JPH05318141A - Method and device for monitoring resistance welded tube welding and resistance welded tube welding controller - Google Patents

Abstract

PURPOSE:To carry out proper discrimination of a welding state and heat input control by catching a heat generating metal part generated on edge parts in opposition of tube raw material as two pictures, picture-processing the feature quantity based on these pictures to obtain an analysis signal and determining the propriety of the welding state based on this analysis signal. CONSTITUTION:In resistance welded tube welding, masking 21 is performed in the vicinity of a welding point 1a, the heat generating metal part generated extending over edge parts 1b and 1c from the welding point 1a is divided into two visually, the feature quantity of these heat generating metal parts 4a and 4b divided into two is analyzed by a picture processing part 10, the welding state is inferred by an inference part 16 based on the analysis signal S5 and a different welding condition signal S9 and an alarm is raised and display is executed on a monitoring part 17 based on the inference result.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a welding state monitoring method in high frequency welding and a monitoring / controlling apparatus therefor.

[0002]

2. Description of the Related Art An electric resistance welded pipe is manufactured by forming a strip into a cylindrical shape by a molding machine, applying a high-frequency current to both V-shaped edges of the strip to heat and melt it, and press-bonding it with a squib roll.

Welding conditions are reflected in almost proper temperature, proper forming condition, proper material and operation level. here,
By grasping changes in measurable elements, determining whether welding is appropriate, determining the cause of inadequacy, and setting the amount of input electric power in a feed-forward manner, and by comprehensively capturing the changes due to other factors and providing feedback. The basic idea of welding monitoring and heat input control is to correct the problem.

FIG. 7 shows induction type high frequency electric resistance welded pipe welding, and FIG. 8 shows contact type high frequency electric resistance welded pipe welding. In the figure, 3a is a work coil for electromagnetic induction, 3b and 3c are tips (contacts) for contact energization, 2a and 2b are squeeze rolls,
Reference numeral 1 is a pipe material to be welded, 1b and 1c are edges forming a V-shaped gap, 1a is a welding point, and 3d is a high-frequency oscillator. The work coil 3a or the chips 3b and 3c are arranged in front of the squeeze rolls 2a and 2b, so that the pipe blank 1 can be formed by a multi-stage forming roll (not shown).
When a high-frequency current i is applied to the opposite edges 1b, 1c of the V-shaped gap formed on the squeeze roll 2a, the opposite edges 1b, 1c are heated by the high-frequency current and reach the maximum temperature at the welding point 1a. It is pressure welded by 2b.

[0005] In such electric resistance welded pipe welding, a phenomenon in which the welding state has a characteristic appears depending on the amount of heat input, the pipe diameter, and the plate thickness.

A conventional method for monitoring the welding condition in the above-mentioned electric resistance welded pipe welding will be described below.

(1) A method of determining whether the operator is sleeping.

(2) A method of measuring the temperature of a welded portion using a radiation thermometer, in which the total radiant energy is converted into temperature, and the ratio of the energy levels of two specific wavelengths of the total radiant energy is calculated. Method to convert to temperature using.

(3) A method of electrically detecting a change in resonance frequency to determine an excessive heat input amount.

(4) A method of grasping the shape by using the protrusions of the beat after welding.

[0011]

None of the above-mentioned methods is intended to monitor the welding state only by determining whether the heat input (temperature) of the welding state is proper, excessive or excessive.

Even if the heat input is appropriate, if the edge is twisted in the molding state, the edge is shaken, or the pressure (upset) of the squib roll is changed, welding failure may result even if the heat input is appropriate. .. Therefore, there has been a demand for a monitoring method capable of grasping such changes in conditions.

The present invention has been made in view of the above-mentioned problems, and an object thereof is to capture, as two images, a heat-generating metal portion generated at opposing edges of a tube material, and perform image processing on a feature amount based on these images. Then, the analysis signal is obtained, and the suitability of the welding state is determined based on this analysis signal, thereby enabling appropriate determination of the welding state and heat input control.

[0014]

In order to achieve the above-mentioned object, the present invention forms a tube material into a tubular shape having a V-shaped gap, and the opposite edges of the V-shaped gap are formed at the edges thereof. In electric resistance welded pipe welding in which welding is continuously performed at the welding points joined with each other, masking is applied to the upper portion in the vicinity of the welding points to visually divide the heat-generating metal portion generated at each edge into two.
Each brightness level or brightness distribution of the divided heating metal part is captured as an image signal by the image pickup means, and the feature amount including the area of each heating metal part, the trapezoidal axial length, the center of gravity position and the inclination is image-processed and created in advance. It is characterized in that the appropriateness of the welding state is determined based on the determined logic, and the cause of the welding failure is determined based on the feature amount.

[0015]

By masking the vicinity of the welding point by the image pickup means, the heat-generating metal portion generated near the weld point is visually divided into two, and the heat-generating metal portion is divided by the luminance level or the luminance distribution of the two-divided heat-generating metal portion. The image processing unit captures the feature amount of 1., the inference unit infers the welding state by the analysis signal of the image processing unit and the preset determination logic, and the welding state is monitored based on the inference result of the inference unit.

[0016]

Embodiments of the present invention will be described below with reference to FIGS.

FIG. 1 shows a supervisory control method and apparatus for electric resistance welded pipe welding according to an embodiment of the present invention, and 20 is a side view of a welded portion in electric resistance welded pipe welding. FIG. 2 is a plan view of the welded portion 20 as seen from above. In FIGS. 1 and 2, reference numeral 5 denotes a camera equipped with a CCD element, which is arranged at the welding point 1 a of the tube material 1. Reference numeral 6 is an analog / digital converter (A / D converter) for converting the analog image signal S ₁ from the camera 5 into a digital image signal S ₂ , 7 is an image memory for storing the digital image signal S ₂ , and 8 is preset. A setting memory for storing the image pattern, and an arithmetic processing unit (CPU) 9 performs arithmetic processing based on the image memory signal S ₃ of the image memory 7 and the setting signal S ₄ to perform image analysis. These A / D converter 6, image memory 7,
Setting memory 7, setting memory 8 and arithmetic processing unit CPU 9
The image processing unit 10 is configured by the image processing unit 1.
Various controls are executed based on the processed signal of 0.

Reference numeral 11 is a signal correction unit for obtaining an appropriate electric control signal based on the analysis signal S ₅ of the CPU 9.
That is, the signal correction unit 11 includes a correction amount calculation circuit 12,
It is composed of a signal conversion circuit 13 and a signal control circuit 14.

The correction amount calculation circuit 12 inputs the image analysis signal S ₅ from the image processing unit 10, the welding state signal S ₉ and the welding reference set value signal S ₀ and calculates the correction amount to be corrected. The welding state signal S ₉ includes a deviation of the forming center, buckling / wave-like phenomenon, step / lap, and deformation, and such a change in state causes a welding defect. Further, the welding reference value setting signal S ₀ basically includes temperature, material, environment and butt state.

The signal conversion circuit 13 includes a correction amount calculation circuit 12
The correction amount signal S ₆ of is converted into an electric signal, and the signal control circuit 1
4 inputs the power control signal S ₈ to the power control unit 15 based on the electric signal S ₇ of the signal conversion circuit 13. Power control unit 1
5 supplies power to the work coil 3a based on the power control unit signals S _8.

The most characteristic feature of the present invention is that the downstream side of the welding point is masked 21 as shown in FIG. Image processing unit 1 as well as image processing
The reason is that the inference unit 16 that infers the cause of welding failure based on the analysis signal S ₅ analyzed by 0 and the monitoring unit 17 that monitors the welding state based on the inference result of the inference unit 16 are provided.

The inference unit 16 makes an inference by the image processing unit 10 using the features such as the area, the axial length, the peripheral length, and the inclination of the heat-generating metal portions 4a and 4b. The monitoring unit 17 uses an alarm circuit 18 that issues an alarm based on the inference result of the inference unit 16, and an image analysis result of the image processing unit 10 and a display circuit 19 that visually displays based on the inference result of the inference unit 16. It is configured.

Further, a feature of the present invention is that the vibration of the detection end is set to 10 μm or less in order to reduce the noise level and the resolution of 100 pixels or more is provided.

The operation of the apparatus shown in FIG. 1 will be described in more detail. First, the camera 5 equipped with n × m CCD elements in the horizontal × vertical direction corresponds to the welding phenomenon mode of the welding point for each CCD element. The image signal S ₁ is output by scanning as a brightness level at the position (that is, the brightness distribution pattern as a whole). The image signal S ₁ is a luminance signal, which is an electric signal from the light receiving body of the camera 5 composed of n × m CCD elements. This electric signal is converted into a digital signal S ₂ by the A / D converter 6 and then n ×
The image memory 7 is used as a digital amount obtained by analyzing the brightness Cd / M ² for each of the m pixels, for example, into 128 levels.
Stored in. The image data S ₃ in the image memory 7 is CP
Input to U9. The CPU 9 receives the image data S ₃ and performs image processing according to the luminance levels of n × m pixels in the horizontal × vertical direction shown in FIG. 4 to obtain the characteristic amount. The patterns P ₁ and P _{2 in} FIG.
It is shown by the positional relationship in the Y-axis direction.

The pattern illustrated in FIG. 4 corresponds to the welding phenomenon mode shown in FIG. 3 in the vicinity of the welding point 1a, and the portion surrounded by P ₁ in FIG. 4 has the highest brightness level. And corresponds to the heat generating portion 4a in FIG. The portion surrounded by P ₂ in FIG. 4 is the portion having the second highest luminance level and corresponds to the heat generating portion 4b in FIG. On the other hand, the CPU 9, on the other hand, determines the judgment standard, the brightness level and the brightness distribution data S ₄ stored in advance in the setting memory 8.
4 and the pattern of FIG. 4 based on the signal S ₃ from the image memory 7 is compared with the reference data S ₄ and the shape (luminance distribution) and luminance level of the welding state relative to the proper level Whether or not the applied power to the facing edge portion forming the V-seam is appropriate is determined by determining whether or not the level is present, and the analysis signal S ₅ is input to the signal correction unit 11, the inference unit 16 and the monitoring unit 17.

In order for the CPU 9 to compare the brightness level and the brightness distribution data S ₃ based on the welding phenomenon mode captured by the camera 5 with a plurality of reference data S ₄ , for example, one of the simplest methods is described above. The length in the X direction of the portion surrounded by P ₁ in the pattern of FIG. 4 may be compared with the reference length of the reference data. That is, this makes it possible to determine the suitability of the welding state (level) from the length of the portion where the heat generating portions 4a and 4b exist in FIG.

More specifically, the image processing unit 10 is shown in FIG.
As shown in, the equivalent trapezoidal spindle length of the heating metal parts 4a and 4b,
Image processing is performed by finding the equivalent trapezoidal auxiliary axis length and the equivalent trapezoidal tilt angles θ ₁ and θ ₂ . The main axis length, auxiliary axis length of a trapezoid having the same area of the same area and the same second moment, and the direction of the trapezoid are detected and used as a quality determination element for welding, and these are used for monitoring and control.

As shown in FIG. 5, heat-generating metal parts 4a and 4b.
As trapezoids 30a and 30b, and trapezoids 30a, 3
0b of the main axis lengths a ₁ and a ₂ , the sub axis lengths b ₁ and b ₂ and the positions of the centers of gravity G ₁ and G ₂ of the trapezoids 30a and 30b.
Inclination angles θ ₁ and θ ₂ are calculated to obtain the feature amount of the heat generating portions 4a and 4b.

The CPU 9 of the image processing unit 10 determines the heat input state and the molding state based on the areas A ₁ and A ₂ , the centers of gravity G1 and G2, and the deviations (tilt angles) θ ₁ and θ ₂ of the heating metal portions 4a and 4b. To analyze. The CPU 9 uses (A ₁ + A ₂ ) / 2 and (A ₁ −A ₂ ) / (A ₁ +) as area data based on the image data.
A ₂ ) ^1/2 is calculated, and as data on the center of gravity (G ₁ +
G ₂ ) / 2 and | G ₁ −G ₂ | / (G ₁ × G ₂ ) ^1/2 , and (| θ ₁ | − | θ ₂ | / (θ ₁ × θ ₂ ) ^1/2 for inclination. The calculated value is multiplied by the analytic signal S ₅ .

The inference unit 16 is the CPU 9 of the image processing unit 10.
If the binary signals of the equivalent trapezoid main axis length, the equivalent trapezoid auxiliary axis length, the equivalent trapezoid inclination angle, and the center of gravity position of the heating metal parts 4a and 4b calculated by The alarm circuit 18 is caused to generate an alarm. The display circuit 19 displays the inference result of the inference unit and the analysis signal S of the CPU 9.
Based on ₅ , the area, center of gravity, equivalent trapezoid main axis length, equivalent trapezoid auxiliary axis length, and equivalent trapezoid inclination angle of the heating metal parts 4a and 4b are visually displayed. By the display of the display circuit 19, the amount of heat input, the size of the plate thickness, and the speed can be determined by the average value (A ₁ + A ₂ ) / 2 of the areas of the heat-generating metal parts 4a and 4b.
The molding imbalance can be determined by the area difference value (A ₁ −A ₂ ) / (A ₁ + A ₂ ) ^1/2 . The average value (G ₁ + G ₂ ) / 2 corresponding to the movement of the center of gravity can be used to determine the magnitude of the pressing force and the plate width, and the difference in the center of gravity (| G ₁ −G ₂ |) /
The molding state can be determined by (G ₁ × G ₂ ) ^1/2 . The deviation of (| D ₁ |-| D ₂ |) / (D ₁ × D ₂ ) ^1/2 can be used to determine the balance of the molding state, and the inclination can be used to determine the stability of molding, the change in plate width and the wear of rolls be able to.

FIG. 6 shows an example of the simplest operation of the monitoring section 17 and shows an operation state with respect to the image center of gravity × position. In FIG. 6, the curve C ₀ shows the variation of the center of gravity × position within the allowable range, and the curve C ₁ shows the actual change state. The curve C ₂ shows the monitoring alarm signal, the straight line L ₁ corresponds to the monitoring lower limit value and the number of pixels is 350.0, and the straight line L ₂ corresponds to the monitoring upper limit value and the number of pixels is 400.0. The wire L ₁ indicates the suitability of the welding state, and the straight line L ₂ or above indicates the inappropriateness of the welding state. If the difference between L ₂ and L ₁ is large, the balance is lost.

As shown in FIG. 6, the hysteresis width is shown between the straight lines L ₁ and L _2, and the time Ts between the times t ₀ and t ₃ is the slightly increasing processing cycle.

As shown in FIG. 6, the hysteresis width is shown between the straight lines L ₁ and L _2, and the time Ts between the times t ₀ and t ₃ is the image processing cycle. When the center of gravity × position exceeds the monitoring hysteresis band at time t ₃ , the inference unit 16 detects this and activates the alarm circuit 18 to generate an alarm signal.
_{At 5} when the center of gravity × position falls below the monitoring hysteresis band,
The inference unit 16 detects this and causes the alarm circuit 18 to stop operating, and the alarm signal stops. Further, in the monitoring unit 17, the display circuit 19 constantly displays the image modes shown in FIG. 6 in time series. As a result, even if the operator changes, it is possible to always judge the quality of the welding state, and it becomes possible to monitor without being subject to human subjectivity.

The monitoring upper limit value, the monitoring lower limit value, and the monitoring hysteresis width can be set to arbitrary values in the inference unit 16.
In addition, the hysteresis bandwidth is necessary to prevent the alarm signal from turning on and off more than necessary.

In the pass / fail judgment method shown in FIG. 6, an example of the X position of the image center of gravity has been shown.
Equivalent trapezoid main axis length, equivalent trapezoid auxiliary axis length, equivalent trapezoid circumference and center of gravity can be performed.

The present invention is effective when a heat-generating metal portion is generated along both edges of the V throat in high frequency welding, and masking the welding abutting point downstream, two images observed by a CCD camera. To do. When the light emission state detected by the camera is subjected to image processing, the brightness (W / sr / M ² ) value can be measured for each pixel in the data of the captured image. The brightness value is binarized, converted into a monochrome image, and the feature amount of the image is obtained. In this case, the image is one image connected by the V throat (abutting point), but by masking from the abutting point or from the entrance side, the image is divided into two black and white images, and both edges of the V throat are heavenly. Each brightness distribution is captured as an image. Each image is obtained as a feature amount, the amount obtained by averaging each feature amount (representing the overall heating state) and the amount obtained by calculating the difference between each feature amount (for the heating state of both edges (Represents the balance state).

In order to monitor the welding state, the detection point of the CCD camera is set so that the entrance side including the welding abutting point can be viewed. The field of view may be about 5 × 5 mm square. When a small amount of high-frequency heat input is applied to the V throat in which the material is stationary, the abutting point is heated to the maximum temperature in a spot shape, so that the abutting point can be recognized.

Further, the number of pixels is 100 pixels or more on one side edge, the plate movement at the observation / detection end is 100 μm, and external light is shielded to take measures against external light noise.

The correction amount calculation circuit 12 of the signal correction unit 11 calculates the correction amount based on the analysis signal S ₅ , the set value signal S ₀ and the welding condition signal S _9, and outputs the correction amount signal S ₆ to the signal conversion circuit 1.
Lead to 3. The signal conversion circuit 13 converts the correction amount signal S ₆ into an electric signal S ₇ , and inputs the electric signal S ₇ to the signal control circuit 14. The signal control circuit 14 compares the electric signal S ₇ of the signal conversion circuit 13 with the power setting reference signal and compares the electric power control signal S ₈ with the electric power control signal S _8.
Is input to the power control unit 15. The power control unit 15 adjusts the voltage of the power supplied to the work coil 3 according to the power control signal S ₈ .

The welding power can be corrected with higher accuracy as the number of reference values stored in advance in the setting memory 8 at the time of the comparison / judgment processing by the CPU 9 in the image processing unit 10 is increased. Alternatively, instead of increasing the number of reference values, processing by a linearizer may be performed.

[0041]

As described above, the present invention maintains appropriate welding conditions (temperature, forming, operation level) during electric resistance welded pipe, and as described above, changes in the plate thickness of the pipe material to be processed and All the effects of disturbance factors such as transfer speed fluctuations are included, and image processing is performed by recognizing the welding phenomenon mode near the welding point, which is a phenomenon that directly corresponds to the final welding level, as the luminance level and its luminance distribution. By performing this, it is possible to judge (guess) this welding state, and it is possible to make a highly accurate and comprehensive determination of whether the welding state is appropriate or not, rather than simply grasping the heat input amount. Abnormal alarm and display can be performed. In addition, for heat input, proper automatic heat input control can be performed.

Therefore, since high-quality electric resistance welded pipe welding can always be carried out under optimum welding conditions, it is possible to stably manufacture a high quality electric resistance welded pipe and to improve the yield of products. It is difficult to measure the temperature, and for this reason, it is also applied to electric resistance welding of aluminum and copper pipes, which has been difficult to perform accurate automatic heat input control in the past, and contributes to automatic heat input control of welding power. Have an effect.

[Brief description of drawings]

FIG. 1 is a block diagram of a supervisory control device for electric resistance welded pipe welding according to an embodiment of the present invention.

FIG. 2 is a plan view showing a welded portion in electric resistance welded pipe welding.

FIG. 3 is an explanatory view showing a welding mode.

FIG. 4 is a pattern measurement diagram.

FIG. 5 is an explanatory diagram showing a measuring method of a heat-generating metal part.

FIG. 6 is an explanatory diagram showing a quality determination method in the vicinity of a welding point.

FIG. 7 is a perspective view showing induction-type high frequency electric resistance welded pipe welding.

FIG. 8 is a perspective view showing melting type high frequency electric resistance welded pipe welding.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 ... Pipe material 1a ... Welding point 1b, 1c ... Edge part 3a ... Work coil 4a, 4b ... Exothermic metal part 5 ... Camera 6 ... Analog / digital conversion circuit 7 ... Image memory 8 ... Setting memory 9 ... Arithmetic processing part 10 ... Image processing unit 11 ... Signal correction unit 12 ... Correction amount calculation circuit 13 ... Signal conversion circuit 14 ... Signal control circuit 15 ... Power control unit 16 ... Inference unit 17 ... Monitoring unit 18 ... Alarm circuit 19 ... Display circuit 21 ... Masking

Claims (3)

[Claims] 1. Electric resistance welded pipe welding in which a pipe material is formed into a tubular shape having a V-shaped gap, and the opposite edges of the V-shaped gap are continuously welded at welding points where the edges join each other. By masking the upper part in the vicinity of the welding point, the heat-generating metal part generated at each edge is visually divided into two.
Each brightness level or brightness distribution of the divided heating metal part is captured as an image signal by the image pickup means, and the feature amount including the area of each heating metal part, the trapezoidal axial length, the center of gravity position and the inclination is image-processed and created in advance. A method for monitoring electric resistance welded pipe welding, characterized in that the appropriateness of the welding state is determined based on the determined determination logic, and the cause of welding failure is determined based on the feature amount. 2. An electric resistance welded pipe welding apparatus in which a pipe material is formed into a tubular shape having a V-shaped gap, and the opposite edges of the V-shaped gap are continuously welded at their joint points. A masking means which is provided in the upper part near the welding point and visually divides the heat-generating metal portion generated at each of the edges into two, and an image is obtained by detecting each luminance level or luminance distribution of the two-divided heat-generating metal portion. An image pickup unit for obtaining a signal, an image processing unit for analyzing the image signal obtained by the image pickup unit to obtain a luminance distribution, an analysis signal (S ₅ ) for the image processing unit, and a welding condition signal (S ₉ ).
It comprises an inference unit (16) that determines the adequacy of the welding state based on a determination logic created in advance based on the above, and a monitoring unit (17) that monitors the welding state based on the inference signal of the inference unit (16). The monitoring unit (17) includes an alarm circuit (18) for issuing an alarm based on the inference signal of the inference unit (16), and the image processing unit.
Each of the heating metal parts (4a, 4) based on the analysis signal of (10)
A monitoring device for electric resistance welded pipe welding, comprising a display circuit 19 for displaying a discrimination result based on the feature amount of b) and the inference signal of the inference unit (16). 3. In electric resistance welded pipe welding, in which a pipe material is formed into a tubular shape having a V-shaped gap, and opposite edges of the V-shaped gap are continuously welded at welding points where the edges join each other. By masking the upper part in the vicinity of the welding point, the heat-generating metal part generated at each edge is visually divided into two.
The respective brightness levels or brightness distributions of the divided heat-generating metal parts are captured as image signals by the image pickup means, and the feature amount including the area of each heat-generating metal part, the trapezoidal axial length, the position of the center of gravity, and the inclination is captured to obtain an analysis signal (S ₅ ) Is obtained, and the analysis signal (S ₅ ) is corrected and calculated based on the reference setting signal (S ₀ ) and the welding condition signal (S ₉ ) to obtain a correction amount signal (S ₆ ). _A method for controlling electric resistance welded pipe, wherein the heat input amount to the pipe material is controlled based on ₆ ).