evaluation sequences in the reference sequence set, record it as n, and obtain the coordinates of the jth evaluation sequence in the reference sequence set in the management plane (q1 j ,q2 j ); Step S505: Calculate the distance lmj ​​between the jth evaluation sequence and the target index. , calculate the sum of the distances between n evaluation sequences and the target index and record it as Ln; Step S506: Calculate the weight uj of the j-th evaluation sequence, uj=1-(lmj / Ln), and obtain the evaluation vector E in the j-th evaluation sequence. j ; Step S507: Gather the evaluation vectors of the weights of all evaluation sequences in the reference sequence set, and calculate the quality prediction vector H corresponding to the target index. .

10. The method for controlling laser welding wire feeding of a robot servo control device according to claim 9, characterized in that: Step S600 includes: Step S601: Calculate the distances between the quality prediction vector and P1, P2 and P3 respectively, wherein the distance between the quality prediction vector and P1 is recorded as ph1, the distance between the quality prediction vector and P2 is recorded as ph2, and the distance between the quality prediction vector and P3 is recorded as ph3; Step S602: setting a threshold value ω, when ph3 ≥ ω, prompting relevant staff to monitor the weld corresponding to the current cycle group; Step S603: When ph1>ph2, the weld surface corresponding to the current cycle group is monitored first; when ph2>ph1, the weld interior corresponding to the current cycle group is monitored first.

Images