CN109014544B - Micro resistance spot welding quality on-line monitoring method - Google Patents
Micro resistance spot welding quality on-line monitoring method Download PDFInfo
- Publication number
- CN109014544B CN109014544B CN201810938371.2A CN201810938371A CN109014544B CN 109014544 B CN109014544 B CN 109014544B CN 201810938371 A CN201810938371 A CN 201810938371A CN 109014544 B CN109014544 B CN 109014544B
- Authority
- CN
- China
- Prior art keywords
- spot welding
- resistance spot
- quality
- monitoring
- photon
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B23—MACHINE TOOLS; METAL-WORKING NOT OTHERWISE PROVIDED FOR
- B23K—SOLDERING OR UNSOLDERING; WELDING; CLADDING OR PLATING BY SOLDERING OR WELDING; CUTTING BY APPLYING HEAT LOCALLY, e.g. FLAME CUTTING; WORKING BY LASER BEAM
- B23K11/00—Resistance welding; Severing by resistance heating
- B23K11/24—Electric supply or control circuits therefor
- B23K11/25—Monitoring devices
- B23K11/252—Monitoring devices using digital means
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B23—MACHINE TOOLS; METAL-WORKING NOT OTHERWISE PROVIDED FOR
- B23K—SOLDERING OR UNSOLDERING; WELDING; CLADDING OR PLATING BY SOLDERING OR WELDING; CUTTING BY APPLYING HEAT LOCALLY, e.g. FLAME CUTTING; WORKING BY LASER BEAM
- B23K11/00—Resistance welding; Severing by resistance heating
- B23K11/10—Spot welding; Stitch welding
- B23K11/11—Spot welding
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B23—MACHINE TOOLS; METAL-WORKING NOT OTHERWISE PROVIDED FOR
- B23K—SOLDERING OR UNSOLDERING; WELDING; CLADDING OR PLATING BY SOLDERING OR WELDING; CUTTING BY APPLYING HEAT LOCALLY, e.g. FLAME CUTTING; WORKING BY LASER BEAM
- B23K31/00—Processes relevant to this subclass, specially adapted for particular articles or purposes, but not covered by only one of the preceding main groups
- B23K31/12—Processes relevant to this subclass, specially adapted for particular articles or purposes, but not covered by only one of the preceding main groups relating to investigating the properties, e.g. the weldability, of materials
- B23K31/125—Weld quality monitoring
Abstract
The invention relates to a method for monitoring the quality of miniature resistance spot welding on line, belonging to the technical field of monitoring the quality of miniature resistance spot welding. Firstly, a single photon detector is utilized to monitor the change condition of the photon number of the contact edge of an electrode and a workpiece along with time in the production process of miniature resistance spot welding in real time, and a photon change curve is drawn. Photon characteristic quantities of the quality of the miniature resistance spot welding joints with different quality grades are extracted by a data mining means, a miniature resistance spot welding quality classification model is generated, a database of miniature resistance spot welding quality monitoring based on photon signals is established, and automatic online monitoring of the quality of the miniature resistance spot welding joints under different process conditions is realized. The method provides a new way for the on-line monitoring of the micro resistance spot welding quality, the used single-photon detector has high detection sensitivity to the change of the spot welding nugget temperature field, and the reliability of the on-line quality monitoring of the micro resistance spot welding is improved.
Description
Technical Field
The invention relates to the technical field of miniature resistance spot welding quality monitoring, in particular to an on-line monitoring method for miniature resistance spot welding quality. Photon change signals of a near nugget region in the micro resistance spot welding process are monitored, photon change characteristics of the near nugget region of different quality levels of the micro resistance spot welding joint are extracted through a data mining method, an online monitoring model of the quality of the micro resistance spot welding is generated, and a database of the quality monitoring of the micro resistance spot welding joint based on the photon signals is established.
Background
The micro resistance spot welding technology is widely applied to the fields of medical instruments, electronic equipment, battery packaging and the like, particularly the power battery packaging of the current new energy automobile, and the micro resistance spot welding technology in series-parallel connection of common lithium batteries is widely applied.
Due to the comprehensive action of multiple physical fields such as thermal power and the like in the spot welding production process, the quality of spot welding joints is unstable, particularly in micro resistance spot welding, the welding time is generally concentrated in 2-3ms, and the extremely short welding time provides higher requirements for the on-line monitoring of the quality of the micro resistance spot welding.
The micro resistance spot welding has the advantages that a workpiece is thin, the electrode displacement and the electrode force change are not obvious in the nugget forming process, and the anti-interference capability of the electrode displacement sensor and the electrode dynamic pressure sensor is poor, so that certain difficulty is brought to an electrode displacement monitoring method and an electrode force monitoring method.
The temperature field of the spot welding nugget area is closely related to the quality of a spot welding joint, and the existing on-line monitoring method of the infrared temperature field has poor response time and photon sensitivity of an infrared temperature monitor, so that the temperature information of the whole process of forming the miniature resistance spot welding nugget is difficult to capture.
Disclosure of Invention
The invention aims to provide an on-line monitoring method for the quality of miniature resistance spot welding, which solves the problems of the existing on-line monitoring technology for the quality of miniature resistance spot welding, establishes an on-line monitoring model for the quality of miniature resistance spot welding based on the photon signal characteristics of a near nugget region, and establishes a miniature resistance spot welding quality monitoring database by combining various miniature resistance spot welding process parameters.
The above object of the present invention is achieved by the following technical solutions:
the method for monitoring the quality of the miniature resistance spot welding on line comprises the steps of firstly utilizing a single-photon detector to monitor the change condition of the number of photons of the contact edge of an electrode and a workpiece along with time in the production process of the miniature resistance spot welding in real time, and drawing a photon change curve; determining the quality grades of the miniature resistance spot welding joints under different process conditions through a metallographic experiment, wherein photon change curves are different in the miniature resistance spot welding process of different qualities, extracting photon characteristic quantities of the miniature resistance spot welding qualities of different quality grades by using a data mining means, generating a miniature resistance spot welding quality classification model, establishing a database of miniature resistance spot welding quality monitoring based on photon signals, and realizing automatic online monitoring of the miniature resistance spot welding quality under different process conditions; the method specifically comprises the following steps:
selecting characteristic quantities of the photon signals;
step (2), analyzing the obtained characteristic quantities, and taking the threshold value of each characteristic quantity as an initial judgment basis;
and (3) establishing a micro resistance spot welding quality model according to the characteristic quantity about photon change obtained in the step (1), and monitoring the micro resistance spot welding quality by adopting a double-model cross validation method in order to ensure the monitoring accuracy of the micro resistance spot welding quality. In the process of establishing the micro resistance spot welding quality monitoring model, dividing an initially input micro resistance spot welding sample number N into two parts, namely a training sample set N1 and a testing sample set N2;
and (4) storing the judgment boundary obtained in the step (2) and the monitoring model established in the step (3) into a database of the miniature resistance spot welding monitoring system corresponding to the process parameters.
The selection of the characteristic quantity of the monitoring signal in the step (1) is as follows:
(1.1) accumulating the photon number in the time period from the beginning of the welding current to the end of the welding current to obtain the total photon number c in the welding time;
(1.2) carrying out data mining on a change curve of the number of photons c along with the welding time t, and extracting photon characteristic quantities reflecting the quality of the miniature resistance spot welding joint: respectively obtaining the photon number c of the welding current starting timestartGenerally corresponding to the minimum photon number, the maximum photon number cmaxThe number of turns m in the photon curve, and the number of photons c corresponding to each turntr1、ctr2、ctr3....ctrmTime t corresponding to two turnstr1、ttr2、ttr3....ttrmAnd the duration t of the number of balanced photonsb；
(1.3) respectively calculating the total number c of photons in each time within the current duration t and every 0.2ms time interval1、c2....cn-1、cnAs a feature amount, wherein n ═ t/0.2;
(1.4) formula according to variance
Calculating the variance characteristic quantity sigma of each micro resistance spot welding photon number2Wherein c is the number of photons which varies with the welding time t, μ is the average number of all photons from the start of the welding current to the end of the welding current, and n is the same as n in (1.3);
(1.5) testing common quality problems in the micro resistance spot welding process and monitoring photon signals in the process to obtain micro resistance spot welding joints with different qualities, determining the joint quality grades of the micro resistance spot welding samples through a metallographic test, and respectively marking the grades as Excellent 2, good 3, qualified 4 and unqualified 0;
(1.6) the quality classification grade in the step (1.5) is only used for classification in the micro resistance spot welding quality model establishing process, and the quality model combines the excellent 2, good 3 and qualified 4 categories into the grade 1 in the actual application process, namely the quality model belongs to the qualified spot welding joint 1, and the unqualified spot welding joint is 0;
and (1.7) performing correlation analysis on the characteristic quantities in the steps (1.1), (1.2), (1.3) and (1.4) and the four quality grades in the step (1.5), and removing part of linearly related characteristic quantities to obtain the characteristic quantities with larger correlation coefficients with the quality grades of the micro resistance spot welding joints.
Analyzing the obtained characteristic quantities, wherein the threshold values of the obtained characteristic quantities are used as an initial judgment basis, and the method specifically comprises the following steps:
(2.1) counting the probability density of each characteristic quantity of different micro resistance spot welding joint quality levels according to Gaussian distribution;
(2.2) according to different application occasions of the miniature resistance spot welding joint, eliminating probability density overlapping sections of all characteristic quantities of different miniature resistance spot welding joint quality levels, and determining a high density section with proper characteristic quantities to serve as an initial judgment threshold;
the quality grades are as follows: excellent 2, good 3, qualified 4, and unqualified 0.
Establishing a miniature resistance spot welding quality monitoring model in the step (3), specifically:
(3.1) establishing a miniature resistance spot welding joint quality monitoring model A based on multiple regression analysis and a least square method error criterion;
(3.1.1) utilizing N1 micro resistance spot welding samples to perform piecewise fitting on photon signals in the micro resistance spot welding process, and searching a fitting curve of the photon signals according to a least square method criterion;
(3.1.2) verifying the fitting curve in the step (3.1.1) based on the N2 micro resistance spot welding samples, and when the testing accuracy is higher than a set value, confirming a photon signal curve of a qualified spot welding sample;
(3.2) establishing a miniature resistance spot welding joint quality monitoring model B based on a combined nearest classifier KNN algorithm;
(3.2.1) utilizing N1 micro resistance spot welding samples, respectively using different KNN algorithm measurement criteria under the condition of determining the classification number, establishing a micro resistance spot welding monitoring model by combining characteristic quantity, verifying different measurement criteria models by using N2 micro resistance spot welding samples, and recording the monitoring results of each model on N2 micro resistance spot welding joints;
and (3.2.2) according to the verification result of the step (3.2.1), giving different weights to the classification models for the monitoring accuracy rates of different measurement criteria so as to optimize the accuracy rate of the miniature resistance spot welding monitoring model.
When the characteristic quantity of a new miniature resistance spot welding sample enters a system, firstly, judging whether all the characteristic quantities of the sample are within the initial judgment threshold interval in the step (2), and directly judging whether the spot welding joint belongs to a qualified joint 1 or an unqualified joint 0;
if all the characteristic quantities are within the initial judgment threshold interval obtained in the step (2), outputting an identifier 1 by the miniature resistance spot welding joint quality monitoring system, namely, indicating that the miniature resistance spot welding joint belongs to a qualified sample;
if the characteristic quantity of the sample is partially not in the initial judgment threshold interval in the step (2), judging the quality grade of the micro resistance spot welding sample by further using the micro resistance spot welding quality monitoring model established in the step (3) for the micro resistance spot welding sample; when the monitoring results of the miniature resistance spot welding joint quality monitoring models A, B are consistent, if the monitoring results of the miniature resistance spot welding joint quality monitoring models A, B all belong to qualified joints 1, the monitoring system automatically outputs 1, and if the monitoring results of the miniature resistance spot welding joint quality monitoring models A, B all belong to unqualified joints 0, the monitoring system automatically outputs 0;
when the monitoring results of the micro resistance spot welding joint quality monitoring model A, B are inconsistent, abnormal treatment needs to be given to the micro resistance spot welding joint, and an operator is reminded to use other detection methods or manually check whether the micro resistance spot welding joint meets the use requirements.
The invention has the beneficial effects that: a high-sensitivity single-photon detector with response time reaching nanosecond level is used for monitoring a micro resistance spot welding process with extremely short welding time, and the quality of a micro resistance spot welding joint is monitored according to the change rule of photons in a near nugget region along with the welding time. The photon number and the temperature change have correlation, the spot welding joint forming process and the temperature field are closely related, and the correlation provides an important theoretical basis for the invention. The invention provides a method for determining probability density space of micro resistance spot welding joints of different levels by Gaussian distribution and eliminating superposition space for determining an initial determination boundary aiming at the problem that the initial determination boundary can not be set according to an empirical value based on a characteristic quantity of spot welding quality of photon signals for the first time. The invention provides a method for ensuring the effective detection of unqualified welding spots by adopting a double-model cross validation method and establishing a micro resistance spot welding quality monitoring database.
Drawings
The accompanying drawings, which are included to provide a further understanding of the invention and are incorporated in and constitute a part of this application, illustrate embodiment(s) of the invention and together with the description serve to explain the invention without limiting the invention.
FIG. 1 is a system block flow diagram of the present invention;
FIG. 2 is an initialization interface of the micro resistance spot welding monitoring system of the present invention;
FIG. 3 is a working interface of the micro resistance spot welding monitoring system of the present invention.
Detailed Description
The details of the present invention and its embodiments are further described below with reference to the accompanying drawings.
Referring to fig. 1 to 3, in the method for monitoring the quality of the micro resistance spot welding on line, firstly, a single photon detector with response time reaching nanosecond level is used for monitoring the change condition of the photon number of the contact edge of an electrode and a workpiece along with time in the production process of the micro resistance spot welding in real time, and a photon change curve is drawn; the quality grades of the miniature resistance spot welding joints under different process conditions are determined through metallographic experiments, photon change reflects the change of a temperature field to a certain extent, photon change curves are different in the miniature resistance spot welding process of different qualities, photon characteristic quantities of the miniature resistance spot welding joint qualities of different quality grades are extracted by using a data mining means, a miniature resistance spot welding joint quality classification model is generated, a database of miniature resistance spot welding joint quality monitoring based on photon signals is established, and automatic online monitoring of the miniature resistance spot welding joint qualities under different process conditions is achieved. The invention provides a new way for the on-line monitoring of the micro resistance spot welding quality, the single photon detector used has high detection sensitivity to the change of the spot welding nugget temperature field, and the reliability of the on-line quality monitoring of the micro resistance spot welding is improved.
Referring to fig. 1 to 3, the online monitoring method for the micro resistance spot welding quality of the present invention comprises the following steps:
step (1), establishing a human-computer interaction information module: the method is used for searching whether the miniature resistance spot welding quality monitoring database has miniature resistance spot welding process parameters to be used or not;
step (2), as for the same spot welding process parameters existing in the database, only selection is needed;
and (3) if the identical technological parameters do not exist, inputting new micro resistance spot welding technological parameters: workpiece material 1, workpiece material 2, and thickness d of workpiece 1 plate1Thickness d of workpiece 2 plate2Selecting a welding current mode (Pc pulse current and Ac alternating current), current continuous electrifying time T, electrode pressure F, environment temperature T and the number N of samples of the micro resistance spot welding established by the initial model
Step (4), in the micro resistance spot welding process, acquiring photon signals in real time: detecting photon signals of a near nugget area by using a single photon detector with extremely quick response time, wherein the spectral response range of the single photon detector selects an infrared band;
step (5), in the process of monitoring the photon signal, the interference of external infrared light waves is avoided, the accuracy of monitoring the quality of the miniature resistance spot welding is influenced, and the interference of clutter on the monitoring result of the quality of the miniature resistance spot welding is reduced through the comprehensive action of three methods: filtering by using an infrared filter matched with the spectral band of the single photon detector in the monitoring device, filtering a photon signal obtained by the photon detector by using a filter circuit to filter frequency components irrelevant to the quality of the miniature resistance spot welding joint, and performing threshold filtering on singular point signals of photon data in a data processing process;
step (6), the change of the photon number detected by the single photon detector along with time can reflect the change of temperature, and a change curve of a photon signal along with the micro resistance spot welding time is drawn, wherein the time interval is from the beginning of welding current to the end of welding current;
a coefficient is introduced for the properties and surface states of different materials, and the relational expression between the radiation capacity, the wavelength and the temperature of an absolute black body can be known according to Planck's law:
M(λ，T)＝*C1λ-5/[exp(C2/λT)-1]
wherein λ -wavelength; t-absolute temperature; c1、C2-the radiation constant. According to the infrared temperature measurement principle, the change of the number of photons detected by the single photon detector along with time can reflect the change of temperature, and a change curve of the photons along with time is drawn, wherein the time interval is from the beginning of welding current to the end of the welding current.
Step (7), selecting characteristic quantity:
(7.1) the size of a nugget and the internal structure of the nugget of the miniature resistance spot welding joint are closely related to the temperature of resistance heat, and the photon number from the beginning of welding current to the end of welding current is accumulated to obtain the total photon number c in the welding time;
(7.2) carrying out data mining on a change curve of the photons c along with the welding duration t, and extracting photon characteristic quantities reflecting the quality of the miniature resistance spot welding joint: respectively obtaining the photon number c of the welding current starting timestartGenerally corresponding to the minimum photon number, the maximum photon number cmaxThe number of turns m in the photon curve, and the number of photons c corresponding to each turntr1、ctr2、ctr3....ctrmTime t corresponding to two turnstr1、ttr2、ttr3....ttrmAnd the duration t of the number of balanced photonsb；
(7.3) calculating the total number of photons c in each period of time within the current duration t and every 0.2ms time interval1、c2....cn-1、cnAs a feature amount, wherein n ═ t/0.2;
(7.4) formula according to variance
Calculating the variance characteristic quantity sigma of each micro resistance spot welding photon number2Wherein c is the number of photons varying with the welding time t, μ is the average number of all photons from the start of the welding current to the end of the welding current, and n is identical to n in (7.3);
(7.5) testing common quality problems in the micro resistance spot welding process and monitoring photon signals in the process to obtain micro resistance spot welding joints with different qualities, determining the joint quality grades of the micro resistance spot welding samples through a metallographic test, and respectively marking the grades as Excellent 2, good 3, qualified 4 and unqualified 0;
(7.6) the quality classification grade in the step 7.5) is only used for classification in the micro resistance spot welding quality model establishing process, and the quality model combines the excellent 2, excellent 3 and qualified 4 categories into the grade 1 in the actual application process, namely the quality model belongs to the qualified spot welding joint 1, and the unqualified spot welding joint is 0;
and (7.7) performing correlation analysis on the characteristic quantities in the steps (7.1), (7.2), (7.3) and (7.4) and the four quality grades in the step (7.5), and removing part of linearly related characteristic quantities to obtain the characteristic quantity with a larger correlation coefficient with the quality grades of the micro resistance spot welding joints.
Step (8), analyzing the obtained characteristic quantities, and taking the threshold value of each characteristic quantity as an initial judgment basis:
(8.1) counting the probability density of each characteristic quantity of different micro resistance spot welding joint quality grades (excellent 2, good 3, qualified 4 and unqualified 0) according to Gaussian distribution;
and (8.2) according to different application occasions of the miniature resistance spot welding joint, eliminating probability density overlapping sections of each characteristic quantity of different miniature resistance spot welding joint quality grades (excellent 2, good 3, qualified 4 and unqualified 0), and determining a high density section with each characteristic quantity as an initial judgment threshold.
Step (9), establishing a micro resistance spot welding quality model according to the characteristic quantity about photon change obtained in the step (7), verifying a micro resistance spot welding joint by adopting a double-model cross verification method in order to ensure the accuracy rate of monitoring the micro resistance spot welding quality, and randomly dividing an initially input micro resistance spot welding sample number N into two parts, namely a training sample set N1 and a test sample set N2, in the establishment process of the micro resistance spot welding joint quality monitoring model:
(9.1) establishing a miniature resistance spot welding joint quality monitoring model A based on multiple regression analysis and a least square method error criterion;
(9.1.1) utilizing N1 micro resistance spot welding samples, performing piecewise fitting on photon signals in the micro resistance spot welding process, and searching a fitting curve of the photon signals according to a least square method criterion;
(9.1.2) verifying the fitted curve in the step (9.1.1) based on the N2 micro resistance spot welding samples, and confirming a photon signal curve of a qualified spot welding sample when the testing accuracy is higher than a set value;
(9.2) establishing a miniature resistance spot welding joint quality monitoring model B based on a combined nearest classifier KNN algorithm;
(9.2.1) with N1 micro resistance spot welding samples, in case of a determined number of classifications, different KNN algorithm metric criteria were used:
euclidean distance formula:
manhattan distance:
and vector space cosine similarity:
establishing a miniature resistance spot welding monitoring model by combining the characteristic quantities, verifying different measurement criterion models by using N2 miniature resistance spot welding samples, and recording the monitoring results of each model on N2 miniature resistance spot welding joints;
and (9.2.2) according to the verification result of the step (9.2.1), giving different weights to the classification models for the monitoring accuracy rates of different measurement criteria so as to optimize the accuracy rate of the miniature resistance spot welding monitoring model.
And (10) storing the judgment boundary obtained in the step (8) and the monitoring model established in the step (9) into a database of the miniature resistance spot welding monitoring system corresponding to the process parameters.
When the characteristic quantity of a new miniature resistance spot welding sample enters a system, firstly judging whether all the characteristic quantities of the sample are within the initial judgment threshold interval in the step (8), and directly judging whether the spot welding joint belongs to a qualified joint 1 or an unqualified joint 0;
if all the characteristic quantities are within the initial judgment threshold interval obtained in the step (8), outputting an identifier 1 by the spot welding system, namely, indicating that the miniature resistance spot welding joint belongs to a qualified sample;
if the characteristic quantity of the sample is partially not in the initial judgment threshold interval in the step (8), judging the quality grade of the micro resistance spot welding sample by further using the micro resistance spot welding quality monitoring model established in the step (9) for the micro resistance spot welding sample; when the monitoring results of the miniature resistance spot welding joint quality monitoring models A, B are consistent, if the monitoring results of the miniature resistance spot welding joint quality monitoring models A, B all belong to qualified joints 1, the monitoring system automatically outputs 1, and if the monitoring results of the miniature resistance spot welding quality monitoring models A, B all belong to unqualified joints 0, the monitoring system automatically outputs 0;
when the monitoring results of the micro resistance spot welding quality monitoring model A, B are inconsistent, exception handling needs to be performed on the micro resistance spot welding joint, and an operator is reminded to use other detection methods or manually check whether the micro resistance spot welding joint meets the use requirements.
Example 1:
taking single photon monitoring in the series and parallel micro resistance spot welding process of the lithium battery in the new energy automobile power battery pack as an example, wherein the positive electrode and the negative electrode of the battery cell are stainless steel shells with the thickness of 0.3mm, the bus bar is a nickel sheet with the thickness of 0.1mm, the welding current is 1.5kA, the welding time is 2ms, and the electrode diameter is 1.5 mm. The method adopts a near-infrared single-photon detector with the working wavelength range of 900-1700 nm, the fastest response speed is up to 10ns, the external dimension of the infrared single-photon detector is a cylinder with the radius of 50mm and the height of 100mm, and the acquisition frequency of a data acquisition module is 500 Mhz.
Referring to fig. 1, which is a flow chart of a data acquisition and processing module of the invention, after entering a micro resistance spot welding single photon monitoring system, a system interface is shown in fig. 2, and the center of the interface displays details of a current welding material and technological conditions of micro resistance spot welding.
And if the parameters of the micro resistance spot welding process to be used do not change, directly clicking the middle button to enter a monitoring system. If the new micro resistance spot welding process parameters are used, the process parameter searching button on the lower left side needs to be clicked, and the existing model in the database is searched.
When the monitoring model under the corresponding process condition is searched in the database, the monitoring state can be directly selected to enter. When the monitoring model of the corresponding process parameters does not exist in the database, the required process parameters are input into the system, and single photon signal monitoring is carried out on the series and parallel micro resistance spot welding process of the lithium battery.
1) Selecting the characteristic quantity of the monitoring signal:
1.1 the nugget size and the nugget internal structure of the miniature resistance spot welding joint are closely related to the temperature of resistance heat, and the photon number in the time period from the beginning of welding current to the end of welding current is accumulated to obtain the total photon number c in the welding time.
1.2, data mining is carried out on a change curve of the photons c along with the time t, and photon characteristic quantities reflecting the quality of the miniature resistance spot welding joint are extracted: respectively obtaining the photon number c of the welding current starting timestart(generally corresponding to the minimum photon count), the maximum photon count cmaxThe number of turns m in the photon curve, and the number of photons c corresponding to each turntr1、ctr2、ctr3....ctrmTime t corresponding to two turnstr1、ttr2、ttr3....ttrmAnd the duration t of the number of balanced photonsb。
1.3 calculating the total number of photons c in each time interval of 0.2ms in the welding time t respectively1、c2....cn-1、cnAs the feature quantity, where n is t/0.2.
1.4 according to the variance equation
Calculating the variance characteristic quantity sigma of each micro resistance spot welding photon number2Wherein c is the number of photons changing with the welding time t, mu is the average number of all photons from the beginning of the welding current to the end of the welding current, and n is consistent with the meaning of n in 1.3;
1.5 testing common quality problems in the micro resistance spot welding process and monitoring photon signals in the process to obtain micro resistance spot welding joints with different qualities, determining the joint quality grade of a micro resistance spot welding sample through a metallographic test, and respectively marking as excellent 2, excellent 3, qualified 4 and unqualified 0.
1.6 the quality classification grade in 1.5) is only used for classification in the process of establishing the micro resistance spot welding quality model, and the quality model combines the excellent 2, good 3 and qualified 4 categories into the grade 1 in the actual application process, namely the quality model belongs to the qualified spot welding joint 1, and the unqualified spot welding joint is 0.
1.7, performing correlation analysis on each characteristic quantity in the 1.1)1.2)1.3)1.4) and four quality grades in the 7.5), and removing part of linearly related characteristic quantities to obtain a characteristic quantity with a large correlation coefficient with the quality grades of the micro resistance spot welding joints.
2) Analyzing the characteristic quantities obtained in the step 1), wherein the threshold value of each characteristic quantity is obtained as an initial judgment basis:
and 2.1 counting the probability density of each characteristic quantity of different micro resistance spot welding joint quality grades (excellent 2, good 3, qualified 4 and unqualified 0) according to Gaussian distribution.
And 2.2 according to different application occasions of the miniature resistance spot welding joint, eliminating probability density overlapping sections of each characteristic quantity of different miniature resistance spot welding joint quality grades (excellent 2, good 3, qualified 4 and unqualified 0), and determining a high density section with appropriate characteristic quantity to serve as an initial judgment threshold.
3) According to various characteristic quantities about photon change obtained in 1), a miniature resistance spot welding quality model is established, in order to ensure the accuracy rate of miniature resistance spot welding quality monitoring, a double-model cross validation method is adopted to carry out on a miniature resistance spot welding joint, and in the process of establishing the miniature resistance spot welding joint quality monitoring model, the number N of initially input miniature resistance spot welding samples is randomly divided into two parts, namely a training sample set N1 and a test sample set N2:
3.1 establishing a miniature resistance spot welding joint quality monitoring model A based on multivariate regression analysis and least square method error criteria.
3.1.1, utilizing N1 micro resistance spot welding samples to perform segmented fitting on photon signals in the micro resistance spot welding process, and searching a fitting curve of the photon signals under a specific process condition according to a least square method criterion.
3.1.2 the fitted curve in 3.1.1 was verified based on N2 micro resistance spot welding samples, and when the test accuracy was higher than the set value, the photon signal curve of the qualified spot welding sample was confirmed.
3.2 establishing a miniature resistance spot welding joint quality monitoring model B based on a combined KNN (nearest neighbor classifier) algorithm.
3.2.1, under the condition that the classification number is determined, respectively using different KNN algorithm measurement criteria, establishing a miniature resistance spot welding monitoring model by combining the characteristic quantity by using N1 miniature resistance spot welding samples, verifying different measurement criteria models by using N2 miniature resistance spot welding samples, and recording the monitoring results of each model on N2 miniature resistance spot welding joints.
3.2.2 according to the verification result of the 3.2.1, giving different weights to the classification model for the monitoring accuracy of different measurement criteria so as to optimize the accuracy of the miniature resistance spot welding monitoring model.
4) And storing the judgment boundary obtained in the step 2) and the monitoring model established in the step 3) into a database of the miniature resistance spot welding monitoring system corresponding to the process parameters.
5) When each characteristic quantity of a new micro resistance spot welding sample enters a system, firstly, whether all the characteristic quantities of the sample are within the initial judgment threshold value interval in the step 8) is judged, and whether the spot welding joint belongs to a qualified joint 1 or an unqualified joint 0 can be directly judged.
6) And if all the characteristic quantities are within the initial judgment threshold interval obtained in the step 2), outputting an identifier 1 by the spot welding system, namely, indicating that the miniature resistance spot welding joint belongs to a qualified sample.
7) If the characteristic quantity of the sample exists and part of the characteristic quantity is not in the initial judgment threshold interval in the step 2), the quality grade of the micro resistance spot welding sample needs to be judged by further utilizing the micro resistance spot welding quality monitoring model established in the step 3). When the monitoring results of the models A, B are consistent, if the monitoring results of the models A, B all belong to qualified joints 1, the monitoring system automatically outputs 1, and if the monitoring results of the models A, B all belong to unqualified joints 0, the monitoring system automatically outputs 0.
When the monitoring results of the model A, B are inconsistent, exception handling needs to be performed on the micro resistance spot welding, and an operator is reminded to use other detection methods or manually check whether the spot welding joint meets the use requirements.
The above description is only a preferred example of the present invention and is not intended to limit the present invention, and various modifications and changes may be made by those skilled in the art. Any modification, equivalent replacement, improvement and the like of the present invention shall be included in the protection scope of the present invention.
Claims (5)
1. A micro resistance spot welding quality on-line monitoring method is characterized in that: firstly, a single photon detector is utilized to monitor the change condition of the photon number of the contact edge of an electrode and a workpiece along with time in the production process of miniature resistance spot welding in real time, and a photon change curve is drawn; determining the quality grades of the miniature resistance spot welding joints under different process conditions through a metallographic experiment, wherein photon change curves are different in the miniature resistance spot welding process of different qualities, extracting photon characteristic quantities of the miniature resistance spot welding qualities of different quality grades by using a data mining means, generating a miniature resistance spot welding quality classification model, establishing a database of miniature resistance spot welding quality monitoring based on photon signals, and realizing automatic online monitoring of the miniature resistance spot welding quality under different process conditions; the method specifically comprises the following steps:
selecting characteristic quantity of a photon signal;
step (2), analyzing the obtained characteristic quantities, and taking the threshold value of each characteristic quantity as an initial judgment basis;
step (3), establishing a micro resistance spot welding quality model according to the characteristic quantity about photon change obtained in the step (1), and monitoring the micro resistance spot welding quality by adopting a dual-model cross validation method in order to ensure the accuracy of monitoring the micro resistance spot welding quality; in the process of establishing the micro resistance spot welding quality monitoring model, dividing an initially input micro resistance spot welding sample number N into two parts, namely a training sample set N1 and a testing sample set N2;
and (4) storing the judgment boundary obtained in the step (2) and the monitoring model established in the step (3) into a database of the miniature resistance spot welding monitoring system corresponding to the process parameters.
2. The on-line monitoring method for the quality of the miniature resistance spot welding according to claim 1, characterized in that: the selection of the characteristic quantity of the photon signal in the step (1) is as follows:
(1.1) accumulating the photon number in the time period from the beginning of the welding current to the end of the welding current to obtain the total photon number c in the welding time;
(1.2) carrying out data mining on a change curve of the number of photons c along with the welding time t, and extracting photon characteristic quantities reflecting the quality of the miniature resistance spot welding joint: respectively obtaining the photon number c of the welding current starting timestartGenerally corresponding to the minimum photon number, the maximum photon number cmaxThe number of turns m in the photon curve, and the number of photons c corresponding to each turntr1、ctr2、ctr3....ctrmTime t corresponding to two turnstr1、ttr2、ttr3....ttrmAnd the duration t of the number of balanced photonsb；
(1.3) respectively calculating the total number c of photons in each time within the current duration t and every 0.2ms time interval1、c2....cn-1、cnAs a feature amount, wherein n ═ t/0.2;
(1.4) formula according to variance
Calculating the variance characteristic quantity sigma of each micro resistance spot welding photon number2Wherein c is the number of photons which varies with the welding time t, μ is the average number of all photons from the start of the welding current to the end of the welding current, and n is the same as n in (1.3);
(1.5) testing common quality problems in the micro resistance spot welding process and monitoring photon signals in the process to obtain micro resistance spot welding joints with different qualities, determining the joint quality grades of the micro resistance spot welding samples through a metallographic test, and respectively marking the grades as Excellent 2, good 3, qualified 4 and unqualified 0;
(1.6) the quality classification grade in the step (1.5) is only used for classification in the micro resistance spot welding quality model establishing process, and the quality model combines the excellent 2, good 3 and qualified 4 categories into the grade 1 in the actual application process, namely the quality model belongs to the qualified spot welding joint 1, and the unqualified spot welding joint is 0;
and (1.7) performing correlation analysis on the characteristic quantities in the steps (1.1), (1.2), (1.3) and (1.4) and the four quality grades in the step (1.5), and removing part of linearly related characteristic quantities to obtain the characteristic quantities with larger correlation coefficients with the quality grades of the micro resistance spot welding joints.
3. The on-line monitoring method for the quality of the miniature resistance spot welding according to claim 1, characterized in that: analyzing the obtained characteristic quantities, wherein the threshold values of the obtained characteristic quantities are used as an initial judgment basis, and the method specifically comprises the following steps:
(2.1) counting the probability density of each characteristic quantity of different micro resistance spot welding joint quality levels according to Gaussian distribution;
(2.2) according to different application occasions of the miniature resistance spot welding joint, eliminating probability density overlapping sections of all characteristic quantities of different miniature resistance spot welding joint quality levels, and determining a high density section with proper characteristic quantities to serve as an initial judgment threshold;
the quality grades are as follows: excellent 2, good 3, qualified 4, and unqualified 0.
4. The on-line monitoring method for the quality of the miniature resistance spot welding according to claim 1, characterized in that: establishing a miniature resistance spot welding quality monitoring model in the step (3), specifically:
(3.1) establishing a miniature resistance spot welding joint quality monitoring model A based on multiple regression analysis and a least square method error criterion;
(3.1.1) utilizing N1 micro resistance spot welding samples to perform piecewise fitting on photon signals in the micro resistance spot welding process, and searching a fitting curve of the photon signals according to a least square method criterion;
(3.1.2) verifying the fitting curve in the step (3.1.1) based on the N2 micro resistance spot welding samples, and when the testing accuracy is higher than a set value, confirming a photon signal curve of a qualified spot welding sample;
(3.2) establishing a miniature resistance spot welding joint quality monitoring model B based on a combined nearest classifier KNN algorithm;
(3.2.1) utilizing N1 micro resistance spot welding samples, respectively using different KNN algorithm measurement criteria under the condition of determining the classification number, establishing a micro resistance spot welding monitoring model by combining characteristic quantity, verifying different measurement criteria models by using N2 micro resistance spot welding samples, and recording the monitoring results of each model on N2 micro resistance spot welding joints;
and (3.2.2) according to the verification result of the step (3.2.1), giving different weights to the classification models for the monitoring accuracy rates of different measurement criteria so as to optimize the accuracy rate of the miniature resistance spot welding monitoring model.
5. The on-line monitoring method of the quality of micro resistance spot welding according to any one of claims 1 to 4, characterized in that: when the characteristic quantity of a new miniature resistance spot welding sample enters a system, firstly, judging whether all the characteristic quantities of the sample are within the initial judgment threshold interval in the step (2), and directly judging whether the spot welding joint belongs to a qualified joint 1 or an unqualified joint 0;
if all the characteristic quantities are within the initial judgment threshold interval obtained in the step (2), outputting an identifier 1 by the miniature resistance spot welding joint quality monitoring system, namely, indicating that the miniature resistance spot welding joint belongs to a qualified sample;
if the characteristic quantity of the sample is partially not in the initial judgment threshold interval in the step (2), judging the quality grade of the micro resistance spot welding sample by further using the micro resistance spot welding quality monitoring model established in the step (3) for the micro resistance spot welding sample; when the monitoring results of the miniature resistance spot welding joint quality monitoring models A, B are consistent, if the monitoring results of the miniature resistance spot welding joint quality monitoring models A, B all belong to qualified joints 1, the monitoring system automatically outputs 1, and if the monitoring results of the miniature resistance spot welding joint quality monitoring models A, B all belong to unqualified joints 0, the monitoring system automatically outputs 0;
when the monitoring results of the micro resistance spot welding joint quality monitoring model A, B are inconsistent, abnormal treatment needs to be given to the micro resistance spot welding joint, and an operator is reminded to use other detection methods or manually check whether the micro resistance spot welding joint meets the use requirements.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810938371.2A CN109014544B (en) | 2018-08-17 | 2018-08-17 | Micro resistance spot welding quality on-line monitoring method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810938371.2A CN109014544B (en) | 2018-08-17 | 2018-08-17 | Micro resistance spot welding quality on-line monitoring method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109014544A CN109014544A (en) | 2018-12-18 |
| CN109014544B true CN109014544B (en) | 2020-07-28 |
Family
ID=64630790
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810938371.2A Active CN109014544B (en) | 2018-08-17 | 2018-08-17 | Micro resistance spot welding quality on-line monitoring method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109014544B (en) |
Families Citing this family (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110227878B (en) * | 2019-06-27 | 2021-02-26 | 天津七所高科技有限公司 | Method for detecting dislocation or loss of nut by using resistance welding equipment |
| CN112453750A (en) * | 2019-09-09 | 2021-03-09 | 英业达科技有限公司 | System and method for establishing detection model according to standard value to confirm welding state |
| CN112557416A (en) * | 2019-09-09 | 2021-03-26 | 英业达科技有限公司 | System and method for detecting whether welding spots are bridged or not by using deep learning model |
| CN111390367B (en) * | 2020-05-13 | 2021-09-03 | 天津七所高科技有限公司 | Method for identifying welding spot splashing in resistance spot welding |
| CN111681235B (en) * | 2020-06-11 | 2023-05-09 | 广东工业大学 | IC welding spot defect detection method based on learning mechanism |
| CN112756759B (en) * | 2021-01-11 | 2022-04-08 | 上海智能制造功能平台有限公司 | Spot welding robot workstation fault judgment method |
| CN114147382B (en) * | 2021-12-14 | 2023-06-30 | 吉林大学 | Method and device for on-line monitoring of electric and ultrasonic signal combined resistance spot welding quality |
| CN117436769A (en) * | 2023-12-20 | 2024-01-23 | 山东方垠智能制造有限公司 | Structural part welding quality monitoring method, system, storage medium and equipment |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US4079225A (en) * | 1976-08-04 | 1978-03-14 | Warner Allan S | Fiber optic/photon detector for brazing machine |
| CN103862136A (en) * | 2014-03-14 | 2014-06-18 | 吉林大学 | Monitoring device and monitoring method of fusion-welding process |
| CN105750754A (en) * | 2016-05-06 | 2016-07-13 | 广州市精源电子设备有限公司 | Method and system for identifying resistance spot welding quality affecting factors |
| CN107160048A (en) * | 2017-05-22 | 2017-09-15 | 哈尔滨工业大学（威海） | A kind of inline diagnosis method of the laser welding defect based on spectral information |
| KR20170136111A (en) * | 2016-05-31 | 2017-12-11 | 피에프케이 (주) | Smart welding machine monitoring system |
| CN108188560A (en) * | 2017-12-12 | 2018-06-22 | 广州亨龙智能装备股份有限公司 | Hand-held electric resistance welding Quality Monitoring Control System based on linux system |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US10207363B2 (en) * | 2014-03-24 | 2019-02-19 | James Eldon Craig | Additive manufacturing temperature controller/sensor apparatus and method of use thereof |
-
2018
- 2018-08-17 CN CN201810938371.2A patent/CN109014544B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US4079225A (en) * | 1976-08-04 | 1978-03-14 | Warner Allan S | Fiber optic/photon detector for brazing machine |
| CN103862136A (en) * | 2014-03-14 | 2014-06-18 | 吉林大学 | Monitoring device and monitoring method of fusion-welding process |
| CN105750754A (en) * | 2016-05-06 | 2016-07-13 | 广州市精源电子设备有限公司 | Method and system for identifying resistance spot welding quality affecting factors |
| KR20170136111A (en) * | 2016-05-31 | 2017-12-11 | 피에프케이 (주) | Smart welding machine monitoring system |
| CN107160048A (en) * | 2017-05-22 | 2017-09-15 | 哈尔滨工业大学（威海） | A kind of inline diagnosis method of the laser welding defect based on spectral information |
| CN108188560A (en) * | 2017-12-12 | 2018-06-22 | 广州亨龙智能装备股份有限公司 | Hand-held electric resistance welding Quality Monitoring Control System based on linux system |
Non-Patent Citations (2)
| Title |
|---|
| 微电阻点焊条件下的质量控制问题研究;万晓东;《中国博士学位论文全文数据库 工程科技Ⅰ辑》;20170815;B022-56 * |
| 电阻点焊质量监测智能化及系统研究;文静;《中国博士学位论文全文数据库 工程科技Ⅰ辑》;20100815;B022-39 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109014544A (en) | 2018-12-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109014544B (en) | Micro resistance spot welding quality on-line monitoring method | |
| JP6145824B2 (en) | Battery diagnostic device and battery diagnostic method | |
| CN104457841B (en) | Resistance spot welding quality on-line monitoring method | |
| CN110824006B (en) | Postweld weld impact quality discrimination method based on intelligent acoustic information identification | |
| CN112487708A (en) | Resistance spot welding quality prediction method based on ensemble learning | |
| CN105866690B (en) | A kind of battery status detection method and system based on capacitance sensor | |
| CN106557637A (en) | Safety performance evaluation method of energy storage product | |
| CN106198749A (en) | A kind of data fusion method of multiple sensor based on Metal Crack monitoring | |
| KR102163828B1 (en) | Machine learning system for spatter prediction from welding waveform and method thereof | |
| CN104002019A (en) | Electric arc chaotic characteristic optimization algorithm based welding material manufacturability evaluation method | |
| CN111331211B (en) | On-line penetration detection method and system for electric spark small hole machining | |
| CN114267150A (en) | Early fire detection and early warning method for thermal runaway of lithium battery | |
| CN114325433A (en) | Lithium ion battery fault detection method and system based on electrochemical impedance spectrum test | |
| CN116520068A (en) | Diagnostic method, device, equipment and storage medium for electric power data | |
| CN108857027B (en) | Quality monitoring device and monitoring method in micro spot welding process | |
| CN201408467Y (en) | Intelligent fire alarm system | |
| CN103675103A (en) | Method for quantitative detection of nugget crack by using energy equivalent of acoustic emission signal in resistance spot-welding process | |
| CN110988700A (en) | Forklift lithium ion battery module health degree evaluation method | |
| EP3888839A1 (en) | Method and system for monitoring and identifying the weld quality on metallic components | |
| CN103994821A (en) | Method for evaluating resistance spot welding spattering quantitatively in real time through acoustic emission signal frequency spectrum | |
| CN105928962B (en) | Alloy designations recognition methods based on membership function | |
| CN104614409B (en) | A kind of milk powder detection method and device based on technology of Internet of things | |
| CN117195134B (en) | Early warning method and device for hydrogen fuel base station power supply | |
| CN117117356B (en) | Energy storage battery pack thermal runaway monitoring and tracing method and system | |
| CN106990096A (en) | Meat meal tankage kind detection method and system based on LIBS |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |