A cold-rolled strip lap welding result evaluation method based on big data

Abstract

The application discloses a lap welding result evaluation method for cold-rolled strip steel based on big data, which takes welding history big data as a sample set, transmits the performance data (welding current, welding speed, lap amount, lap compensation amount, welding wheel pressure and temperature) of the current weld recorded by a digital steel coil to a software system through a data interface, collects the welding performance data (welding current, welding wheel pressure, welding speed, lap amount, lap compensation amount and temperature) through inherent acquisition functions of a welding machine PLC, collects weld height information through an external laser range finder, then the hardware system including the welding machine PLC and the laser range finder is in communication with an FDDA acquisition system, the FDDA acquisition system is in interaction with a digital steel coil system, the digital steel coil system is in interaction with an OLW, and the welding result is determined by integrated welding result determination logic (embedded in the OLW), welding quality quantitative evaluation is realized, and the determination result (release or re-welding) is transmitted to the welding machine through the software system, so that welding quality evaluation and quantitative results are given.

Description

TECHNICAL FIELD

[0001] The present application relates to thin strip steel welding control technology, and in particular to a cold-rolled strip lap welding result evaluation method based on big data. BACKGROUND

[0002] There are many factors affecting the quality of the lap welding seam, which can be summarized into three categories: first, the rationality of the welding process parameters, mainly including the set value of the welding process parameters, commonly known as L2, such as current setting, roller pressure setting, welding speed, lap amount and compensation amount setting; second, the influence of the workpiece material itself, mainly including the steel grade (material code) of the strip, width, thickness and surface state; third, the equipment state parameters of the welding machine itself, such as clamping force and its uniformity, strip positioning accuracy, welding wheel state, and post-weld flattening wheel pressure stability.

[0003] At present, the quality evaluation of the lap welding seam of the cold-rolled strip welding machine mainly relies on manual work. After welding, the on-site operator first confirms the output results of the QCDS system, and then uses the method of manual on-site confirmation at the crescent shear, such as early hammering, the commonly used cupping test and bending test, and then the operator makes a "decision" on whether to release. However, for this mode of operation, on the one hand, there is a lack of data platform for effective analysis of welding historical data; on the other hand, in the welding quality evaluation process, it largely relies on subjective judgment and manual test, which not only has a one-sided knowledge acquisition and inheritance, but also cannot form a quantitative evaluation of the welding quality and provide effective guidance for the maintenance of the welding machine. SUMMARY

[0004] In order to solve the above problems of the prior art, the present application provides a cold-rolled strip lap welding result evaluation method based on big data, which takes the lap welding historical big data of the cold-rolled strip as a sample set, analyzes the welding historical big data to induce the welding rules, establishes a welding quality logical judgment process, and realizes welding quality evaluation and gives quantitative results through logical judgment.

[0005] A cold-rolled strip lap welding result evaluation method based on big data, comprising the following steps:

[0006] S1. According to the lap welding historical big data of the cold-rolled strip, the corresponding welding parameters associated with high welding temperature are screened out, and the threshold values of each parameter are preset;

[0007] S2. The actual welding parameters are collected by the welding machine PLC, if all the welding parameters meet the set threshold value, it is determined to be qualified and scored, and the next step is entered; if any parameter does not meet the requirement, the welding seam is determined to be unqualified;

[0008] S3. The upper and lower limits of the measured height of the weld are judged, if it is qualified, then the score is given and the next step is entered; if it is not qualified, then the weld is judged unqualified;

[0009] S4. The temperature effective zone data of the weld is obtained, and the trend, volatility and temperature distribution of the weld temperature are judged and scored in turn;

[0010] S5. If the total score of steps S2-S4 is higher than the set value, then the weld judgment result and score are sent to the welding machine.

[0011] In step S1, the welding parameters include welding current, welding speed, overlap amount / overlap compensation amount, and welding wheel pressure.

[0012] In step S1, the set threshold value is the deviation value of the corresponding set parameter and the actual parameter ±4%.

[0013] In step S2, the score of the qualified judgment is 10 points per parameter.

[0014] In step S3, the upper and lower limit judgment includes whether it is three-piece welding and whether the height is inconsistent on the weld length; the upper limit is 1.5 times the maximum thickness of the front and rear coils, and the lower limit is the minimum thickness of the front and rear coils; the score of the qualified judgment is 10 points.

[0015] In step S4, the trend and volatility of multiple weld temperatures are counted, and the density function of the trend and volatility distribution of the weld temperature is used to develop the judgment and scoring criteria for the trend and volatility as follows:

[0016] The trend of temperature is between [-1, 0.2] and gets 20 points; the trend of temperature is between [-1.2, -1), (0.2, 0.5] and gets 10 points; the trend of temperature is between [-2, -1.2), (0.5, 2] and gets 5 points; the absolute value of the trend of temperature exceeds 2, then NG;

[0017] The volatility of temperature is between [0, 25] and gets 20 points; the volatility of temperature is between (25, 40] and gets 10 points; the volatility of temperature is between (40, 60] and gets 5 points; the volatility of temperature exceeds 60, then NG.

[0018] In step S4, the temperature mean ±3SD (standard deviation) is used to evaluate the temperature distribution of each point in the effective zone: the temperature points falling within the mean ±3SD interval are judged to be qualified; the points exceeding the interval are unqualified; using the deduction system, a total of 10 points, each unqualified point is deducted one point, until it is reduced to 0 points.

[0019] In step S5, the set value is 60 points.

[0020] The cold-rolled strip lap welding result evaluation method based on big data can analyze welding historical big data and establish welding process logical judgment, and realize quantitative scoring of the weld of known input parameters (welding current, welding wheel pressure, lap amount, lap compensation amount, welding speed, temperature). The method can determine the weld quality in time after welding according to welding historical big data, and realize quantitative evaluation of the weld quality, so that the welding quality evaluation process greatly relies on subjective judgment of people and auxiliary artificial test, knowledge acquisition and inheritance are relatively one-sided, and quantitative evaluation of the welding quality cannot be formed. BRIEF DESCRIPTION OF DRAWINGS

[0021] Figure 1 is a principle block diagram of the evaluation method of the present application;

[0022] Figure 2 is a correlation coefficient thermodynamic diagram of welding process parameters and welding temperature of the present application;

[0023] Figure 2 In the formula, H1: front row coil thickness; H2: rear row coil thickness; V: welding speed; P1: welding pressure; A: welding current; M1: lap amount; P2: rolling wheel pressure; M2: lap compensation amount. DETAILED DESCRIPTION

[0024] The cold-rolled strip lap welding result evaluation method based on big data of the present application is further described below, and the principle of the method is shown in Figure 1

[0025] The current weld performance data (welding current, welding speed, lap amount, lap compensation amount, welding wheel pressure, temperature) recorded by the digital steel coil is transmitted to the software system through the data interface (the welding performance data (welding current, welding wheel pressure, welding speed, lap amount, lap compensation amount, temperature) is collected through the inherent collection function of the welding machine PLC, the laser range finder is externally connected to collect the weld height information, then the welding machine PLC and the laser range finder and other hardware systems are in communication with the FDDA collection system, the FDDA collection system interacts with the digital steel coil system, and the digital steel coil system interacts with the OLW), and the integrated welding result judgment logic (embedded in the OLW) is used for judgment, so as to realize quantitative evaluation of the welding quality, and the software system judgment result (pass or re-weld) is transmitted to the welding machine.

[0026] Judgment logic

[0027] After welding, the judgment of the OLW on the weld is subject to the following logic:

[0028] Table 1 judgment logic table

[0029] ​ Step Judgment item Judgment basis Judgment result A Welder state Predefined parameter Pass (score) A-1 Welding speed ±4% of set value Pass (10 points) A-2 Wheel pressure ±4% of set value Pass (10 points) A-3 Lap amount / compensation ±4% of set value Pass (10 points) A-4 Welding current ±4% of set value Pass (10 points) B Weld height [Hmin, 1.5Hmax] Pass (10 points) C Welding temperature Predefined upper and lower temperature limits Score, OK / NK C-1 Temperature trend 3000-weld trend statistics Pass (20 points) C-2 Temperature fluctuation 3000-weld fluctuation statistics Pass (20 points) C-3 Temperature level 30° above lower limit Pass (10 points)

[0030] Note: H represents the maximum thickness of the cold-rolled steel strip to be welded.

[0031] Specifically as follows:

[0032] S1. Based on historical big data of lap welding of cold-rolled strip steel, select the corresponding welding parameters that are highly correlated with the welding temperature, and preset the threshold values ​​for each parameter.

[0033] S2. Collect the above actual welding parameters through the welding machine PLC. If all welding parameters meet the set threshold, the weld is deemed qualified and scored, and the process proceeds to the next step. If any parameter does not meet the threshold, the weld is deemed unqualified.

[0034] First, the stability and rationality of the welding parameters must be determined, as this is a prerequisite for welding. Therefore, the welding data should be monitored softly to determine whether the welding parameters used are reasonable. The score for passing the test is 10 points per parameter. All parameters passing: OK (40 points); if any parameter fails, the weld is considered unqualified.

[0035] The specific approach is as follows:

[0036] Further research into the relationship between welding process parameters and welding temperature will help in determining the welding result (temperature). For example... Figure 2 As shown, according to correlation analysis, all parameters are almost highly correlated with temperature. Figure 2 This is a heat map showing the correlation coefficients between welding parameters and temperature for welding performance data of a certain unit over the past six months. Among these parameters, welding current, welding wheel pressure, and rolling pressure are crucial welding machine parameters and directly affect welding quality (closely related to the temperature generated during welding). Based on analysis of historical data of welds deemed unqualified by manual on-site inspection, a pre-set threshold (±4%) was established between the set value and the actual welding value. Only if all welding performance parameters are qualified is the next step of evaluation proceeded; if any parameter is unqualified, the weld is deemed unqualified.

[0037] S3. Determine the upper and lower limits of the measured height of the weld. If it meets the limits, it is deemed qualified and scored, and proceeds to the next step; if it does not meet the limits, it is deemed unqualified.

[0038] The weld height (welding result) is determined by analyzing the weld height data. Is it a three-piece weld (normally the front and rear coils are welded together, but a three-piece weld is three steel plates welded together, which increases the weld height)? Is the height inconsistent along the weld length? None: OK (10 points), Yes: NG.

[0039] Weld height determination method: Install laser rangefinders (one on top and one on the bottom) on one side of the welding machine's grinding wheel to monitor the weld height in real time. A Modbus TCP to OPC DA gateway converts the laser rangefinder signal into an OPC DA interface signal, which is then transmitted to the FDA (Fast-To-Area Controller) via network cable. The FDA then transmits the signal to the OLW (Low-Installation Wrapper) via the digital steel coil data terminal. The OLW has built-in software for converting electrical signals to height signals, thus receiving the weld height information. Specifically, the logic involves determining upper and lower limits for the weld height: the upper limit is 1.5 times the maximum thickness of the preceding and following coils, and the lower limit is the minimum thickness of the preceding and following coils (a threshold based on field experience).

[0040] S4. Obtain the effective temperature zone data of the weld, and judge and score the weld trend, fluctuation and temperature distribution in turn;

[0041] The weld temperature (welding result) is judged. If it passes, it is OK; otherwise, it is NG. At the same time, a weld quality score is given.

[0042] Temperature is the most important criterion for judging weld quality, as it is the result of welding. First, the effective zone data of the weld is determined, and temperature data that does not conform to the norm, such as obvious sharp points, are eliminated. Then, the trend and fluctuation of the weld temperature are judged (the optimal weld temperature curve should be a flat horizontal line, that is, without a gradual increasing or decreasing trend; in other words, the trend is the slope of the linear temperature fit, and the least squares method is used here to obtain the slope for linear fitting). Then, the fluctuation of the weld temperature is judged (measured by standard deviation).

[0043] Based on the above, the average temperature of the weld is calculated. If the average temperature is within the upper and lower limits of the weld temperature, the temperature condition is met. According to the scoring rules, the final temperature of the weld is obtained.

[0044] Preliminary data processing stage:

[0045] Temperature spikes (abnormal data points) are filtered out (the first and last data points collected by the welding machine PLC after welding begins are deleted because they deviate from normal values; this stage is also called weld effective zone data filtering). Then, the data from the weld effective zone is used to determine whether the upper and lower temperature limits are exceeded. If they are exceeded, it is NG (as required on site). If it is qualified, the valid welding data is retained for the following judgments.

[0046] Judgment Phase:

[0047] The trends and fluctuations of 3000 weld seams were statistically analyzed. Then, based on the density function of the distribution of the trends and fluctuations of the 3000 weld seams (intervals with higher distribution scores and lower distribution scores), and after consultation with the on-site staff, the scoring criteria were established as follows:

[0048] I. Trend Determination

[0049] 1. A trend between [-1, 0.2] scores 20 points;

[0050] 2. A trend between [-1.2, -1) and (0.2, 0.5] scores 10 points;

[0051] 3. A trend between [-2, -1.2) and (0.5, 2] scores 5 points;

[0052] 4. If the absolute value of linearity exceeds 2, then it is NG.

[0053] II. Fluctuation Determination

[0054] 1. Volatility within the range [0, 25] scores 20 points;

[0055] 2. Volatility within the range of (25, 40) scores 10 points;

[0056] 3. Volatility within (40, 60) scores 5 points;

[0057] 4. If the volatility exceeds 60, then it is NG.

[0058] III. Temperature Distribution Judgment

[0059] The temperature distribution at each point within the valid range is evaluated using the mean ± 3 SD (standard deviation). Temperature points falling within the mean ± 3 SD range are considered acceptable (as agreed upon in discussions with the site team), while those exceeding the range are considered unacceptable and thus subject to a point deduction system. A total of 10 points are awarded, with one point deducted for each additional point (exceeding the mean ± 3 SD range), until a total of 0 points are reached.

[0060] S5. If the total score of steps S2-S4 is higher than the set value, the weld judgment result and score will be sent to the welding machine.

[0061] Finally, the OLW software communicates with the welding machine to send the weld judgment results and scores:

[0062] The OLW communicates directly with the welding machine, sending the judgment result and score (a score below 60 indicates an NG, while a score of 60 or higher indicates an OK) to the welding machine. The welding machine's PLC receives the result and displays it on the welding machine's operating terminal, allowing the operator to approve the weld or request a re-welding.

[0063] Example 1

[0064] 1) Before welding, the welding parameters are displayed on the computer terminal;

[0065] In this embodiment, the first strip has a coil number of 521697100, a steel grade of 2, a strip width of 1409 mm, and a strip thickness of 0.7 mm; the second strip has a coil number of 521698100, a steel grade of 2, a strip width of 1276 mm, and a strip thickness of 0.8 mm.

[0066] In this embodiment, the welding parameters preset by the welding machine are: welding current 16KA, welding speed 10cm / s, welding wheel pressure 9kN, overlap 1mm, overlap compensation 0.6mm, upper temperature limit 1100℃, and lower temperature limit 820℃.

[0067] 2) At the same time, the digital steel coil data platform sends information such as the coil number and thickness of the preceding and following coils to the OLW software.

[0068] 3) When welding begins, the infrared detection device synchronously sends the real-time weld height to the OLW software; the welding machine PLC synchronously sends the real-time welding setting data to the digital steel coil data platform, and then the data platform synchronously forwards the setting data to the OLW software.

[0069] In step 3), the welding performance data includes: welding current, welding wheel pressure, overlap amount, overlap compensation amount, welding speed, and welding temperature.

[0070] Table 2 Welding Performance Data

[0071]

[0072]

[0073]

[0074]

[0075] 5) After the above steps are completed, the OLW software will make a judgment.

[0076] First: Check the condition of the welding machine. If all are qualified: OK. If any one is unqualified, the weld is unqualified.

[0077] Based on the pre-set threshold (the deviation between the set value and the actual welding value) being ±4%, all the welding parameters of the above welding data are qualified, and the next step of judgment is carried out.

[0078] Then: Determine the weld height.

[0079] Judge the weld defects such as three-piece welding according to the height of the whole weld. The specific logic is to judge the upper and lower limits of the weld height. The upper limit is 1.5 times the maximum value of the thickness of the forward coil and the backward coil, and the lower limit is the minimum value of the thickness of the forward coil and the backward coil (the threshold given by on-site working experience). If three-piece welding occurs, the weld height increases; if the overlap amount is insufficient, the weld height is on the small side. The average weld height is 0.9 mm, which is within the preset range, so it is judged as qualified and the next judgment is carried out.

[0080] Finally: Judge the weld temperature.

[0081] If the temperature is within the preset upper and lower limits (the upper limit is 1100 °C and the lower limit is 870 °C), it is OK; otherwise, it is NG, and the weld quality score is given at the same time. The temperature is the welding result and also the most important criterion for judging weld quality.

[0082] The specific logic is as follows: First, determine the data in the effective area of the weld temperature, eliminate the abnormal temperature data such as obvious sharp points in the effective area data of the temperature, then judge the trend of the weld temperature, then judge the weld volatility, and finally score the temperature distribution of each point.

[0083] That is, the welding process parameters are qualified, getting 40 points; the weld height is qualified, getting 10 points; the trend is -0.76, getting 20 points; the volatility is 23.73 °C, getting 20 points; the mean ± SD interval is [859, 1001], and there are two temperatures not in this interval, getting 8 points. The overall score of the weld: 40 + 10 + 20 + c20 + 8 = 98 points, which is qualified, so it is OK.

[0084] According to the scoring and judgment of the OLW software, the score of this weld is 98 points and it is judged as qualified (OK).

[0085] 6) In step 5), the judgment result of the OLW is transmitted back to the welder PLC and displayed on the computer terminal;

[0086] 7) When the judgment of the welder PLC is inconsistent with the displayed result in step 6), the operator will conduct a re-inspection for further judgment.

[0087] In summary, lap welding machines are typically equipped with relatively simple QCDS systems. The common practice for evaluating the quality of lap welds is that after welding, on-site operators first confirm the output of the QCDS (based on welding machine process parameters) system, supplemented by manual on-site verification methods at the crescent shear, such as the early drop hammer test, and the now widely used cupping test and bending test, before the operator makes a "decision" on whether to release the weld. This method relies heavily on subjective human judgment and supplementary manual testing, resulting in a somewhat one-sided knowledge acquisition and transmission, and failing to provide effective guidance for quantitative assessment of welding quality and maintenance of the welding machine. However, the welding result evaluation method of this invention enables online evaluation of welding quality and quantitative scoring. During a one-month trial run on-site, the number of re-weldings due to software misjudgments was less than 1 / 3 of a day, and the number of time-lapses due to software misjudgments was less than 0.

[0088] However, those skilled in the art should recognize that the above embodiments are only used to illustrate the present invention and are not intended to limit the present invention. Any changes or modifications to the above embodiments that are within the essential spirit of the present invention will fall within the scope of the claims of the present invention.

Claims

1. A method for evaluating the welding results of lap welds on cold-rolled strip steel based on big data, characterized in that... , including the following steps: S1. According to the big data of the lap welding history of cold-rolled strip steel, screen out the corresponding welding parameters highly correlated with the welding temperature, and preset the thresholds of each parameter; S2. Collect the above actual welding parameters through the welder PLC. If all the welding parameters meet the set thresholds, it is judged as qualified and scored, and proceed to the next step; if any parameter does not meet the requirements, the weld is judged as unqualified; S3. Judge the upper and lower limits of the measured height of the weld. If it meets the requirements, it is judged as qualified and scored, and proceed to the next step; if it does not meet the requirements, the weld is judged as unqualified; S4. Obtain the data of the effective temperature area of the weld, and judge and score the trend, volatility and temperature distribution of the weld temperature in turn; S5. If the total score of the scoring in steps S2 - S4 is higher than the set value, the weld is judged as qualified, and the judgment result and the specific score are sent to the welder; In step S4, the trends and volatilities of the temperatures of multiple welds are statistically analyzed, and the judgment and scoring criteria for trends and volatilities are formulated according to the density functions of the trends and volatilities of the weld temperatures.

2. The method for evaluating the welding results of cold-rolled strip lap welds based on big data as described in claim 1, characterized in that: In step S1, the welding parameters include welding current, welding speed, lap amount / lap compensation amount, and wheel pressure.

3. The method for evaluating the welding results of cold-rolled strip lap welds based on big data as described in claim 1 or 2, characterized in that: In step S1, the set threshold is the deviation value of ±4% between the corresponding set parameter and the actual parameter.

4. The method for evaluating the welding results of cold-rolled strip lap welds based on big data as described in claim 1, characterized in that: In step S2, the score for judging qualified is 10 points for each parameter.

5. The method for evaluating the welding results of cold-rolled strip lap welds based on big data as described in claim 1, characterized in that: In step S3, the upper and lower limit judgment includes whether it is a three-piece weld and whether the height is inconsistent along the weld length; the upper limit is 1.5 times the maximum value of the thickness of the forward coil and the backward coil, and the lower limit is the minimum value of the thickness of the forward coil and the backward coil; the score for judging qualified is 10 points.

6. The method for evaluating the welding results of cold-rolled strip lap welds based on big data as described in claim 1, characterized in that: In step S4, the trends and volatilities of the temperatures of multiple welds are statistically analyzed, and the judgment and scoring criteria for trends and volatilities formulated according to the density functions of the trends and volatilities of the weld temperatures are as follows: If the trend of the temperature is between [-1, 0.2], it gets 20 points; if the trend of the temperature is between [-1.2, -1), (0.2, 0.5], it gets 10 points; if the trend of the temperature is between [-2, -1.2), (0.5, 2], it gets 5 points; if the absolute value of the trend of the temperature exceeds 2, it is NG; If the volatility of the temperature is between [0, 25], it gets 20 points; if the volatility of the temperature is between (25, 40], it gets 10 points; if the volatility of the temperature is between (40, 60], it gets 5 points; if the volatility of the temperature exceeds 60, it is NG.

7. A method for evaluating the welding results of cold-rolled strip lap welds based on big data, as described in claim 1 or 6, characterized in that: In step S4, the temperature distribution of each point in the effective area is evaluated by using the temperature mean ± 3 times the standard deviation: the temperature points falling within the interval of the mean ± 3 times the standard deviation are judged as qualified; those exceeding the interval are judged as unqualified for the temperature of this point; a point deduction system is adopted, with a total of 10 points, and 1 point is deducted for each point judged as unqualified until it is reduced to 0 point.

8. The method for evaluating the welding results of cold-rolled strip lap welds based on big data as described in claim 1, characterized in that: In step S5, the set value is 60 points.

Images